The relationship between the CD cotton effect and the absolute configuration of FD-838 and its seven stereoisomers

Article abstract of DOI:10.1021/jo100496f

Three new metabolites having a spiro-heterocyclic γ-lactam core, cephalimysins B-D (1-3), as well as FD-838 (4) were isolated from a culture broth of Aspergillus fumigatus that was originally separated from the marine fish Mugil cephalus. Compounds 1-3 are the diastereomers of 4. Compounds 2 and 3 exhibit an opposite absolute configuration at a spiro carbon to that of other known naturally occurring spiro-heterocyclic γ-lactams. In addition, we succeeded in the chemical transformation of the four natural products (1-4) into their epimers (1′-4′) at C-8 to afford all the stereoisomers of FD-838 (4) with three stereogenic centers. Consequently, the relationship between the absolute configuration at stereogenic centers and the CD Cotton effects for these compounds could be unambiguously established. All of the compounds except 1 moderately inhibited the growth of cultured P388 and HL-60 cell lines.

Full text of DOI:10.1021/jo100496f

pubs.acs.org/joc
The Relationship between the CD Cotton Effect and the Absolute
Configuration of FD-838 and Its Seven Stereoisomers
Takeshi Yamada,* Hiroki Kitada, Tetsuya Kajimoto, Atsushi Numata, and
Reiko Tanaka
Osaka University of Pharmaceutical Sciences, 4-20-1, Nasahara, Takatsuki, Osaka 569-1094, Japan
yamada@gly.oups.ac.jp
Received March 23, 2010
Three new metabolites having a spiro-heterocyclic γ-lactam core, cephalimysins B-D (1-3), as well
as FD-838 (4) were isolated from a culture broth of Aspergillus fumigatus that was originally
separated from the marine fish Mugil cephalus. Compounds 1-3 are the diastereomers of 4. Com-
pounds 2 and 3 exhibit an opposite absolute configuration at a spiro carbon to that of other known
naturally occurring spiro-heterocyclic γ-lactams. In addition, we succeeded in the chemical trans-
formation of the four natural products (1-4) into their epimers (10-40) at C-8 to afford all the
stereoisomers of FD-838 (4) with three stereogenic centers. Consequently, the relationship between
the absolute configuration at stereogenic centers and the CD Cotton effects for these compounds
could be unambiguously established. All of the compounds except 1 moderately inhibited the growth
of cultured P388 and HL-60 cell lines.
Introduction
fungi of marine animals in our search for active antitumor
compounds,4 such as cephalimysin A from A. fumigatus.5
The absolute configuration of most of these γ-lactams was
determined by asymmetric total synthesis;6 however, that of
pseurotin1a was determined by X-ray diffraction analysis of a
dibromo derivative,1a while the absolute configurations of
synerazol2b and cephalimycin A5 were confirmed using the
modified Mosher’s method after the derivation of hydroxy
derivatives.2b,5 Although one of the most effective means of
A variety of natural products with a spiro-heterocyclic
γ-lactam structure have been isolated from diverse micro-
organisms, including (-)-pseurotin A from Psudeurotium
ovalis,1 (þ)-synerazol from Aspergillus fumigatus,2 and azas-
pirene from Neosartorya sp.3 We also reported the structure
of novel spiro-heterocyclic γ-lactams isolated from parasitic
(1) (a) Bloch, P.; Tamm, C.; Bollinger, P.; Petcher, T. J.; Weber, H. P.
Helv. Chim. Acta 1976, 59, 133–137. (b) Weber, H. P.; Petcher, T. J.; Bloch,
P.; Tamm, C. Helv. Chim. Acta 1976, 59, 137–140. (c) Bloch, P.; Tamm, C.
Helv. Chim. Acta 1981, 64, 301–315.
(2) (a) Ando, O.; Satake, H.; Nakajima, M.; Sato, A.; Nakamura, T.;
Kinoshita, T.; Furuya, K.; Haneishi, T. J. Antibiot. 1991, 44, 382–389.
(b) Igarashi, Y.; Yabuta, Y.; Furumai J. Antibiot. 2004, 57, 537–540.
(3) T. Asami, Y.; Kakeya, H.; Onose, R.; Yoshida, A.; Matsuzaki, H.;
Osada, H. Org. Lett. 2002, 4, 2845–2848.
(4) (a) Yamada, T.; Iritani, M.; Doi, M.; Minoura, K.; Ito, Y.; Numata,
A. J. Chem. Soc., Perkin Trans. 1 2001, 3046–3053. (b) Yamada, T.; Iritani,
M.; Minoura, K.; Kawai, K.; Numata, A. Org. Biomol. Chem. 2004, 2131–
2135. (c) Yamada, T.; Iritani, M.; Ohishi, H.; Tanaka, K.; Minoura, K.; Doi,
M.; Numata, A. Org. Biomol. Chem. 2007, 3979–3986. (d) Yamada, T.; Doi,
M.; Shigeta, H.; Muroga, Y.; Hosoe, S.; Numata, A.; Tanaka, R. Tetra-
hedron Lett. 2008, 49, 4192–4195.
(5) Yamada, T.; Imai, E.; Nakatsuji, K.; Numata, A.; Tanaka, R.
Tetrahedron Lett. 2007, 48, 6294–6296.
