Structural corrections of photinides A, B and their novel derivatives

Article abstract of DOI:10.1016/j.tet.2012.06.108

Stereochemistries of the C2C8 double bond in photinides A and B were corrected to be reversed forms. The present studies also revised their absolute configurations by comparing the CD spectra of the degradation product with the corresponding synthetic sample. Discosia sp. SH 125 produced novel derivatives photinides X and Y, of which stereochemistry was established by NMR and CD analyses as well as theoretical calculations of the CD spectra.

Full text of DOI:10.1016/j.tet.2012.06.108

Tetrahedron 68 (2012) 7991e7996
Contents lists available at SciVerse ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Structural corrections of photinides A, B and their novel derivatives
,
Ryoko Yasumura a, Kazuaki Tanaka a, Tatsuo Nehira b, Masaru Hashimoto a
*
a Department of Agriculture and Bioscience, Hirosaki University, 3-Bunkyo-cho, Hirosaki 036-8561, Japan
b Graduate School of Integrated Arts and Sciences, Hiroshima University, 1-7-1, Kagamiyama, Higashi-Hiroshima 739-8521, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 5 June 2012
Received in revised form 28 June 2012
Accepted 29 June 2012
Available online 7 July 2012
Stereochemistries of the C2]C8 double bond in photinides A and B were corrected to be reversed forms.
The present studies also revised their absolute configurations by comparing the CD spectra of the
degradation product with the corresponding synthetic sample. Discosia sp. SH 125 produced novel
derivatives photinides X and Y, of which stereochemistry was established by NMR and CD analyses as
well as theoretical calculations of the CD spectra.
Ó 2012 Elsevier Ltd. All rights reserved.
Keywords:
Photinides
Stereochemical correction
Circular dichroism
Chemical shift calculations
Novel natural isomers
1. Introduction
In the course of our studies exploring metabolites from fungi
with unique ecologies in northern Japan,1e4 we found that Discosia
sp. SH 125 produces photinides A and B, which had been reported
by Che et al. in 2009. Our studies revealed structural errors in their
C2]C8 double bonds as well as the absolute configurations to re-
vise their structures as shown in Fig. 1. We also isolated their di-
astereomers photinides X (3) and Y (4) as their novel derivatives
from the same source. Combination analyses of the CD spectral
comparisons of 1 with 3 and their theoretical calculations led us to
establish their absolute configurations. This paper reports the detail
of these studies.
In our screening with Cochlibolus miyabeanus,5 we found a po-
tent growth inhibition in the EtOAc extracts from the culture broth
of Discosia sp. SH 125. This fungus had been isolated by Tanaka, one
of the present authors, from fallen leaves of Castanea crenata at
Aomori prefecture, Japan in 2003. Chromatographic separation
provided a mixture of photinides A, B, X, and Y (1e4, 650 mg) from
2 L of the culture broth. The ODS HPLC employing LiChrospherÒ RP-
Fig. 1. Structures of photinides A, B, X and Y, as well as the old structures of photinides
A and B.
later, only part of the crude sample was subjected to the HPLC
purification, and used for structural and biological studies.
18e (4.0
f
ꢀ250 mm) separated them (1: tR¼37 min, 3: tR¼40 min,
4, tR¼43 min, and 2: tR¼56 min) by eluting with MeOH/H2O (30:70,
1.0 mL/min flow). Other conditions (other column, eluents etc.)
remarkably reduced the separation efficiency, so far we examined.6
Due to such separation difficulty and their instability as described
2. Stereochemical corrections of photinide A and B
The 1H and 13C NMR data of 1 and 2 (Table 1) showed very good
accordance with those of Che’s photinides A and B,7 respectively, in
acetone-d6. Our samples gave the protonated ions (1: m/z 305.1022,
2: m/z 305.1029) in the ESI-MS, while Che reported sodium adduct
* Corresponding author. E-mail address: hmasaru@cc.hirosaki-u.ac.jp (M.
Hashimoto).
