Stereostructure reassignment and determination of the absolute configuration of pericosine Do by a synthetic approach

Article abstract of DOI:10.1021/np100843j

A combination of chemical synthesis and NMR methods was used to reassign the structure of pericosine D_o (8), a cytotoxic marine natural product produced by the fungus Periconia byssoides OUPS-N133 that was originally derived from the sea hare A

Full text of DOI:10.1021/np100843j

NOTE
pubs.acs.org/jnp
Stereostructure Reassignment and Determination of the Absolute
Configuration of Pericosine Do by a Synthetic Approach
Yoshihide Usami* and Koji Mizuki
Laboratory of Pharmaceutical Organic Chemistry, Osaka University of Pharmaceutical Sciences, 4-20-1 Nasahara, Takatsuki,
Osaka 569-1094, Japan
S Supporting Information
b
ABSTRACT: A combination of chemical synthesis and NMR methods was used to
reassign the structure of pericosine Do (8), a cytotoxic marine natural product produced
by the fungus Periconia byssoides OUPS-N133 that was originally derived from the sea
hare Aplysia kurodai. Chemical synthesis was used to prepare pericoisne Do (8) from a
known chlorohydrin that was in turn derived from (-)-quinic acid. The absolute
configuration of natural pericosine Do (8) was determined to be methyl (3R,4S,5S,6S)-
6-chloro-3,4,5-trihydroxy-1-cyclohexene-1-carboxylate. HPLC analyses using a chiral-
phase column indicated that pericosine Do (8) exists in an enantiomerically pure form in
nature.
ericosines A-E (1-5) are a family of cytotoxic metabolites
isolated from Periconia byssoides OUPS-N133, a fungus that
and in understanding the biosynthetic pathway of these natu-
ral products.
P
was collected from the sea hare Aplysia kurodai.1 We report
herein our ongoing efforts to elucidate the structures of members
of this family of carbasugars. Interest in this family of molecules
arises in part due to their unique and highly functionalized C-7
cyclohexenoid structures, significant bioactivity, and unknown
absolute configurations. The latter poses an interesting challenge
for asymmetric synthesis.2-7 Through our synthetic studies, we
elucidated the absolute configuration of naturally occurring 1 and
also prepared the compound that corresponds to the structure
originally reported as pericosine D (4). However, the data for
synthetic 4 did not agree with those appearing in the literature.1
In contrast, deprotection of acetonide 4a, which was supplied by
the original authors, provided a product whose spectroscopic
data matched satisfactorily with synthetic 4. The structure of
sample 4a used in the deprotection experiment was reconfirmed
by comparison of 1H NMR data with those reported, eliminating
the possibility that there was a mix-up of samples. The fact that 4a
was originally derived from a natural compound indicates that
the structure of 4 does represent a natural product, but the data
originally reported for pericosine D actually correspond to an
undefined diastereomer of pericosine D. Therefore, the deter-
mination of the correct structure for the original pericosine D is
still needed. We have designated this unassigned molecule
pericosine Do to indicate the natural product originally named
pericosine D. This compound should be another diastereomer
of 1 and 4. It is important to elucidate the structure of peri-
cosine Do because of its significant cytotoxicity to the murine
P388 leukemia cell line (ED50 value: 3.0 ug/mL). In this paper
we describe the revision of the relative configuration and the
determination of the absolute configuration of pericosine Do
by a synthetic approach. This is significant for natural product
chemistry both in the discovery of new drug lead compounds
All possible diastereomers of C-6 chlorinated pericosines (1,
4, 6-11) are shown in Figure 1. In prior studies2-4 we com-
pleted the synthesis of the four diastereomers (1, 4, 6, and 7)
containing a cis-configuration between C-3 and C-4 and noted
that each of these isomers did not correspond to pericosine Do.