(6) (a) Hayashi, Y.; Shoji, M.; Yamaguchi, J.; Sato, K.; Yamaguchi, S.;
Mukaiyama, T.; Sakai, K.; Asami, Y.; Kakeya, H.; Osada, H. J. Am. Chem.
Soc. 2002, 124, 12078–12079. (b) Aoki, S.; Oi, T.; Shimizu, K.; Shiraki, R.;
Takao, K.; Tadano, K. Bull. Chem. Soc. Jpn. 2004, 77, 1703–1716.
(c) Hayashi, Y.; Shoji, M.; Yamaguchi, S.; Mukaiyama, T.; Yamaguchi, J.;
Kakeya, H.; Osada, H. Org. Lett. 2003, 5, 2287–2290. (d) Hayashi, Y.; Shoji,
M.; Mukaiyama, T.; Gotoh, H.; Yamaguchi, S.; Nakata, M.; Kakeya, H.;
Osada, H. J. Org. Chem. 2005, 70, 5643–5654. (e) Hayashi, Y.; Sankar, K.;
Ishikawa, H.; Nozawa, Y.; Mizoue, K.; Kakeya, H. Bioorg. Med. Chem. Lett.
2009, 19, 3863–3865. See also: Orellana, A.; Rovis, T. Chem. Commun. 2008,
730–732.
4146 J. Org. Chem. 2010, 75, 4146–4153
Published on Web 05/27/2010
DOI: 10.1021/jo100496f
r
2010 American Chemical Society

Yamada et al.
JOCArticle
FIGURE 1. Structures of natural products from A. fumigatus.
determining the absolute configuration of organic com-
pounds, CD spectroscopy has never been employed to
confirm the absolute configuration of spiro-heterocyclic
γ-lactams.
oxygen-bearing sp3-methine (C-9), two oxygen-bearing
quaternary sp3-carbons (C-5 and C-8), seven sp2-methines
(C-11, C-12, C-19, C-20, C-21, C-22, and C-23), five qua-
ternary sp2-carbons (C-2, C-3, C-10, C-13, and C-18) includ-
ing three oxygen-bearing quaternary carbons (C-2, C-10,
and C-13), two conjugated carbonyl groups (C-4 and C-17),
one amido (C-6 and N-7), and one hydroxy group (9-OH).
Meanwhile, we succeeded in the isolation of all four
diastereoisomers of FD-838, a spiro-heterocyclic γ-lactam
having three stereogenic centers, from culture broth of A.
fumigatus OUPS-T106B-5, which was originally obtained
from the marine fish Mugil cephalus. Moreover, we prepared
a set of four unnatural forms of FD-838 by chemical treat-
ment of FD-838 and its naturally occurring diastereomers.
With all eight stereoisomers in hand, CD spectra could be
analyzed precisely. We herein report the structural determi-
nation of these stereoisomers using CD spectra as well as 1H
and 13C NMR spectra. In addition, we describe their cyto-
toxic activity against the murine P388 leukemia, human HL-
60 leukemia, murine L1210 leukemia, and human KB epi-
dermoid carcinoma cell lines.
FIGURE 2. Selected 1H-1H COSY and HMBC correlations of 1.
Results and Discussion
1H-1H COSY analysis of 1 gave four partial structural
units as shown by bold lines in Figure 2. The connection of
these units with the remaining functional groups was deter-
mined on the basis of HMBC correlations summarized in
Figure 2. The above data revealed that the planar structure
of 1, elucidated as shown in Figure 2, is the same as that of
FD-838 (4), which was identified by comparison with data in
the literature.7 The general features of the 1H and 13C NMR
spectra of 1 closely resembled those of 4 except for a chemical
shift of the proton signal for H-9 (δH 4.87) and the carbon
signals for C-8 (δC 94.9) and C-9 (δC 77.0). In NOESY
experiments, no correlation that demonstrated the difference
in stereochemistry between 1 and 4 was observed.
Bioassay-directed fractionation (cytotoxicity in P388
cells) of an ethyl acetate extract of the culture broth of A.
fumigatus OUPS-T106B-5 as reported previously,5 employ-
ing stepwise a combination of Sephadex LH-20 and silica gel
column chromatographies, followed by reversed-phase
HPLC, afforded cephalimysin B-D (1-3) together with
FD-838 (4) (Figure 1).7
Cephalimysin B (1) had the molecular formula C22H21NO7,
as established from the [M þ H]þ peak in HRFABMS. Its IR
spectrum exhibited bands at 3304, 2932, 1732, 1695, 1621,
1568, and 1509 cm-1, characteristic of a hydroxy group and
an amide. Close inspection of the 1H and 13C NMR spectra
(Table 1) of 1 using DEPT and H-13C correlation spec-
troscopy (HSQC) revealed the presence of one primary
methyl (C-15), one olefinic methyl (C-16), one methoxy
group (8-OCH3), one sp3-hybridized methylene (C-14), one
Cephalimysins C (2) and D (3) were assigned the same
molecular formula as 1 and 4 based on deduction made from
HRFABMS data, and analyses of HMBC correlations con-
firmed that 2 and 3 had the same planar structure as the
1
1
above cephalimysins. The H and 13C NMR spectra of 2
(Table 1) showed features closely resembling those of 4
except for a difference in the chemical shifts of the carbon
signals for C-4 (δC 198.1) and C-5 (δC 86.6). On the other
(7) Mizoue, D.; Okazaki, T.; Hanada, K.; Amamoto, T.; Yamagashi, M.;
Omura, S. Eur. Pat. Appl. EP 216607, J. P. Appl. 205278, 1987; Chem.