0040-4020/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tet.2012.06.108

7992
R. Yasumura et al. / Tetrahedron 68 (2012) 7991e7996
Table 1
1H and 13C NMR data for photinides A (1), B (2), X (3), and Y (4) in acetone-d6
Photinide A (1)
d1H (J in Hz)
Photinide B (2)
d1H (J in Hz)
Photinide X (3)
d1H (J in Hz)
Photinide Y (4)
d1H (J in Hz)
Position
d13C
d13C
d13C
d13C
2
3
3a
4
5
6
7
7a
8
9
146.5
182.4
112.3
159.7
106.5
140.3
105.1
167.2
128.7
80.5
146.8
182.4
112.2
159.8
106.6
140.4
105.1
167.2
129.6
83.1
145.7
182.4
112.4
159.6
106.4
140.2
105.2
167.5
129.3
78.9
145.9
182.2
112.2
159.8
106.6
140.3
105.1
167.0
130.2
80.6
6.79 (d, 8.3)
7.68 (t, 8.3)
6.84 (d, 8.3)
6.79 (d, 8.3)
7.69 (t, 8.3)
6.83 (d, 8.3)
6.78 (d, 8.3)
7.67 (t, 8.3)
6.85 (d, 8.3)
6.80 (d, 8.2)
7.69 (t, 8.2)
6.83 (d, 8.2)
6.19 (d, 7.3)
5.28 (d, 7.2)
6.42 (d, 8.7)
5.74 (d, 8.1)
10
11
2.86 (m)
2.30 (dd, 8.4, 17.0)
2.77 (dd, 8.5, 17.0)
37.5
37.2
2.82 (m)
2.33 (dd, 9.1 17.0)
2.79 (dd, 8.6, 17.0)
37.0
36.9
3.13 (m)
2.45 (dd, 8.9, 17.2)
2.65 (dd, 9.2, 17.2)
36.0
34.3
3.12 (m)
2.44 (dd, 7.6, 17.0)
2.74 (dd, 8.8, 17.0)
36.5
34.8
12
14
15
176.8
18.0
56.6
176.7
17.7
54.4
177.4
15.6
56.5
176.9
15.5
53.4
1.21 (d, 6.5)
4.40 (d, 12.6)
4.69 (d, 12.6)
3.95 (s)
1.23 (d, 6.6)
4.39 (d, 13.0)
5.11 (d, 13.0)
3.96 (s)
0.98 (d, 7.1)
4.34 (d, 12.0)
4.69 (d, 12.6)
3.95 (s)
1.05 (d, 7.1)
4.36 (d, 13.5)
5.10 (d, 13.5)
3.96 (s)
16
56.6
56.6
56.6
56.7
ions ([MþNa]þ) by the same ionization method. However, both
indicated the same molecular formula C16H16O6 for these mole-
cules. Our own analyses involving 2D HMBC and HSQC spectra of
these molecules consisted well with Che’s photinides A and B ex-
cept for the stereochemistry at C2]C8 double bonds. In spite of
a (Z)-stereochemistry for 1 in Che’s report, we concluded (E)-form
based on the characteristically high frequency chemical shift of the
C9H resonance (6.19 ppm), compare to that of 2 (5.28 ppm). We
assumed that carbonyl oxygen at C3 carbonyl withdraws the
electrons of the spatially close C9H, which magnetically deshields
C9H of 1.8 However, Che considered a shielding effect by the C3
carbonyl to explain this high frequency shift. Since our assignment
contradicted their conclusion, we further studied that with their
theoretical chemical shifts.9,10 Prior to these calculations, confor-
mational searches for both 1 and 2 were performed using semi-
empirical AM1 method on Spartan1011 to find 12 and 14 stable
conformers, respectively. Each conformer was precisely equili-
brated using the EDF2/6-31G* density functional model12 and then
subjected to chemical shifts calculations with the same theory.
Application of this scheme has been known to reproduce a wide
range of experimental chemical shifts.13 Theoretical chemical shifts
were given after correction by Boltzmann distributions. As expec-
ted, the C9H of 1 localizes geometrically close position to the C3
carbonyl in all stable conformers within 10 kcal/mol steric energy
from the global minimum conformation. These calculations sug-
gested the C9H chemical shift of 1 to be 6.42 ppm, while that of 2
was estimated to be 5.12 ppm. These results accorded with our
assignment as described. The experimental 13C signal for the C9 of 1
(80.5 ppm) appeared 2.6 ppm lower frequency than that of 2
(83.1 ppm), while the C15 of 1 (56.6 ppm) resonated at 2.2 ppm
higher frequency than that of 2 (54.4 ppm). These differences can
be explained by steric compression effect14,15 by the C3 carbonyl
group. Calculations also reproduced those well [C9: 78.5 ppm (1),
80.3 ppm (2), Dd ꢁ2.3 ppm, C15: 60.1 ppm (1), 57.2 (2) Dd 2.9 ppm].
As described, the relative structure of photinides A and B should be
revised as shown Fig. 1.
by heating with benzyl alcohol in the presence of p-toluenesulfonic
acid. We sometimes experience low reliability in specific rotation
values in cases of small quantity of sample and/or small [
Since only small amount of the 6 was available, we investigated that
by CD spectra. With expecting increment of the CD intensity,18
a]D value.
6
was converted into tris-O-(2-naphthoyl) derivative nat-7 by LiAlH4
reduction followed by treatment with 2-naphthoyl chloride in
pyridine. Strong UV absorption of 2-naphthoyl group was also
helpful in the preparative TLC. As expected, nat-7 provided a typical
splitting positive Cotton effect [
see supplemental data] with reliable quality. Then we synthesized
syn-7 (2S,3S-form) from acetone
-glyceraldehyde 819 as shown in
Dε þ118 (241 nm), ꢁ80 (227 nm),
D
Scheme 1. Although 10 (4S,5S-form) was obtained along with its
(4R,5S)-diastereomer epi-10 in almost 1:1 ratio, these could be
separated by silica gel chromatography.20 Stereochemistry of 10
was established by observing a characteristic NOE between the C4-
methyl and the C5-methine protons. The isomer epi-10 showed no
corresponding NOE between those protons but gave that between
C4-methyl and C5-methylene protons, which consisted with above
assignments. The 1H NMR spectra of nat-7 and syn-7 were identical
to correlate their relative stereochemistry. It was found that syn-7
gave quite similar CD profile [
D
ε þ88 (241 nm), ꢁ63 (228 nm)] to
that of nat-7. The j
Dεj values of syn-7 were slightly smaller than
those of nat-7. It could be explained by partial racemization in the
Wittig reaction giving 9, because the reaction required three days
for completion to allow that. The (2S,3R)-isomer of 7, prepared with
epi-10, showed considerably different 1H NMR spectra from that of
nat-7, which led us to undoubtedly establish the (Z,9S,10S)-con-
figuration for 2.