Therefore, pericosine Do must have a trans-configuration be-
tween C-3 and C-4, and the correct structure must be either
compound 8, 9, 10, or 11. Among the four possible 3,4-trans-
diastereomers, we also synthesized compound 92 and found it to
be different from pericosine Do as well. The three remaining
diastereomers 8, 10, and 11 were potential targets for synthesis.
There are few reports of highly functionalized cyclohexenoids
possessing such relative configurations as 10 or 11 in natural
compounds.8-12 Because we had a sample of precursor 8a from
our previous work, 8 was the most accessible molecule. The
synthesis of 8 is summarized in Scheme 1. An inseparable mixture
of epoxides 14 and 15 (ca. 3:2) was obtained through a five-step
synthesis from (-)-quinic acid (12) via unstable diene 13.4
Treatment of the mixture of 14 and 15 with HCl afforded a
mixture of 6a, 4a0, 8a, and 16a. As mentioned in our previous
work,4 the relative configuration of 8a was determined by
comparing its 1H NMR coupling constants with those of struc-
turally related compounds. The NOESY spectrum of 8a did not
give any crucial cross-peaks. Chlorohydrins 6a, 4a0, and 8a
should be derived from epoxide 14. The configuration of 14
was confirmed as described in our previous paper.4 This was also
supported by the fact that 14 is different from the diastereomeric
isomer 17 that appeared in another recent work of ours.6 There-
fore, the configuration of C-3 in 8a was assigned as R, and the
Received: November 18, 2010
Published: March 10, 2011
r
Copyright
2011 American Chemical Society and
877
dx.doi.org/10.1021/np100843j J. Nat. Prod. 2011, 74, 877–881
American Society of Pharmacognosy
|

Journal of Natural Products
NOTE
Figure 1. Structures of pericosines A-E (1-5) and possible chlorine-containing diastereomers (1, 4, and 6-11).
Scheme 1. Synthesis of (-)-Pericosine Do (8)
*Racemic 6a, 4a0, 8a, and 16a were obtained via the same procedure from a racemic mixture of 14 and 15 (ca.3:1)13 in 47, 1, 4, and 7% yields, respectively.
See Experimental Section.
878
dx.doi.org/10.1021/np100843j |J. Nat. Prod. 2011, 74, 877–881

Journal of Natural Products
Table 1. NMR Spectroscocic Data of Synthesized and Natural Pericosine Do (8) (1H 500 MHz, 13C 125 MHz, acetone-d6)
NOTE
synthesized pericosine Do (8)
δC, mult. δH, (J in Hz)
natural pericosine Do (8)a
δH, (J in Hz)
position
HMBC
δC, mult.
1
2
3
4
5
6
7
8
129.1, C
144.8, CH
69.8, CH
71.1, CH
75.2, CH
56.1, CH
165.8, C
52.4, CH3
129.06, C
144.78, CH
69.79, CH
71.10, CH
75.19, CH
56.02, CH
165.86, C
52.38, CH3
6.91, d (2.5)
1, 4, 6, 7
1, 2, 4
3
6.90, d (2.5)
4.47, br dd (8.5, 2.5)
3.99, dd (8.5, 2.6)
4.20, dd (3.0, 2.5)
4.82, dd (3.0, 0.8)
4.46, ddd (8.5, 6.2,b 2.5)
3.99, ddd (8.5, 6.0,b 2.6)
4.21, dt (3.1,b 2.6)
4.83, dd (2.6, 1.1)
1, 3, 4
1, 2, 4, 5, 7
3.78, s
1, 7
3.77, s
a Reference 1. b Coupling with -OH.
Figure 2. Structures of reference compounds.
remaining undefined configuration of C-6 in 8 must be S. If C-6
in 8 had the R-configuration, 8a should be converted into known
compound 9.2 Chlorohydrin 8a was treated with TFA in MeOH
at room temperature for deprotection of the 3,4-O-cyclohexyli-
dene moiety to afford a product in 96% yield. This product was
different from 9.