Abstr., 1987, 107, 132627x.
J. Org. Chem. Vol. 75, No. 12, 2010 4147

Yamada et al.
JOCArticle
TABLE 1. NMR Data for 1-4 in CDCl3
1
2
3
4
a
a
δH
a
δH
a
position
δH
J/Hz
δC
J/Hz
δC
J/Hz
δC
δH
J/Hz
δC
l
2
3
4
5
170.1 (s)
108.1 (s)
195.8 (s)
87.9 (s)
171.2 (s)
109.3 (s)
198.1 (s)
86.6 (s)
172.8 (s)
109.3 (s)
199.5 (s)
85.6 (s)
172.5 (s)
107.8 (s)
195.6 (s)
89.6 (s)
6
168.0 (s)
167.1 (s)
167.9 (s)
166.2 (s)
7
8
9
10
11
12
13
14
15
16
7.57 s
4.87 d
7.42 s
4.59 d
7.40 s
4.90 d
7.44 s
4.69 d
94.9 (s)
77.0 (d)
143.1 (s)
117.9 (d)
107.7 (d)
163.8 (s)
21.8 (t)
92.4 (s)
73.4 (d)
143.4 (s)
117.7 (d)
107.6 (d)
163.8 (s)
21.8 (t)
96.5 (s)
79.1 (d)
143.3 (s)
118.9 (d)
107.9 (d)
164.5 (s)
21.9 (t)
91.5 (s)
74.2 (d)
143.3 (s)
118.3 (d)
107.9 (d)
163.7 (s)
21.8 (t)
3.9
12.4
4.1
12.7
7.02 d
6.19 d
3.4
3.4
7.10 d
6.26 d
3.4
3.4
7.17 d
6.29 d
3.4
3.4
7.04 d
6.23 d
3.4
3.4
2.74 q
1.28 t
2.03 s
7.6
7.6
2.79 q
1.32 t
2.02 s
7.6
7.6
2.80 q
1.32 t
2.04 s
7.6
7.6
2.75 q
1.28 t
2.01 s
7.6
7.6
11.7 (q)
6.2 (q)
11.7 (q)
6.2 (q)
11.7 (q)
6.0 (q)
11.8 (q)
6.2 (q)
17
18
19
20
21
22
23
8-OCH 3
9-OH
193.3 (s)
133.9 (s)
129.8 (d)
128.6 (d)
134.0 (d)
128.6 (d)
129.7 (d)
51.4 (q)
193.8 (s)
133.0 (s)
130.6 (d)
128.6 (d)
134.5 (d)
128.6 (d)
130.6 (d)
51.7 (q)
192.4 (s)
133.9 (s)
129.4 (d)
128.8 (d)
134.0 (d)
128.8 (d)
129.4 (d)
51.4 (q)
194.6 (s)
132.4 (s)
130.6 (d)
128.7 (d)
134.5 (d)
128.7 (d)
130.6 (d)
51.6 (q)
8.26 d
7.48 t
7.61 t
7.48 t
8.26 d
3.35 s
2.98 d
8.7
7.6
7.6
7.6
8.7
8.32 d
7.48 t
7.63 t
7.48 t
8.32 d
3.37 s
3.37 d
8.5
7.7
7.7
7.7
8.5
8.23 d
7.49 t
7.62 t
7.49 t
8.23 d
3.30 s
5.36 d
8.5
7.6
7.6
7.6
8.5
8.32 d
7.49 t
7.64 t
7.49 t
8.32 d
3.40 s
4.13 d
8.5
7.6
7.6
7.6
8.5
3.9
12.4
4.1
12.7
a1H chemical shift values (δ ppm from SiMe4) followed by multiplicity and then the coupling constants (J /Hz). Numbers in parentheses indicate the
proton coupling with that position.
FIGURE 3. CD spectra of 4, 6, and 8.
hand, the difference in chemical shift between 3 [the proton
signal: H-9 (δH 4.90); the carbon signals: C-4 (δC 199.5), C-5
(δC 85.6), C-8 (δC 96.5) and C-9 (δC 79.1)] and 4 [the proton
signal: H-9 (δH 4.69); the carbon signals: C-4 (δC 195.6), C-5
(δC 89.6), C-8 (δC 91.5) and C-9 (δC 74.2)] was the same as
that of both 1 and 2 as described above.
The spectroscopic study suggested 1-3 to be the diastereo-
mers of FD-838 (4), whose absolute configuration was
recently confirmed by enantioselective synthesis.6e Thus,
assignment of the CD spectrum of FD-838 (4) (Figure 3)
was attempted. Compound 4 was reduced with sodium
borohydride to afford 5a,b (20.2%), in which the C-4 and
C-17 carbonyl groups were reduced to alcohols, 6a,b
(39.9%), and 7 (7.8%), where the C-17 carbonyl group was
converted into a benzyl alcohol moiety (Scheme 1).
spectra of which were identical with those of 3. On the basis of
these results, 6a,b and 7 having the same planar structure
were identified as stereoisomers at the C-8 position, i.e., the
8β-OCH3 and 8R-OCH3 derivatives, respectively (vide infra).