We next investigated their absolute sterochemistry. Che judged
(9R)-configuration for both their photinides A and B by observing
a negative Cotton effect at 210 and 219 nm, respectively, in the CD
spectra. In fact, we also observed the effect. However their con-
clusion involves ambiguity because of only a few examples repor-
ted for their conclusion.16,17 Thus, we determined it by chemical
correlations. When 2 was oxidized with ozone, furanonecarboxylic
acid 5 was obtained after oxidative work up. Since 5 was hardly
detected by TLC, it was isolated in a form of benzyl ester 6 (2.5 mg)
Scheme 1. Preparations of nat-7 and syn-7 (2S,3S-form).

R. Yasumura et al. / Tetrahedron 68 (2012) 7991e7996
7993
Since 1 could be isolated in smaller amount (3.5 mg) than 2
(14.3 mg) due to separation efficiency in the HPLC, 1 could not be
applied to configurational determination in the similar manner.
However, it was determined as (E,9S,10S)-form as described below.
Readily isomerization was observed between 1 and 2, for example,
in acetone-d6. This interconversion became faster in CDCl3 and in
aqueous CH3CN solution containing 0.1% TFA after HPLC purifica-
tion.21 These suggested that protonation on the C2]C8 double
bond generates cation I, which allows the bond rotation as shown
in Scheme 2. The solvent CDCl3 is gradually decomposed to result
HCl, which would catalyze the isomerization even it was trace
amount. However, this mechanism does not induce epimerization
at C9 nor C10. These led us conclude the same configuration for the
C9eC14 furanone moiety of 2. As described, these experiments
revised both relative and absolute stereochemistry of old photi-
nides A and B by Che et al.
similar discussion led the same configuration for the C9eC14 fur-
anone moieties of 3 and 4. Accordingly stereochemistry of 3 and 4
were concluded to be (E,9R,10S)- and (Z,9R,10S)-forms, respectively.
Since C9H is expected be relatively acidic because of C3 C]O group,
interconversion might be possible between 1 and 3. However, we
have not observed that so far.
Fig. 2. Overlaid the most stable conformers of 1 [(E,9S,10S)-form, green] and (9S,10R)-
enantiomer of 3 (purple) by DFT EDF2 with 6-31þG*.
Scheme 2. Plausible isomerization mechanism.
3. Structure of photinide X (3) and Y (4)
Discosia sp. SH 125 also produces photinide X (3) and Y (4) as the
minor components. Similarly to 1 and 2, these were separable but
gradually interconverted each other. Thus, we needed to complete
the separation within a day. Also due to separation difficulty, only
analytical samples could be prepared (2.1 and 1.6 mg, respectively).
These molecules showed the same molecular formulas (C16H16O6)
as both 1 and 2. The 1H and 13C NMR profiles of 3 and 4 (Table 1)
showed considerable differences for the C9eC14 furanone moieties
from those of 1 and 2, respectively, but very resembled for other
parts. These suggested that 3 and 4 are diastereomers about the
furanone moiety but have the same 2-methylenebenzofuran-
3(2H)-one substructure, id est (E,9R*,10S*)-form for
3 and
(Z,9R*,10S*)-form for 4. The relative stereochemistry of the C9eC14
furanone part was confirmed by employing 3. Irradiation of C9H in
3 induced a characteristic NOE at C10H, while 1 gave the NOE be-
tween C9H and C14H3.
Fig. 3. Experimental and calculated CD spectra of 1 and 3.
Absolute chemistries of these minor components were
investigated by combination analyses of molecular modeling and CD
4. Biological assay
spectra. Conformational search was performed for
3 by the
Finally their biological activity was studied to disclose that
same scheme as those for 1 as described. When the most stable
conformers of 1 [being established to be (9S,10S)] and (9S,10R)-en-
photinide A, B and X (1, 2, and 3) showed potent growth inhibition
against C. miyabeanus with 10 mg/mL (IC50), whereas photinide Y (4)
showed considerably weaker activity (IC50 100 mg/mol).
antiomer of
3 were overlaid, the chromophores (O1eC8 2-
methylenebenzofuranone moiety, and C12 carbonyl) in 3 nearly co-
incided with those in 1, as shown in Fig. 2. Conformational searches
also provided other stable conformers for both 1 and 3. In each case,
these were conformational isomers due UV unabsorbent C4OMe and
C15OH groups but showed no considerable conformational differ-
ence in their UV absorbent frameworks. Since geometrical relation-
ship of chromophores affect in CD,22 these isomers should provide
the similar CD spectra and not disturb above discussions. Contrary to
these assumption, 1 and 3 gave CD spectra in nearly inverse re-
lationship, suggesting that 3 is the mirror image, (9R,10S)-enantio-
mer. This was verified by their theoretical CD spectra with DFT B3LYP
with 6-31þG* basis set followed by spectral correction based on the
Boltzmann distributions of the conformers. As shown in Fig. 3, these
calculations reproduced the experimental CD spectra of both 1 and 3.
Compounds 3 and 4 were also readily isomerized each other. The
5. Conclusion
As described, we investigated structures of 1 and 2. Compari-
sons of the experimental and theoretical NMR chemical shifts led
us to revise their stereochemistry of C2]C8 double bonds. Absolute
chemistry of old photinides A and B by Che was also corrected by
CD spectral correlation between the degradation product and the
synthetic sample. We also isolated 3 and 4 as the minor analogues.
Spectral analyses disclosed that these are C9-epimers of 1 and 2.
Although their configurations were not determined directly, those
could be disclosed by a combination of conformational and CD
spectral comparisons between 1 and 3. Theoretical CD calculations
confirmed them.

7994
R. Yasumura et al. / Tetrahedron 68 (2012) 7991e7996
6. Material, methods, and experiments
[MþH]þ: 305.1025), 287.0910 (100, calcd for C16H15Oþ5 ,
[MþHeH2O]þ: 287.0914).