8 ([R]D þ1.9).1 Because of the difference in specific rotation
values between synthesized (-)-8 and natural 8, in addition to
the difference in their signs, we questioned whether or not
natural 8 also exists as an enantiomeric mixture, similar to 2, 3,
and 5.1,13 Therefore, we synthesized racemic 8 starting from the
Diels-Alder adducts derived from the reaction of furan and
methyl acrylate via racemic 13. According to the procedures from
our previous work,5,6,13 an approximately 3:1 mixture of racemic
14 and 15 was obtained from rac-13 by the oxidation with
DMDO (dimethyldioxirane). The mixture of rac-14 and rac-15
was similarly treated with HCl in dry ether to afford rac-6a
(47%), rac-4a0 (1%), rac-8a (4%), and rac-16a (7%). The impro-
ved yields of the chlorohydrins were due to the improved ratio of
14:15.13 From rac-8a, rac-8 was prepared in the same way as
described above. Synthesized rac-8 was subjected to chiral-phase
HPLC to determine a suitable column and conditions that would
yield clear enantioseparation in a 50:50 ratio. After many trials,
CHIRALPAK IA (Daicel) was selected.13 Natural 8 and synthe-
sized (-)-8 derived from 12 were subjected to HPLC analysis
with CHIRALPAK IA under the optimum conditions. To our
surprise, they were identical, suggesting that natural 8 exists in an
enantiomerically pure form. Both HPLC samples of synthetic
and natural 8 had the same retention time and showed positive
optical rotation at 426 nm. Then, the specific rotation (at
589 nm) of natural 8 was remeasured using the same conditions,
and an acceptable [R]25D value of -6.6 was obtained. From these
experiments, the absolute configuration of natural pericosine Do
was assigned as methyl (3R,4S,5S,6S)-6-chloro-3,4,5-trihydroxy-
1-cyclohexene-1-carboxylate (8).
The spectroscopic data of synthesized 8 satisfactorily agreed
with the reported data of pericosine Do1 except for the specific
1
rotation. The assignments of the H and 13C NMR signals for
synthesized 8 were confirmed by 2D NMR experiments. Similar
to 8a, the NOESY spectrum of 8 did not yield any crucial
information on the relative configuration. Details of the NMR
data of 8 are provided in Table 1.
Interestingly, synthesized (-)-8 corresponds to the core part
of both the chlorine-containing natural product 18 and cyathi-
formine C (19) (Figure 2). Compounds 18 and 19 are metabo-
lites of other fungi.15,16 The fact that 19 was synthesized from
cyathiformine B (20)17,18 implies the possibility of a similar epoxide
precursor in the biosynthetic pathway of the pericosines.6
The similarity of the 1H NMR coupling constants between 8
and 18 measured with acetone-d6 also supported our assignment
of the relative configuration of 8. Another non-natural regioi-
somer of pericosine, (-)-16, which was derived from 16a, also
showed comparable coupling constants in the same solvent. In
addition, the coupling constants for 1915 and (6S)-6-hydroxy-5-
epishikimic acid (21)12 were also similar to those for 8, despite
having NMR data recorded in CDCl3 and methanol-d4, respec-
tively. This implies that 19 and 21, which possess the same relative
configuration as 8, take similar conformations in different solvents.
The specific rotation of synthesized 8, which was derived from
12, had the opposite sign ([R]25D -6.5) of the reported datum of
Now that the structure of pericosine Do has been established,
we must discuss the possibility of a transformation between 8 and
879
dx.doi.org/10.1021/np100843j |J. Nat. Prod. 2011, 74, 877–881

Journal of Natural Products
NOTE
Figure 3. Theoretical possibility of conversion of 8 into 4a via a [1a,3s]-sigmatropic rearrangement and the experimental result.
acetonide 4a as described in the literature.1 Theoretically, it is
possible via a [1a,3s]-sigmatropic rearrangement under thermal
conditions according to the Woodward-Hoffmann rules.18
Important orbitals in the hypothetical [1a,3s]-sigmatropic pro-
cess are illustrated in Figure 3. When acetonide formation from 8
was re-examined with 2,2-dimethoxypropane in dry CH2Cl2 at
room temperature in the presence of a catalytic amount of
pyridinium para-toluenesulfonate (PPTS),1 no reaction occur-
red. The same reaction at higher temperature (60-80 °C)
resulted in the formation of a complex mixture that did not
include 4a. Therefore, the formation of 4a as originally reported1
is still unexplained. It is possible that the starting compound as
presented in the prior work was not actually 4, but rather 8.