A plausible mechanism forthis reaction is shown in Scheme1.
The 8-OCH3 group was eliminated as a result of the abstrac-
tion of an amide proton by the methoxide anion, and then,
the methoxide anion attacked C-8 as a nucleophile, accom-
panying the synchronous reduction of the carbonyl group at
C-17. In principle, the CD spectrum of 6a or 6b should show
the absorption band attributable to the enone moiety, while
that of 8 would exhibit the absorption band ascribed to the
benzophenone moiety of 4 (Figure 3). Since, in practice, 8 was
not obtained in the reduction of 4, the CD spectrum of 8 was
determined by subtracting the CD spectrum of 6 from that
of 4. In conclusion, the positive (Δε352 þ3.7) and negative
(Δε320 -10.1) Cotton effects in the CD spectrum of 4 indicate
S configuration at C-5 and C-8 (Figure 3).
The reduction products 6a,b could be reconverted to 4 by
oxidation with manganese dioxide in dichloromethane, while
the oxidation of 7 gave the product 30, the 1H and 13C NMR
4148 J. Org. Chem. Vol. 75, No. 12, 2010

Yamada et al.
JOCArticle
SCHEME 1. Plausible Mechanism for the Reduction of 4
Compound 30, the oxidation product of 7, was identical to
the natural product 3 in terms of IR, 1H, and 13C NMR data;
however, its specific rotation and CD spectrum (Figure 4)
were the reverse of those of 3. This finding revealed that 3 was
the enantiomer of 30, being the diastereomer of 4 at C-5 and
C-9. Next, as expected, treatment of 4 with conc. H2SO4 in
MeOH also gave the product 30 at a constant ratio (14.6%)
as shown in Scheme 2. The same acid treatment was applied
to 3 to produce 40, which was identified as the enantiomer of
4 on the basis of accordance of all spectroscopic data expect
the CD data (Figure 4). The above evidence revealed the
absolute configuration of 3 and 40.
Split Cotton effects (λmax ∼350 and 320 nm) were observed
for the 5S,8S isomer 4, i.e., positive and negative Cotton
FIGURE 4. CD spectra of 4, 40, 3, and 30.
J. Org. Chem. Vol. 75, No. 12, 2010 4149

Yamada et al.
JOCArticle
SCHEME 2. Acid-Catalyzed Treatment of Compound 4
effects. The 5R,8R isomer 40 showed negative and positive
Cotton effects, while the 5R,8S isomer 3 and the 5S,8R
isomer 30 showed a negative Cotton effect and a positive
Cotton effect (λmax ∼345 nm), respectively (Figure 4).
Compound 1 showed a CD spectrum resembling that of 4
(Figure 5). Since the rule concerning the Cotton effect did not
involve the configuration at C-9, the similarity of the CD
spectra suggested 1 to be an epimer of 4 at C-9; i.e., 1 pos-
sessed the 5S,8S,9S absolute configuration. Treatment of 1
with concd H2SO4 in MeOH gave compound 20 (Scheme 3),
which was identical with 2 in all spectroscopic data expect the
CD data (Figure 5) and had a [R]D value opposite that of 2.
This evidence showed that the natural product 2 was the
enantiomer of the reaction product 20 obtained from 1. As
expected, the epimerization at C-8 of 2 with concd H2SO4
gave the product 10, the enantiomer of 1. Thus, the absolute
configurations of 10, 2, and20 could be deduced as (5R,8R,9R),
(5R,8S,9R), and (5S,8R,9S), respectively.
We succeeded in the isolation of all stereoisomers of FD-
838 (4) including four reaction products and, therefore,
could establish the relationship between the absolute configu-
rations at C-5 and C-8 in the spirofuranone-lactam skeleton
and the CD Cotton effects. In addition, we found that the
coupling constants of H-9 and 9-OH in the 1H NMR spectra
FIGURE 5. CD spectra of 1, 10, 2, and 20.
4150 J. Org. Chem. Vol. 75, No. 12, 2010

Yamada et al.
JOCArticle
SCHEME 3. Acid Treatment of 1 and 2
FIGURE 6. Hydrogen-bond interactions in 4.
TABLE 2. Cytotoxicity Assay against P388, HL-60, L1210, and KB
Cells
leukemia, and the human KB epidermoid carcinoma cell lines.
All the compounds except 1exhibited moderate activity against
the P388 and HL-60 cell lines (Table 2). The cancer cell growth
inhibitory properties did not imply a structure-activity rela-
tionship, and so molecular target screening for inhibitory
activities of histone deacetylase, protein kinase, telomerase,
and fenecyltransferase are in progress.
IC50 (μM)a
cell line
P388
cell line
HL-60
cell line
L1210
cell line
KB
compd
1
2
3
4
10
20
30
40
>200
53.5
51.1
56.0
53.5
60.8
70.6
58.4
2.5
>200
58.4
48.7
60.8
56.0
51.1
53.5
60.8
2.2
>200
>200
>200
70.5
>200
184.2
150.9
111.9
2.1
>200
>200
>200
73.0
>200
>200
>200
>200
7.7
Conclusion
All diastereomers of FD-838 (4), referred to as cephali-
mysins B-D (1-3), were obtained from the broth of A.
fumigatus. Since acid treatment of 1-4 led to the epimer-
ization of the C-8 acetal carbon to afford all stereoisomers
of 4, the CD cotton effects could be studied comprehen-
sively. The results concerning the relationship between CD
spectra and spiro-hetero γ-lactams will be valuable for
determining the absolute configuration of related com-
pounds.