6.1. General
6.4.3. Photinide
X
(3) UV (1.5ꢀ10ꢁ3 mg/mL, CH3CN, nm) 218
Ultraviolet (UV) spectra were obtained by a HITACHI U-2010
spectrometer. Circular dichroism (CD) spectra were measured on
a JASCO J-725 spectropolarimeter. NMR spectra were recorded on
a JEOL JNM-ECX500 spectrometer. In the case acetone-d6 was used,
(ε 13,000), 272 (ε 11,000), 304 (ε 3900), 367 (ε 4900). IR (NaCl) 3440
(br), 2925, 1775, 1610, 1260 cmꢁ1. The CD (5.1ꢀ10ꢁ5 mol/L, CH3CN)
and NMR spectral data (acetone-d6) are shown in Fig. 3 and Table 1,
respectively. ESIMS (rel int, %) m/z 305.1033 (60, calcd for C16H17O6
[MþH]þ: 305.1019), 287.0927 (100, calcd for C16H15Oþ5 ,
[MþHeH2O]þ: 287.0914).
CHD2COCD3 (1H: 2.03 ppm) and 13CD3COCD3 13C: 29.9 ppm) were
(
employed as the standards. In the cases CDCl3 was used, tetrame-
thylsilane (0.0 ppm for 1H and 13C NMR) was used as the internal
standard. Mass spectra were measured by electrospray-ionization
(ESI) mass spectrometryon a HITACHI NanoFrontier LDspectrometer.
6.4.4. Photinide Y (4) UV (1.4ꢀ10ꢁ3 mg/mL, CH3CN, nm) 218 nm
(ε 15,000), 271 (ε 11,000), 307 (ε 3500), 366 (ε 4300). IR (NaCl) 3400,
2925, 1780, 1610, 1260 cmꢁ1. CD (4.6ꢀ10ꢁ5 mol/L, CH3CN) 213 nm
(
D
ε þ1.5), 273 nm (
D
ε ꢁ2.0), 332 nm (
D
ε þ0.7), 362 nm ( ε þ1.7).
D
6.2. Fungus
The CD spectrum of this sample may not be not reliable due to the
impurity. The NMR spectral data (acetone-d6) are shown in Table 1.
ESIMS (rel int, %) m/z 305.1025 (15, calcd for C16H17O6 [MþH]þ:
305.1019), 287.0923 (100, calcd for C16H15Oþ5 , [MþHeH2O]þ:
287.0914).
Discosia sp. SH 125 was collected from fallen leaves of C. crenata
Aomori Prefecture, Japan in 2003, and the fungal isolate was de-
posited at the National Institute of Agrobiological Sciences, Japan as
MAFF 242783.
6.5. Preparation of tris(2-naphthoyl)ester of (2S,3S)-3-
methylpentane-1,2,5-triol (nat-7) from 2
6.3. Calculations
Conformational searches and chemical shift calculations were
performed with Spartan 10 (Wavefunction, Irvine, CA) using
a hand-made PC (operating System: Windows7 Professional, CPU:
AMD Phenom(tm) X3 945 processor 3.00 GHz, RAM 8 GB). Theo-
retical CD spectra were calculated with Gaussian 09 (Revision A.02
by Gaussian, Wallingford, CT) with a PC (Operating System: CentOS
a Linux, CPU: 2 Intel Xeon 3 5550 processors 2.67 GHz, RAM 24 GB).
6.5.1. (2S,3S)-Benzyl 3-methyl-5-oxotetrahydrofuran-2-carboxylate
(6) Ozone was bubbled to a solution of 2 (10 mg, 33 mmol) in
MeOH (8.0 mL) at ꢁ78 ꢂC and the mixture was stirred at the same
temperature for 30 min. Aqueous H2O2 solution (30%, 50 mL) was
added in order to decompose the ozonide, the mixture was further
stirred at room temperature for additional 30 min. After the mix-
ture was diluted with H2O (20 mL), methanol was removed by ro-
tary evaporator to obtain the aqueous solution. It was extracted
with EtOAc (50 mL) and the organic layer was washed with 1.0 M
aqueous HCl solution (30 mL). The organic solution was dried over
MgSO4, and then concentrated under reduced pressure. The residue
6.4. Isolation
Discosia sp. was cultured in potato-sucrose medium (200 mL in
500 mL baffled Erlenmeyer flaskꢀ10) on a rotary shaker (110 rpm)
at 25.8 ꢂC for 10 days. After filtration in suction, the filtrate was
concentrated bellow 25 ꢂC under reduced pressure until the whole
volume became 1.0 L. The resulting solution was extracted with
EtOAc (500 mLꢀ2) and the organic layer was dried over MgSO4 and
concentrated under reduced pressure. Silica gel column chroma-
tography of the residue eluting with 0, 1, and 3% acetone/chloro-
form and the fraction eluted with 1% acetone/chloroform was
recovered and concentrated to give a mixture of photinides AeD
(650 mg). Part of the mixture was subjected to HPLC [LiChrospherÒ
was dissolved in benzyl alcohol (100
solution was stirred with p-TsOH (3.0 mg) at 90 ꢂC for 10 h. After
Et3N (50 L) was added, the mixture was concentrated and the
mL) and benzene (1.0 mL). The
m
resulting oily material was further concentrated by bulb to bulb
distillation. Silica gel column chromatography of the residue with
EtOAc/benzene (3:97) gave 6 containing benzyl alcohol as the im-
purity (total 2.5 mg) 1H NMR (CDCl3)
d
1.24 (3H, d, J¼6.9 Hz), 2.14
(1H, dd, J¼5.6, 17.4 Hz), 2.63 (1H, m), 2.74 (1H, dd, J¼8.4, 17.4 Hz),
4.50 (1H, d, J¼4.8 Hz), 5.22 and 5.23 (each 1H, d (AB), J¼12.3 Hz),
7.35 (aromatic protons), Other than above signals, benzyl proton
(4.67 ppm, s, 2H proportion) was observed. Irradiation of the
doublet methyl at 1.25 ppm induced NOEs at the signals of 2.15,
2.63, 4.52 ppm. LC-ESIMS m/z 273.0533 (25, calcd for C13H14O4K,
[MþK]þ: 273.0529), 257.0796 (35, calcd for C13H14O4Na, [MþNa]þ:
257.0790), 252.1239 (100, calcd for C13H18O4N, [MþNH4]þ:
252.1236), 235.0972 (20, calcd for C13H15O4, [MþH]þ: 235.0970).