However, it is clear from our studies that pericosine Do (8) and
pericosine D (4) are independent compounds.
In conclusion, we have elucidated the relative configuration
and the absolute configuration of pericosine Do (8) to be methyl
(3R,4S,5S,6S)-6-chloro-3,4,5-trihydroxy-1-cyclohexene-1-carboxy-
late through this synthetic approach. Chiral-phase HPLC analysis
proved that natural 8 exists as a pure enantiomer. We also conclude
that pericosine Do is a compound distinct from the structure
originally assigned as pericosine D.
trifluoroacetic acid (0.5 mL). After stirring for 1 h at room temperature,
the solvent was removed under reduced pressure directly to give a crude
product, which was purified by silica gel column chromatography
(EtOAc/hexane, 1:1 to 2:1) to afford (-)-8 (5.2 mg, 96%). (-)-8:
Colorless oil, [R]25D -6.5 (c 0.56, EtOH); IR (liquid film) νmax 3409
(OH), 1716 (CdO), 1644 (CdC) cm-1; 1H and 13C NMR data are
provided in Table 1; HRMS m/z 223.0365 [M þ H]þ (calcd for
C8H12O535Cl, 223.0372). Reported data1 of natural 1 (designated as
pericosine D in the original paper): [R]D þ1.9 (c 1.05, EtOH)
(remeasured specific rotation in this study; [R]D -6.6 (c 0.51, EtOH));
oil; IR (liquid film) νmax 3332 (OH), 1720 (CO), 1635 (CdC) cm-1
;
1H and 13C NMR data are provided in Table 1; HRMS m/z 223.0365
[M þ H]þ (calcd for C8H1235ClO5, 223.0372).
Synthesis of Racemic Chlorohydrins 4a0, 6a, 8a, and 16a.
To a mixture of rac-epoxides 14 and 15 (ca. 3:1, 285.2 mg, 1.1 mmol) in
dry Et2O (1 mL), which was prepared following the procedure in a
previous report,13 was added 1 M HCl in Et2O (1.2 mL, 1.2 mmol) at
0 °C. After stirring overnight, the reaction mixture was evaporated
directly to afford a crude mixture, which was purified by column
chromatography (EtOAc/hexane/MeOH, 2:7:1) and preparative TLC
(EtOAc/hexane/MeOH, 2:7:1) to give rac-6a (153.0 mg, 47%), rac-4a0
(2.9 mg, 1%), rac-8a (14.3 mg, 4%), rac-16a (23.7 mg, 7%), and
recovered rac-15 (12.9 mg, 5%). rac-8 was prepared from rac-8a in
the same way as described above.
Chiral-Phase HPLC Analysis of Synthetic Racemic 8, (-)-8,
and Natural 8. HPLC enantioseparation of 8 was carried out with
Chiralpak IA as the analytical column (0.46 cm i.d. ꢀ 25 cm L) at 40 °C
under isocratic elution using n-hexane/EtOH (70:30%, v/v) at a flow
rate of 1.0 mL/min and UV detection at 216 nm. Analytical samples of
8 were injected in 30 μL portions (500 ppm in n-hexane/EtOH, 70:30
(v/v)). Optical rotations were monitored at the same time at 426 nm by
LED (light-emitting diode).