5-fluorouracilb
aDMSO was used as vehicle. bPositive control.
show the orientations of 9-OH and 8-OCH3, i.e., 9-OH
oriented cis to 8-OCH3 for a large coupling constant (J =
12 Hz) and trans to 8-OCH3 for a small coupling constant
(J = 4 Hz). This would occur by a hydrogen bond, which
held the conformation of 9-hydroxy as shown in Figure 6.
Hayashi and co-workers had reported that the hydrogen
bond prevented the racemization at C-9 in the last step, the
elimination of the protecting group, of the synthesis process
of synerazol.6d Such findings are useful for examining the
stereochemistry of spirofuranone-lactams.
Experimental Section
General Methods. NMR spectra were recorded at 27 °C with
TMS as internal reference. Liquid chromatography over silica
gel (mesh 230-400) was performed at medium pressure and
detected with TLC [the solvent system CH2Cl2-MeOH (90:
10)], and compounds were viewed under a UV lamp, sprayed
with 10% H2SO4, and heated. ODS HPLC was detected with a
differential refractometer.
As a primary screening for antitumor activity, cancer cell
growth inhibitory properties of the natural products 1-4 and
their epimers 10-40 were examined using the murine P388
leukemia, the human HL-60 leukemia, the murine L1210
Culture and Isolation of Metabolites. A strain of A. fumigatus
was initially isolated from the marine fish M. cephalus captured
J. Org. Chem. Vol. 75, No. 12, 2010 4151

Yamada et al.
JOCArticle
in Katsuura Bay, Japan, in October 2000. The fish was disin-
fected with EtOH and its gastrointestinal tract applied to the
surface of nutrient agar layered in a Petri dish. Serial transfers of
one of the resulting colonies provided a pure strain of A.
fumigatus. The fungal strain was cultured at 27 °C for 6 weeks
in a liquid medium (75 L) containing 1% soluble starch and
0.1% casein in 50% artificial seawater adjusted to pH 7.4. The
culture was filtered under suction, and the culture filtrate was
extracted three times with EtOAc. The combined extracts
were evaporated in vacuo to afford a mixture of crude metabo-
lites (18.8 g) that exhibited cytotoxicity against P388 cell line
(IC50 < 1 μg/mL). The EtOAc extract was passed through a
Sephadex LH-20 column using CHCl3-MeOH (1: 1) as the
eluent. The second fraction (14.6 g), exhibiting strong activity,
was chromatographed on a silica gel column with CHCl3-
MeOH gradient as the eluent. The 100% CHCl3 eluate (1.2 g)
was purified by ODS HPLC using MeOH-H2O (70: 30) as the
eluent to afford fraction 1 (20.3 mg) and fraction 2 (100.7 mg)
exhibiting cytotoxicity. Fraction 1 was further purified by ODS
HPLC using MeCN-H2O (45: 55) as the eluent to afford 1
(7.4 mg) and 2 (8.2 mg). Fraction 2 was further purified by ODS
HPLC using MeCN-H2O (45: 55) as the eluent to afford 4
(64.0 mg) and 3 (1.1 mg).
(rel int) 415 ([M]þ, 55.6); HRFABMS [M]þ m/z 415.1639
(C22H25NO7, calcd [M]þ 415.1631); H NMR δ ppm (CDCl3)
1
1.23 (3H, t, J=7.5 Hz, H-15), 1.98 (3H, s, H-16), 2.04 (1H, d, J=
11.8 Hz, 4-OH), 2.45 (1H, d, J=4.2 Hz, 17-OH), 2.61 (1H, d, J=
7.8 Hz, 9-OH), 2.66 (2H, q, J = 7.5 Hz, H-14), 3.35 (3H, s,
8-OCH3), 4.63 (1H, d, J=7.8 Hz, H-9), 4.87 (1H, d, J=4.2 Hz,
H-17), 5.10 (1H, d, J =11.8 Hz, H-4), 6.02 (1H, d, J =3.0 Hz,
H-12), 6.47 (1H, d, J = 3.0 Hz, H-11), 6.68 (1H, br s, H-7),
7.37-7.47 (5H, m, Ar-H).
5b: pale yellow oil; [R]22D -150.5 (c 0.09, EtOH); IR (liquid)
ν
max 3376, 2939, 1721, 1612, 1563, 1532 cm-1; UV (EtOH) λmax
(log ε) 220 (3.97), 274 (3.95), 335 (3.81) nm; FABMS m/z (rel int)
415 ([M]þ, 52.6); HRFABMS [M]þ m/z 415.1629 (C22H25NO7,
calcd [M]þ 415.1631); H NMR δ ppm (CDCl3) 1.22 (3H, t,
1
J = 7.5 Hz, H-15), 1.98 (3H, s, H-16), 2.09 (1H, d, J = 11.8
Hz, 4-OH), 2.58 (1H, br s, 17-OH), 2.65 (2H, q, J = 7.5 Hz,
H-14), 2.74 (1H, d, J = 7.6 Hz, 9-OH), 3.50 (3H, s, 8-OCH3),
4.39 (1H, d, J=7.6 Hz, H-9), 4.96 (1H, d, J=4.2 Hz, H-17), 5.15
(1H, d, J = 11.8 Hz, H-4), 5.99 (1H, d, J = 3.0 Hz, H-12), 6.40
(1H, d, J=3.0 Hz, H-11), 6.65 (1H, br s, H-7), 7.36-7.43 (5H, m,
Ar-H).