RP-18e (4.0
f
ꢀ250 mm) MeOH/H2O (30:70), 1.0 mL/min flow] to
give photinide A (1, 3.5 mg, tR 37 min), photinide B (2, 14.3 mg, tR
56 min), photinide X (3, 2.1 mg, tR 40 min) and photinide Y (4,
1.6 mg, tR 43 min) (photinide D).
6.4.1. Photinide A (1) UV (2.6ꢀ10ꢁ3 mg/mL in CH3CN, nm) 218
(ε 12,500), 271 (ε 11,000), 309 (ε 3600), 377 (ε 4600). IR (NaCl) 3450,
2850,1780,1610,1495,1090 cmꢁ1. The CD spectrum (8.6ꢀ10ꢁ5 mol/
L in CH3CN) and NMR spectral data (acetone-d6) are shown in Fig. 3
and Table 1. The 1H and 13C NMR data showed good accordance
with those photinide A in Che’s report.7 ESIMS (rel int, %) m/z
305.1022 (85, calcd for C16H17O6 [MþH]þ: 305.1025), 287.0910 (100,
calcd for C16H15O5þ, [MþHeH2O]þ: 287.0914).
6.5.2. Tris-O-(2-naphthoyl) ester of (2S,3S)-3-methylpentane-1,2,5-
triol (nat-7) The obtained benzyl ester 6 containing benzyl alco-
hol (2.5 mg) was stirred with LiAlH4 (3.0 mg, 79 mmol) in THF
(2.0 mL) at 0 ꢂC for 1 h. Saturated aqueous NH4Cl solution (two
drops) and Celite (ca. 50 mg) were added and the mixture was
further stirred at room temperature for 30 min. After filtration
through Celite pad, the filtrate was concentrated in vacuo. Silica gel
column chromatography of the residue with MeOH/EtOAc (3:97)
6.4.2. Photinide B (2) UV (5.9ꢀ10ꢁ3 mg/mL in CH3CN, nm) 218
(ε 14,000), 271 (ε 11,000), 309 (ε 3600), 377 (ε 5700). IR (NaCl) 3450,
2970, 1780, 1490, 1260, 1090 cmꢁ1. CD (5.9ꢀ10ꢁ3 mol/L in CH3CN,
gave the triol (ca. 2.0 mg). 1H NMR (CDCl3)
d
0.88 (3H, d, J¼6.8 Hz),
1.5e1.7 (3H, m), 3.46 (2H, m), 3.61 (1H, ddd, J¼4.5, 7.7, 10.9 Hz), 3.67
nm) 219 (
D
ε ꢁ2.4), 271 (
D
ε þ1.3), 327(
D
ε 0), 385 (
D
ε þ1.0). The NMR
(1H, dd, J¼2.1, 9.9 Hz), 3.75 (1H, ddd. J¼4.5, 6.6, 10.9 Hz). The
spectral data (acetone-d6) are shown in Table 1. The 1H and 13C NMR
data showed good accordance with those of photinide B in Che’s
report.7 ESIMS (rel int, %) m/z 305.1029 (85, calcd for C16H17O6
obtained triol (2.0 mg, 15 mmol) was stirred with 2-naphthoyl
chloride (14 mg, 75
mmol) in pyridine (1.0 mL) at room tempera-
ture for 12 h in the dark. MeOH (100
mL) was added to decompose

R. Yasumura et al. / Tetrahedron 68 (2012) 7991e7996
7995
the excess reagent. The mixture was poured into saturated aqueous
NaHCO3 solution (20 mL) and extracted ether (20 mLꢀ3). Com-
bined ethereal solution was washed with brine, dried over MgSO4,
and then concentrated in vacuo. Preparative TLC of the residue with
EtOAc/hexane (20:80) gave nat-7 (4.0 mg, 6.7ꢀ10ꢁ3 mmol). CD
calcd 5.70 ppm). Calculations were performed in the similar man-
ner as described in the text. The ESI method did not ionize this
sample.