’ EXPERIMENTAL SECTION
General Experimental Procedures. The optical rotation was
measured on a JASCO DIP-1000 polarimeter. IR spectra were obtained
with a JEOL FT/IR-680 Plus spectrometer. NMR spectra were recorded
at 27 °C on a Varian UNITY INOVA-500 spectrometer (1H at 500 MHz
and 13C at 125 MHz) in acetone-d6 with tetramethylsilane (TMS) as the
internal reference. HRMS values were determined using a JEOL JMS-
700 (2) mass spectrometer. Liquid column chromatography was con-
ducted over silica gel (Silicycle, SiliaFlash F60, 230-400 mesh).
Analytical TLC was performed on precoated Merck aluminum sheets
(DC-Alufolien Kieselgel 60 F254), and the compounds were viewed by
spraying an EtOH solution of phosphomolybdic acid, followed by
heating. Chiral-phase HPLC analysis was carried out with a Shimadzu
Prominance equipped with a Shimadzu CTO-20AC (50 μL sample
loop) and a Shimadzu SPD-20A UV-vis detector. Chiralpak IA (Daicel
Chemical Industries, Japan) was used as the analytical column.
Methyl (3R,4S,5S,6S)-3-Chloro-4,5,6-trihydroxy-1-cyclohex-
ene-1-carboxylate (16). Chlorohydrin 16a (12.0 mg, 0.040 mmol) was
dissolved in MeOH (0.45 mL) and trifluoroacetic acid (0.5 mL). After
stirring for 3 h at rt, the solvent was removed under reduced pressure
directly to give a crude product that was purified by preparative TLC
(eluent: EtOAc) to afford (-)-16 (4.4 mg, 50%). (-)-16: colorless oil;
[R]25 -8.1 (c 0.44, EtOH); IR (liquid film) νmax 3377 (OH), 1700
D
1
(CdO), 1644 (CdC) cm-1; H NMR (acetone-d6, 500 MHz) δ 6.76
(1H, d, J = 2.7 Hz, H-2), 4.71 (1H, ddd, J = 8.0, 2.7, 0.6 Hz, H-3), 4.55 (1H,
dd, J = 3.6, 0.8 Hz, H-6), 4.04 (1H, dd, J = 8.0, 2.5 Hz, H-4), 4.01 (1H, dd,
J = 3.6, 2.5 Hz, H-5), 3.76 (3H, s, COOCH3); 13C NMR (acetone-d6,
125 MHz) δ 166.8 (C, C-7), 139.2 (CH, C-2), 132.9 (C, C-1), 74.7 (CH,
(-)-Pericosine Do: Methyl (3R,4S,5S,6S)-6-Chloro-3,4,5-
trihydroxy-1-cyclohexene-1-carboxylate (8). Chlorohydrin 8a
(7.4 mg, 0.024 mmol) was dissolved in MeOH (0.5 mL) and
880
dx.doi.org/10.1021/np100843j |J. Nat. Prod. 2011, 74, 877–881

Journal of Natural Products
NOTE
C-5), 72.7 (CH, C-4), 68.7 (CH, C-6), 60.4 (CH, C-3), 52.2 (CH3, C-8);
HRMSm/z223.0371 [M þ H]þ (calcd for C8H1235ClO5, 223.0372);m/z
225.0334 [M þ H]þ (calcd for C8H1237ClO5, 225.0334).
(16) Arnone, A.; Cardillo, R.; Nasini, G.; Vajna de Pava, O. Tetra-
hedron 1993, 49, 7251–7258.
(17) Meier, R.-M.; Ganem, B. Tetrahedron 1994, 50, 2715–2720.
(18) Tu, C.; Berchtold, G. A. J. Org. Chem. 1993, 58, 6915–6916.
(19) Woodward, R. B.; Hoffmann, R. The Conservation of Orbital
Symmetry; Verlag Chemie, 1970.
’ ASSOCIATED CONTENT
1
Supporting Information. Copies of H and 13C NMR
S
b
spectra of natural and synthesized 8 and compound 16. This
material is available free of charge via the Internet at http://pubs.
acs.org.