6a: pale yellow oil; [R]22 -78.4 (c 0.10, EtOH); IR (liquid)
D
max 3467, 2940, 1730, 1705, 1614, 1567, 1508 cm-1; UV (EtOH)
ν
Cephalimysin B (1): pale yellow oil; [R]22 þ129.7 (c 0.09,
λmax (log ε) 216 (4.00), 284 (3.90), 317 (3.98), 355 (3.95) nm;
FABMS m/z (rel int) 414 ([M þ H]þ, 31.1); HRFABMS [M þ
H]þ m/z 414.1561 (C22H24NO7, calcd [M þ H]þ 414.1553); 1H
NMR δ ppm (CDCl3) 1.30 (3H, d, J=7.4 Hz, H-15), 1.97 (3H, s,
H-16), 2.40 (1H, br s, 17-OH), 2.77 (2H, q, J = 7.4 Hz, H-14),
3.20 (3H, s, 8-OCH3), 3.61 (1H, d, J=11.3 Hz, 9-OH), 4.85 (1H,
br s, H-17), 4.97 (1H, d, J=11.3 Hz, H-9), 6.24 (1H, d, J=3.6
Hz, H-12), 6.60 (1H, br s, H-7), 7.09 (1H, d, J=3.6 Hz, H-11),
7.38-7.44 (3H, m, Ar-H), 7.46-7.49 (2H, m, Ar-H); 13C NMR
δ ppm (CDCl3) 6.1 (C-16), 11.8 (C-15), 21.8 (C-14), 52.6
(8-CH3), 73.4 (C-9), 74.3 (C-17), 87.5 (C-5), 91.6 (C-8), 107.7
(C-12), 108.0 (C-3), 118.0 (C-11), 127.2 (C-19, C-23), 128.7
(C-20, C-22), 129.0 (C-21), 137.8 (C-18), 143.6 (C-10), 163.5
(C-13), 167.5 (C-6), 172.3 (C-2), 196.0 (C-4).
D
EtOH); IR (liquid) νmax 3304, 2932, 1732, 1695, 1621, 1568,
1509 cm-1; UV (EtOH) λmax (log ε) 210 (3.85), 249 (3.95), 318
(3.85), 355 (3.85) nm; NMR data, see Table 1 and Table S1
(Supporting Information); FABMS m/z (rel int) 412 ([M þ H]þ,
52.2), 380 ([M - OCH3]þ, 100); HRFABMS m/z 412.1394 [M þ
H]þ (calcd for C22H22NO7 412.1397); CD (c 2.37 ꢀ 10-4 M,
EtOH) λ (Δε) 386 (0), 352 (5.5), 337 (0), 318 (-7.7), 283 (0) nm.
Cephalimysin C (2): pale yellow oil; [R]22 -209.2 (c 0.09,
D
EtOH); IR (liquid) νmax 3347, 2933, 1724, 1704, 1617, 1566, 1507
cm-1; UV (EtOH) λmax (log ε) 212 (3.79), 249 (3.88), 318 (3.79),
358 (3.76) nm; NMR data, see Table 1 and Table S2 (Supporting
Information); FABMS m/z (rel. int. %) 412 ([MþH]þ, 100), 380
([M-OCH3]þ, 13.2); HRFABMS m/z 412.1395 [MþH]þ (calcd
for C22H22NO7: 412.1397); CD (c 2.90ꢀ10-4 M, EtOH) λ (Δε)
391 (0), 346 (-12.7), 296 (0) nm.
6b: pale yellow oil; [R]22D þ145.6 (c 0.10, EtOH); IR (liquid)
ν
max 3418, 2941, 1730, 1698, 1616, 1568, 1509 cm-1; UV (EtOH)
Cephalimysin D (3): pale yellow oil; [R]22 -236.6 (c 0.11,
λmax (log ε) 215 (3.86), 284 (3.75), 312 (3.85), 357 (3.82) nm;
D
FABMS m/z (rel int) 414 ([M þ H]þ, 17.9); HRFABMS [M þ
EtOH); IR (liquid) νmax 3289, 2937, 1732, 1691, 1612, 1563,
1507 cm-1; UV (EtOH) λmax (log ε) 210 (3.85), 248 (3.95), 315
(3.87), 356 (3.83) nm; NMR data, see Table 1 and Table S3
(Supporting Information); FABMS m/z (rel int) 412 ([M þ H]þ,
100), 380 ([M - OCH3]þ, 44.2); HRFABMS m/z 412.1399 [M þ
H]þ (calcd for C22H22NO7 412.1397); CD (c 2.55 ꢀ 10-4 M,
EtOH) λ (Δε) 389 (0), 344 (-10.2), 293 (0) nm.