6.6.2. (4S,5S)-5-Benzoyloxymethyl-3,4-dihydro-3-methyl-furanone
(10) and its (4R)-isomer(epi-10) (Z)-Isomer of 9 (30 mg, 150 mmol)
(6.0
mmol/
mL, MeOH)
Dε þ118 (241 nm), ꢁ82 (227 nm)]. IR (NaCl)
2920, 2850, 1715, 1280, 1195 cmꢁ1. 1H NMR 1.26 (3H, d, J¼6.9 Hz),
1.86, 2.28, and 2.43 (each 1H, m), 4.51 (1H, ddd, J¼6.1, 8.2, 11.0 Hz),
4.58 (1H, ddd, J¼5.5, 6.7, 11.0 Hz), 4.64 (1H, dd, J¼7.2, 11.9 Hz), 4.78
(1H, dd, J¼3.6, 11.9 Hz), 5.61 (1H, ddd, J¼3.6, 6.2, 7.2 Hz), 7.49 (3H,
m), 7.52 (3H, m), 7.73e7.98 (9H), 7.95, 8.02, and 8.07 (each 1H, dd,
J¼1.7, 8.6 Hz), 8.50, 8.57, 8.62 (each 1H, br s). ESIMS (rel int, %) m/z
614.2526 (100, calcd for C39H36O6N, [MþNH4]þ: 614.2537),
619.2084 (35, calcd for C39H32O6Na, [MþNa]þ: 619.2091), 597.2265
(10, calcd for C39H33O6, [MþH]þ: 597.2272).
was stirred vigorously with 10% Pd/C (7.0 mg) in MeOH (2.0 mL)
under H2 atmosphere (5.0 atm) at room temperature for 12 h. The
mixture was filtered and the filtrate was concentrated in vacuo. The
residue was diluted with MeOH (2.0 mL) again and stirred with p-
TsOH (2.0 mg) at room temperature for 12 h. After Et3N (two drop)
was added to neutralize the mixture, then it was concentrated
under reduced pressure. The residue was stirred with benzoyl
chloride (100 mL) in pyridine (1.0 mL) at room temperature for
12 h. MeOH was added to destroy the excess reagent. The mixture
was poured into saturated aqueous NaHCO3 solution (20 mL) and
then extracted with ether (20 mLꢀ2). The combined organic solu-
tion was washed with brine (20 mL), dried over MgSO4, and then
concentrated in vacuo. Silica gel column chromatography of the
residue (EtOAc/hexane¼80:20) gave a mixture of 10 [(4S,5S)-form]
and epi-10 [(4R,5S)-form] (28 mg, 79%). Employing part of the
mixture (14 mg), preparative silica gel TLC (EtOAc/hexane¼40:60)
was performed to provide 10 (6.0 mg, Rf 0.5) and epi-10 (7.0 mg, Rf
0.45). The reaction employing the (E)-9 gave the same products in
the almost same diastereomeric ratio.
6.6. Preparation of syn-7 (2S,3S-form)
6.6.1. (S)-Methyl 3-(2,2-dimethyl-1,3-dioxolan-4-yl)but-2-enoate
(9) Crude acetone
D-glyceraldehyde, prepared from 1,2,5,6-di-O-
isopropylidene- -mannitol (500 mg, 1.90 mmol) by NaIO4 oxida-
D
tion under standard conditions,19 was stirred with methyl-
magesium bromide (0.99 M in THF, 6.5 mL) in THF (10 mL) at 0 ꢂC
for 2 h. The mixture was poured into water and extracted with ether
(20 mLꢀ3). Combined ethereal solution was washed with brine
(30 mL), dried over MgSO4 and the concentrated under reduced
pressure. Silica gel column chromatography of the residue with
EtOAc/hexane (12:88) gave the alcohol (458 mg, 3.13 mmol, 82% in
two steps) as 2:1 diastereomeric mixture about the methylcarbinol
moiety. IR (NaCl) 3450, 2990, 2890, 1370, 1066 cmꢁ1. 1H NMR (the
6.6.2.1. Physical data for 10 [(4S,5S)-form]. [
CHCl3). IR (NaCl) 2970, 1720, 1470, 1270, 1120 cmꢁ1. 1H NMR (CDCl3)
1.89 (3H, d, J¼7.0 Hz), 2.20 (1H, dd, J¼8.7, 17.5 Hz), 2.43 (1H, m),
a]
þ50 (c 1.0,
D
d
2.72 (1H, dd, J¼8.4, 17.5 Hz), 4.32 (1H, ddd, J¼3.1, 5.4, 8.3 Hz), 4.39
(1H, dd, J¼5.4, 12.2 Hz), 4.51 (1H, dd, J¼3.1, 12.2 Hz), 7.39 (2H, br t,
J¼8 Hz), 7.52 (1H, br t, J¼8 Hz), 7.97 (2H, br d, J¼8 Hz). Irradiation of
the methyl doublet (1.89 ppm) induced an NOE at the oxymethine
proton (4.32 ppm). ESIMS (rel int, %) m/z 257.0809 (40, calcd for
C13H14O4Na, [MþNa]þ: 257.0784), 252.1254 (30, calcd for
C13H18O4N, [MþNH4]þ: 252.1230), 235.0989 (100, calcd for
C13H15O4, [MþH]þ: 235.0965).
sample was composed of diastereomers. CDCl3)