’ AUTHOR INFORMATION
Corresponding Author
*Tel: þ81-72-690-1087. Fax: þ81-72-690-1087. E-mail: usa-
mi@gly.oups.ac.jp.
’ ACKNOWLEDGMENT
We are grateful to Dr. J. Onouchi of ICR Center, Daicel
Chemical Industries, Niigata, Japan, for chiral-phase HPLC
analysis. We also thank Dr. K. Minoura, Ms. M. Fujitake, and
Dr. T. Yamada of this University for NMR and MS measure-
ments and the gift of natural pericosine Do (8) sample and its 1H
and 13C NMR spectra, plus useful discussions. In addition, we
thank Prof. A. G. Griesbeck of the Institute of Organic Chem-
istry, Cologne, Germany, for useful information on (6S)-6-
hydroxy-5-epishikimic acid (21). This work was supported in
part by a Grant-in-Aid from “Dousoukai” of Osaka University of
Pharmaceutical Sciences to Y.U.
’ REFERENCES
(1) Yamada, T.; Iritani, M.; Ohishi, H.; Tanaka, K.; Doi, M.;
Minoura, K.; Numata, A. Org. Biomol. Chem. 2007, 5, 3979–3986.
(2) Usami, Y.; Ueda, Y. Synthesis 2007, 3219–3225.
(3) Usami, Y.; Takaoka, I.; Ichikawa, H.; Horibe, Y.; Tomiyama,
Y.; Otsuka, M.; Imanishi, Y.; Arimoto, M. J. Org. Chem. 2007,
68, 6127–6134.
(4) Usami, Y.; Mizuki, K.; Ichikawa, H.; Arimoto, M. Tetrahedron:
Asymmetry 2008, 19, 1461–1464.
(5) Usami, Y.; Suzuki, K.; Mizuki, K.; Ichikawa, H.; Arimoto, M. Org.
Biomol. Chem. 2009, 7, 315–318.
(6) Usami, Y.; Ohsugi, M.; Mizuki, K.; Ichikawa, H.; Arimoto, M.
Org. Lett. 2009, 11, 2699–2701.
(7) Boyd, D. R.; Sharma, N. D.; Acaru, C. A.; Malone, J. F.; O’Dowd,
C. R.; Allen, C. C.; Stevenson, P. J. Org. Lett. 2010, 12, 2206–2209.
(8) Bach, G.; Breiding-Mack, S.; Grabley, S.; Hammann, P.; Huetter,
K.; Thiericke, R.; Uhr, H.; Wink, J.; Zeeck, A. Liebigs Ann. Chem.
1993, 241–50.
(9) Shing, T. K.; Cheng, H. M. Synlett 2010, 142–144.
(10) Zhang, C.-S.; Stratmann, A.; Block, O.; Br€uckner, R.; Podeschwa,
M.; Altenbach, H.-J.; Wehmeier, U. F.; Piepersberg, W. J. Biol. Chem.
2002, 277, 22853–22862.
(11) Zhang, C.-S.; Podeschwa, M.; Block, O.; Altenbach, H.-J.;
Piepersberg, W.; Wehmeier, U. F. FEBS Lett. 2003, 540, 53–57.
(12) Griesbeck, A. G.; Maria, C.; Neudoerfl, J. ARKIVOC 2007
216–223.
(13) Usami, Y.; Okada, Y.; Yamada, T. Chirality, in press, DOI:
10.1002/chir.20937.
(14) Huntley, C.; Frederick, M.; Wood, H. B.; Ganem, B. Tetra-
hedron Lett. 2000, 41, 2031–2034.
(15) Kitamura, E.; Hirora, A.; Nakagawa, M.; Nakamura, M.;
Nozaki, H.; Tada, T.; Nukina, M.; Hirota, H. Tetrahedron Lett. 1990,
30, 4605–4608.
881
dx.doi.org/10.1021/np100843j |J. Nat. Prod. 2011, 74, 877–881