1
H]þ m/z 414.1554 (C22H24NO7, calcd [M þ H]þ 414.1553); H
NMR δ ppm (CDCl3) 1.27 (3H, d, J=7.7 Hz, H-15), 1.95 (3H, s,
H-16), 2.73 (2H, q, J=7.7 Hz, H-14), 2.93 (1H, br s, 17-OH), 3.52
(1H, d, J=11.0 Hz, 9-OH), 3.58 (3H, s, 8-OCH3), 4.64 (1H, d, J=
11.0 Hz, H-9), 4.89 (1H, br s, H-17), 6.20 (1H, d, J = 3.6 Hz,
H-12), 7.01 (1H, d, J=3.6 Hz, H-11), 7.05 (1H, br s, H-7), 7.36 -
7.42 (5H, m, Ar-H); 13C NMR δ ppm (CDCl3) 6.1 (C-16), 11.7
(C-15), 21.7 (C-14), 52.8 (8-CH3), 73.8 (C-9), 75.9 (C-17), 88.0
(C-5), 91.5 (C-8), 107.6 (C-12), 108.0 (C-3), 118.0 (C-11), 127.2
(C-19, C-23), 128.5 (C-20, C-22), 128.8(C-21), 137.6 (C-18), 143.5
(C-10), 163.6 (C-13), 168.1 (C-6), 172.2 (C-2), 195.8 (C-4).
FD-838 (4): pale yellow oil; [R]22D þ40.5 (c 0.10, EtOH); IR
(liquid) νmax 3433, 2921, 1734, 1681, 1619, 1566, 1506 cm-1; UV
(EtOH) λmax (log ε) 210 (3.82), 252 (3.92), 319 (3.86), 356
(3.83) nm; NMR data, see Table 1; FABMS m/z (rel int) 412
([M þ H]þ, 100), 380 ([M - OCH3]þ, 32.5); HRFABMS m/z
412.1398 [M þ H]þ (calcd for C22H22NO7 412.1397); CD
(c 2.47 ꢀ 10-4 M, EtOH) λ (Δε) 385 (0), 352 (3.7), 343 (0), 320
(-10.1), 290 (0) nm.
7: pale yellow oil; [R]22 þ52.9 (c 0.11, EtOH); IR (liquid)
D
max 3348, 2939, 1729, 1698, 1625, 1569, 1509 cm-1; UV (EtOH)
ν
λmax (log ε) 212 (3.92), 284 (3.79), 316 (3.93), 358 (3.91) nm;
FABMS m/z (rel int) 414 ([M þ H]þ, 100); HRFABMS [M þ
H]þ m/z 414.1554 (C22H24NO7, calcd [M þ H]þ 414.1553); 1H
NMR δ ppm (CDCl3) 1.31 (3H, d, J=7.6 Hz, H-15), 2.05 (3H, s,
H-16), 2.78 (2H, q, J = 7.6 Hz, H-14), 3.35 (1H, br s, 17-OH),
3.37 (3H, s, 8-OCH3), 4.52 (1H, br s, 9-OH), 5.02 (1H, s, H-9),
5.44 (1H, br s, H-7), 5.90 (1H, br s, H-17), 6.26 (1H, d, J =
3.7 Hz, H-12), 7.11 (1H, d, J=3.7 Hz, H-11), 7.35-7.40 (3H, m,
Ar-H), 7.52-7.44 (2H, m, Ar-H); 13C NMR δ ppm (CDCl3) 6.2
(C-16), 11.8 (C-15), 21.8 (C-14), 49.3 (8-CH3), 74.7 (C-9), 76.1
(C-17), 89.7 (C-5), 91.2 (C-8), 107.8 (C-12), 108.1 (C-3), 118.3
(C-11), 128.3 (C-19, C-23), 128.5 (C-20, C-21, C-22), 137.0
(C-18), 143.3 (C-10), 163.8 (C-13), 165.4 (C-6), 172.2 (C-2),
196.4 (C-4).
Reduction of FD-838 (4) to 5-7. NaBH4 (4.0 mg) was added to
a MeOH solution (1.5 mL) of FD-838 (4) (41.2 mg, 0.1 mmol),
and the reaction mixture was stirred at room temperature for 30
min. The mixture was diluted with water and extracted with
CHCl3, and the CHCl3 was evaporated under reduced pressure.
The residue was purified by HPLC using MeCN-H2O (45: 55)
as the eluent to afford fraction 1 (20.5 mg) and 6 (6a: 6.3 mg
and 6b: 10.2 mg). Fraction 1 was purified by HPLC using
MeCN-H2O (35: 65) as the eluent to afford 5 (5a: 5.1 mg and
5b: 3.3 mg) and 7 (3.8 mg).
5a: pale yellow oil; [R]22D -158.6 (c 0.10, EtOH); IR (liquid)
ν
λ
max 3433, 2938, 1721, 1614, 1619, 1530, 1508 cm-1; UV (EtOH)
max (log ε) 216 (3.96), 274 (3.97), 334 (3.71) nm; FABMS m/z
4152 J. Org. Chem. Vol. 75, No. 12, 2010

Yamada et al.
JOCArticle
Oxidation of 6a to 4. MnO2 (4.4 mg) was added to a CH2Cl2
solution (4.0 mL) of the reduction product 6a (2.5 mg, 0.005
mmol), and the reaction mixture was refluxed overnight. The
solution of the mixture in CH2Cl2 was evaporated under redu-
ced pressure. The residue was purified by HPLC using MeCN-
H2O (45: 55) as the eluent to afford 4 (0.3 mg, yield 15.0%).