d
1.08 (3Hꢀ0.33
(for minor) d, J¼6.2 Hz),1.10 (3Hꢀ0.67 (for major), d, J¼6.4 Hz),1.30
and 1.37 (each 3H, br s) 3.62 (1Hꢀ0.33 (for minor), dd, J¼6.3,
8.2 Hz), 3.85e3.97 (4Hꢀ0.67 (for major) and 3Hꢀ0.33 (for minor)).
The obtained alcohol (200 mg, 1.37 mmol) was oxidized under
standard Swern’s condition23 in CH2Cl2 (15 mL) employing dime-
thylsulfoxide (500 mL), oxalyl chloride (240 mL), and Et3N (2.0 mL) to
give the methyl ketone (190 mg, 96%) after silica gel column chro-
matography. IR (NaCl) 2985, 1720, 1460, 1380, 1065 cmꢁ1. 1H NMR
6.6.2.2. Physical data for epi-10 [(4R,5S)-form]. [
CHCl3). IR (NaCl) 2970, 2850, 1790, 1450, 1270, 1110 cmꢁ1. 1H NMR
(CDCl3)
1.15 (3H, d, J¼6.7 Hz), 2.34 (1H, dd, J¼7.0, 17.5 Hz), 2.70
a
]
þ71 (c 1.0,
D
(CDCl3)
d
1.36, 1.46, and 2.22 (each 3H, s), 3.97 (1H, dd, J¼5.6,
8.7 Hz), 4.17 (1H, dd, J¼7.8, 8.7 Hz), 4.38 (1H, dd, J¼5.6, 7.8 Hz).
Whole amount of the ketone thus obtained was dissolved in MeOH
(6.0 mL) and stirred with methyl (triphenylphosphoranylidene)
acetate (670 mg, 2.8 mmol) at room temperature for 3 days. After
concentration, the residue was purified with silica gel column
chromatography (EtOAc/hexane 4:96) to give the (Z)-isomer of 9
(112 mg, 42%) and the (E)-isomer of 9 (94.0 mg, 36%).
d
(1H, dd, J¼8.3, 17.5 Hz), 2.83 (1H, m), 4.45 (1H, dd, J¼5.5, 12.6 Hz),
4.55 (1H, dd, J¼3.5, 12.6 Hz), 4.78 (1H, ddd, J¼3.5, 5.5, 7.0 Hz), 7.43
(2H, br t, J¼8 Hz), 7.56 (1H, br t, J¼8 Hz), 7.99 (2H, br d, J¼8 Hz).
Irradiation of the methyl doublet signal (1.19 ppm) induced an NOE
at the oxymethylene protons (4.41 and 4.51 ppm). ESIMS (rel int, %)
m/z 257.0799 (45, calcd for C13H14O4Na, [MþNa]þ: 257.0784),
252.1245 (25, calcd for C13H18O4N, [MþNH4]þ: 252.1230), 235.0977
(100, calcd for C13H15O4, [MþH]þ: 235.0965).
6.6.1.1. Physical data for the (E)-isomer. [
IR (NaCl) 2990, 2930, 1720, 1660, 1465, 1230 cmꢁ1. 1H NMR (CDCl3)
1.39, 1.44 (each 3H, s), 2.08 (3H, d, J¼1.0 Hz), 3.62 (1H, dd, J¼7.9,
a]
þ27 (c 1.0, CHCl3).
D
d
6.6.3. Tris(2-naphthoyl) ester of (2S,3S)-3-methylpentane-1,2,5-triol
8.1 Hz), 3.69 (3H, s), 4.19 (1H, dd, J¼6.9, 8.1 Hz), 4.50 (1H, dd, J¼8.0,
7.9 Hz), 6.02 (quint, J¼1.0 Hz). (E)-Stereochemistry for the double
bond was established by a comparison of the experimental and the
theoretical chemical shift for the oxymethine proton (exp.:
4.50 ppm, calcd 4.46 ppm). Calculations were performed in the
similar manner as described in the text. The ESI method did not
ionize this sample.
(syn-7) Benzoyl ester 10 (40 mg, 171 mmol) in THF (1.0 mL) was
stirred with LiAlH4 (5.0 mg) at 0 ꢂC for 30 min. The work up and
purification were performed in the similar manner as the described
in Section 6.5.2 to give the triol (20 mg, 147
spectrum of the crude sample was identical with that from natural
sample. Triol thus obtained (20 mg, 147 mol) was stirred with 2-
naphthoyl chloride (140 mg, 750 mol) and pyridine (1.5 mL) in
m
mol). The 1H NMR
m
m
the similar manner as described in the Section 6.5.2 to give syn-7
(8.0 mg, 20%) after silica gel column chromatography. Chromato-
graphic behavior and the 1H NMR spectrum of syn-7 were identical
6.6.1.2. Physical data for the (Z)-isomer. [
a
]
þ49 (c 1.0, CHCl3).
D
IR (NaCl) 2990, 2850, 1720, 1450, 1160 cmꢁ1. 1H NMR (CDCl3)
d
1.37,
1.46 (each 3H, s), 1.97 (3H, d, J¼1.0 Hz), 3.54 (1H, dd, J¼6.7, 8.1 Hz),
3.66 (3H, s), 4.37 (1H, dd, J¼7.5, 8.1 Hz), 5.70 (1H, br t, J¼8 Hz), 5.74
(1H, quint, J¼1.0 Hz). (Z)-Stereochemistry for the double bond
was established by a comparison of experimental and theoretical
chemical shift for the oxymethine proton (exp.: 5.70 ppm,
with that of nat-7. CD (9.6ꢀ10ꢁ3 mmol/mL, MeOH)
D
ε þ88
(241 nm), ꢁ63 (228 nm). IR (NaCl) 3060, 2925, 2850, 1715, 1280,
1195 cmꢁ1. The 1H NMR spectrum of this sample was identical to
that from natural nat-7. ESIMS (rel int, %) m/z 597.2275 (15, calcd for
C39H33O6, [MþH]þ: 597.2271), 614.2545 (100, calcd for C39H36O6N,

7996
R. Yasumura et al. / Tetrahedron 68 (2012) 7991e7996
[MþNH4]þ: 614.2537), 619.2095 (50, calcd for C39H33O6Na,
[MþNa]þ: 619.2097).
References and notes
1. Murakami, T.; Morikawa, Y.; Hashimoto, M.; Okuno, T.; Harada, Y. Org. Lett. 2004,6,157.
2. Hashimoto, M.; Tsushima, T.; Murakami, T.; Nomiya, M.; Takada, N.; Tanaka, K.
Bioorg. Med. Chem. Lett. 2008, 18, 4228.
6.6.3.1. (2S,3R)-isomer of syn-7. The (2S,3R)-isomer epi-10
(40 mg, 170
with LiAlH4 (5.0 mg) in THF (1.0 mL) and then 2-naphthylated with
2-naphthoyl chloride (140 mg, 750 mol) and pyridine (1.5 mL) in
m
mol) obtained in Section 6.6.2.2 was also reduced
3. Yamamoto, K.; Noguchi, S.; Takada, N.; Miyairi, K.; Hashimoto, M. Carbohydr.
Res. 2010, 345.
4. Tayone, W. C.; Honma, M.; Kanamaru, S.; Noguchi, S.; Tanaka, K.; Nehira, T.;
Hashimoto, M. J. Nat. Prod. 2011, 74, 425.
5. Nihei, K.; Itoh, H.; Hashimoto, K.; Miyairi, K.; Okuno, T. Biosci. Biotechnol. Bio-
m
the same manner as described in Section 6.6.3 to afford (2S,3R)-
isomer of 7 (10 mg). CD (7.2ꢀ10ꢁ3 mmol/mL, MeOH)
D
ε þ56
chem. 1998, 62, 858.
6. We failed in scaling up with semi-preparative column (LiChrospher RP-18e,
(242 nm), ꢁ32 (228 nm). IR (NaCl) 3060, 2920, 1715, 1280,
10 mm 10
fꢀ250 mm).
1190 cmꢁ1. 1H NMR (CDCl3)
d
1.29 (3H, d, J¼6.9 Hz), 1.84 (1H, m),
7. Ding, G.; Zheng, Z.; Liu, S.; Zhang, H.; Guo, L.; Che, Y. J. Nat. Prod. 2009, 77, 942.
8. Silverstein, R. M.; Webster, F. X.; Kiemle, D. J. Spectrometric Identification of
Organic Compounds Hoboken; John Wiley & Sons: 2005.
2.21 (1H, m), 2.41 (1H, m), 4.53 (1H, ddd, J¼5.9, 6.4, 11.2 Hz), 4.57
(1H, dd, J¼6.1, 11.2 Hz), 4.68 (1H, dd, J¼7.3, 11.8 Hz), 4.73 (1H, dd,
J¼4.7, 11.8 Hz), 5.71 (1H, ddd, J¼4.3, 4.6, 7.3 Hz), 7.41e7.58 (6H),
7.71e7.97 (9H), 7.94, 7.96, 8.06 (each 1H, dd, J¼1.7, 8.6 Hz), 8.51,
8.52, 8.61 (each 1H, s). ESIMS (rel int, %) m/z 597.2259 (5, calcd for
C39H33O6, [MþH]þ: 597.2271), 614.2526 (100, calcd for C39H36O6N,
[MþNH4]þ: 614.2537), 619.2078 (55, calcd for C39H33O6Na,
[MþNa]þ: 619.2097).
9. Murakami, T.; Takada, N.; Hehre, W. ,J.; Hashimoto, M. Bioorg. Med. Chem. Lett.
2008, 18, 4547.
10. Honma, M.; Kudo, S.; Takada, N.; Miura, T.; Hashimoto, M. Bioorg. Med. Chem.
Lett. 2010, 20, 709.
11. Wavefunction, 18401 Von Karman Avenue, Suite 370, Irvine, CA 92612.
12. Lin, C. Y.; George, M. W.; Gill, P. M. W. Aust. J. Chem. 2004, 57, 365.
13. Wavefunction, Tutorial and User’s Guide, Spartan ’10 for Windows, Macintosh,
and Linux, page 293.
14. Dalling, D. K.; Grant, D. M. J. Am. Chem. Soc. 1972, 94, 5318.
15. Prashad, M.; Seth, M.; Bhaduri, A. P.; Banerji, A. Org. Magn. Reson. 1980, 14, 74.
16. Vigneron, J. P.; Meric, R.; Larcheveque, M.; Debal, A.; Kunesch, G.; Zagatti, P.;
Gallois, M. Tetrahedron Lett. 1982, 28, 5051.
Acknowledgements
17. Vigneron, J. P.; Meric, R.; Larcheveque, M.; Debal, A.; Lallemand, J. Y.; Kunesch,
We appreciate Prof Warren Hehre (Wavefunction Inc.) for kind
and valuable discussions.
G.; Zagatti, P.; Gallois, M. Tetrahedron 1984, 40, 3521.
18. Kanamaru, S.; Honma, M.; Murakami, T.; Tsushima, T.; Kudo, S.; Tanaka, K.;
Nehira, T.; Hashimoto, M. Chirality 2012, 24, 137.
19. Schmid, C. R.; Bryant, J. D. 1998; Collect. Vol. No. 9, p 450.
20. Benzoyl group was introduced for effective separation.
21. Due to this isomerization, we failed in HPLC purification in the presence of acids.
22. Harada, N.; Nakanishi, K. Circular Dichroic Spectroscopy-exciton Coupling in Or-
ganic Stereochemistry; University Science Books: Mill Valley, CA, 1984.
23. Mancuso, A. J.; Huang, S. L.; Swern, D. J. Org. Chem. 1978, 43, 2480.
Supplementary data
Supplementary data related to this article can be found online at
http://dx.doi.org/10.1016/j.tet.2012.06.108.