Oxidation of 6b to 4. Using the same procedure as above with
6a, a solution of 6b (1.8 mg, 0.004 mmol) in CH2Cl2 (4.0 mL) was
treated with MnO2 (3.9 mg) and purified by HPLC using MeCN-
H2O (45: 55) as the eluent to afford 4 (0.2 mg, yield 12.1%).
Oxidation of 7 to 30. Using the same procedure as above with
compound 6a, a solution of 7 (3.5 mg, 0.009 mmol) in CH2Cl2
(4.0 mL) was treated with MnO2 (7.4 mg) and purified by HPLC
using MeCN-H2O (45:55) as the eluent to afford 30 (0.7 mg,
yield 20.3%). The optical rotation and CD data of 30 are
described later.
Transformation of 1 to 20. To a solution of cephalimysin B (1)
(18.5 mg, 0.05 mmol) in MeOH (1.0 mL) was added concd
H2SO4 (0.01 mL), and the reaction mixture was left at room
temperature for 8 h. The mixture was diluted with water and
extracted with CHCl3, the extract was evaporated under re-
duced pressure, and then the residue was purified by HPLC
using MeCN-H2O (45: 55) as the eluent to afford 1 (8.0 mg) and
20 (1.2 mg).
purified by HPLC using MeCN-H2O (45:55) as the eluent to
afford 3 (0.15 mg) and 40 (0.05 mg).
40: pale yellow oil; [R]22D -39.7 (c 0.05, EtOH); CD (c 1.22ꢀ
10-4 M, EtOH) λ (Δε) 386 (0), 353 (-3.8), 343 (0), 321 (9.7), 290
(0) nm.
Assay for Cytotoxicity. Cytotoxic activities of cephalimysins
B-D (1-3) and FD-838 (4) and epimers (10-40) were examined
with the 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazo-
lium bromide (MTT) method. P388, HL-60, L1210, and KB
cells were cultured in Eagle’s Minimum Essential Medium (10%
fetal calf serum) at 37 °C in 5% CO2. The test material was
dissolved in DMSO to give a concentration of 10 mM, and the
solution was diluted with the Essential Medium to yield con-
centrations of 200, 20, and 2 μM, respectively. Each solution was
combined with each cell suspension (1 ꢀ 105 cells/mL) in the
medium, respectively. After incubation at 37 °C for 72 h in 5%
CO2, grown cells were labeled with 5 mg/mL of MTT in
phosphate-buffered saline (PBS), and the absorbance of for-
mazan dissolved in 20% sodium dodecyl sulfate (SDS) in 0.1 N
HCl was measured at 540 nm with a microplate reader. Each
absorbance value was expressed as percentage relative to that of
the control cell suspension that was prepared without the test
substance using the same procedure as that described above. All
assays were performed three times, semilogarithmic plots were
constructed from the averaged data, and the effective dose of the
substance required to inhibit cell growth by 50% (IC50) was
determined.
20: pale yellow oil; [R]22D þ201.7 (c 0.12, EtOH); CD (c 2.99ꢀ
10-4 M, EtOH) λ (Δε) 385 (0), 345 (9.7), 298 (0) nm.
Transformation of 2 to 10. Using the same procedure as above
with 1, a solution of cephalimysin C (2) (10.0 mg, 0.02 mmol) in
MeOH (1.0 mL) was treated with concd H2SO4 (0.01 mL) and
purified by HPLC using MeCN-H2O (45: 55) as the eluent to
afford 2 (1.2 mg) and 10 (0.4 mg).
Acknowledgment. We thank Dr. T. Ito (National Institute
of Technology and Evaluation, Biological Resource Center)
and Dr. Endo (Kanazawa University) for identification of
the fungal strain and for supply of the cancer cells, respectively.
We are grateful to Ms. M. Fujitake and Dr. K. Minoura of this
university for MS and NMR measurements, respectively. This
study was supported by a Grant-in-aid for High Technology
from the Ministry of Education, Culture, Sports, Science and
Technology, Japan.
10: pale yellow oil; [R]22D -133.6 (c 0.08, EtOH); CD (c 1.50ꢀ
10-4 M, EtOH) λ (Δε) 389 (0), 352 (-6.0), 333 (0), 315 (5.3), 284
(0.4) nm.
Transformation of 4 to 30. Using the same procedure as above
with 1, a solution of FD-838 (4) (12.4 mg, 0.03 mmol) in MeOH
(1.0 mL) was treated with concd H2SO4 (0.01 mL) and purified
by HPLC using MeCN-H2O (45: 55) as the eluent to afford 4
(2.1 mg) and 30 (1.8 mg).
30: pale yellow oil; [R]22D þ233.4 (c 0.16, EtOH); CD (c 3.19ꢀ
10-4 M, EtOH) λ (Δε) 388 (0), 346 (10.7), 292 (0) nm.
Transformation of 3 to 40. Using the same procedure as above
with 1, a solution of cephalimysin D (3) (0.8 mg, 0.002 mmol) in
MeOH (1.0 mL) was treated with concd H2SO4 (0.01 mL) and
Supporting Information Available: Copies of spectra and
chromatograms for these metabolites and the reaction products.
This material is available free of charge via the Internet at http://
pubs.acs.org.
J. Org. Chem. Vol. 75, No. 12, 2010 4153