Synthesis and structural revision of phomopsin B, a novel polyketide carrying a 10-membered cyclic-ether ring

Article abstract of DOI:10.1021/ol2011117

Total synthesis of the proposed structure 2 for phomopsin B was achieved by using an intramolecular olefin metathesis as a key step. The spectral data, however, did not match with those of the natural product reported. Re-examination of the reported NMR data led to the structural revision of phomopsin B to known dothiorelone A 18. The R configuration of dothiorelone A was determined by total synthesis through a cross-metathesis with a chiral olefin 19.

Full text of DOI:10.1021/ol2011117

ORGANIC
LETTERS
2011
Vol. 13, No. 13
3360–3363
Synthesis and Structural Revision of
Phomopsin B, a Novel Polyketide Carrying
a 10-Membered Cyclic-Ether Ring
Yuko Izuchi,†,‡ Hiroyuki Koshino,*,† Yayoi Hongo,† Nobuhiro Kanomata,‡ and
Shunya Takahashi*,†
RIKEN, Wako, Saitama 351-0198, Japan, and Department of Chemistry and
Biochemistry, Waseda University, 3-4-1, Ohkubo, Shinjuku-ku, Tokyo 169-8555, Japan
koshino@riken.jp; shunyat@riken.jp
Received April 27, 2011
ABSTRACT
Total synthesis of the proposed structure 2 for phomopsin B was achieved by using an intramolecular olefin metathesis as a key step. The spectral
data, however, did not match with those of the natural product reported. Re-examination of the reported NMR data led to the structural revision of
phomopsin B to known dothiorelone A 18. The R configuration of dothiorelone A was determined by total synthesis through a cross-metathesis
with a chiral olefin 19.
In 2008, Lin et al. isolated novel aromatic polyketides
from the mangrove endophytic fungus, Phomopsis sp.
ZSU-H76 obtained from the South China Sea, and named
them phomopsins A and B.1 Their structures were eluci-
oxonine or oxecine ring fused to resorcinol, respectively, as
shown in Figure 1. Although there are many macrocyclic
lactones such as zearalenone in nature,2 medium-sized
cyclic phenol ethers such as 1 and 2 were virtually
unknown.3 In connection with our studies on heterocyclic
natural products,4 the unique structure stimulated our
interest and the total synthesis was planned. Described
herein is the first total synthesis of phomopsin B that
dictates revision of the formula 2 to 18.
In the synthetic study of this unique molecule, construction
of the oxecine skeleton5 fusedtoethylm-hydroxyphenylacetate
(3) (a) Raduranin A: Asakawa, Y.; Toyota, M.; Takemoto, T.
Phytochemistry 1978, 17, 2005–2010. (b) Helianane: Harrison, B.;
Crews, P. J. Org. Chem. 1997, 62, 2646–2648. (c) Halogenated deriva-
Figure 1. Structures of phomopsins A (1) and B (2).
ꢀ ꢀ
tives of helianane: Martın, M. J.; Berrue, F.; Amade, P.; Fernandez, R.;
´
Francesch, A.; Reyes, F.; Cuevas, C. J. Nat. Prod. 2005, 68, 1554–1555.
(4) (a) Izuchi, Y.; Kanomata, N.; Koshino, H.; Hongo, Y.; Nakata,
T.; Takahashi, S. Tetrahedron: Asymmetry 2011, 22, 246–251. (b) Ye, Y.-
Q.; Koshino, H.; Onose, J.; Yoshikawa, K.; Abe, N.; Takahashi, S. Org.
Lett. 2009, 11, 5074–5077. (c) Takahashi, S.; Takahashi, R.; Hongo, Y.;
Koshino, H.; Yamaguchi, K.; Miyagi, T. J. Org. Chem. 2009, 74, 6382–
6385. (d) Nishii, Y.; Higa, T.; Takahashi, S.; Nakata, T. Tetrahedron
Lett. 2009, 50, 3597–3601.
(5) Simple medium-sized ring ethers fused to an aromatic ring were pre-
pared by an intramolecular Williamson reaction of a phenol derivative
tethering an alkyl chain with a halide atom at the terminal position. However,
the method involved a terminal olefin obtained by E2 elimination as a side
product. See: (a) Illuminati, G; Mandolini, L; Masci, B. J. Org. Chem. 1974,39,
2598–2600. (b) Mandolini, L.; Masci, B. J. Org. Chem. 1977, 42, 2840–2843.
dated by spectroscopic methods, mainly by the 1D and 2D
NMR spectroscopic technique, to be 1 and 2 possessing an
† RIKEN.
‡ Waseda University.
(1) Huang, Z.; Cai, X.; Shao, C.; She, Z.; Xia, X.; Chen, Y.; Yang, J.;
Zhou, S.; Lin, Y. Phytochemistry 2008, 69, 1604–1608.
(2) For review, see: (a) Betina, V. Zearalenone and its Derivatives in
Mycotoxins: Chemical, Biological and Environmental Aspects; Elsevier:
Amsterdam, 1989. (b) Shipchandler, M. T. Heterocycles 1975, 3, 471–520
and references cited therein.
r
10.1021/ol2011117
Published on Web 05/26/2011
2011 American Chemical Society

is the main problem throughout the synthetic course. We
planned to utilize an intramolecular olefin metathesis
reaction6 as a key step to build up the skeleton and designed
a synthetic strategy as shown in Scheme 1. Cleavage of
the 10-membered ring in the target molecule leads to the
tetrasubstituted benzene 3. Releasing the pentenyl group
and changing the oxidation level can revert 3 back to the
benzaldehyde derivative 4. This would be synthesized from
methyl benzoate derivative 57 through an O-protection and
selective reduction. These retrosynthetic analyses allowed us
to prepare a real structure 18 of phomopsin B by exchanging
the O-alkyl group at the C-2 position of 5.
elongation reaction alsogaveunsatisfactoryresults. There-
fore, the saturated carbonyl was tentatively reduced and
Scheme 2
Scheme 1
Synthesis of the proposed structure 2 began with selec-
tiveO-alkylation of2,4-dihydroxybenzoicacidderivative5
(Scheme 2). Thus, 5 was initially treated with 1.0 equiv of
benzylalcohol in the presence of diisopropyl azodicarbox-
ylate (1.0 equiv)8 and triphenylphosphine (1.0 equiv) to
afford 4-O-benzyl derivative 6. Second O-alkylation into
the 2-position of 6 was performed by using 3-buten-2-ol in
the presence of a large amount of Mitsunobu reagent to
provide 7 in good yield. For our purpose, a chain elonga-
tion at the unsaturated carbonyl group was required.
However, discrimination of the two carbonyl group in 7
by conventional methods was difficult. For example,
MnO2 or BaMnO4 oxidation of a diol obtained from
DIBALH reduction of 7 resulted in a mixture of a desired
hydroxy aldehyde, its cyclized hemiacetal, and the corre-
sponding lactone in a variety of ratios. Attempts to
prepare9 the corresponding homophthalic anhydride or
isocoumarin derivative from 7 and their use for the chain
the formyl group in the resulting unstable compound was
protected as dimethylacetal to give 8. This was trans-
formed into 4 via 9 through a reductionꢀoxidation pro-
cess. The chain elongation reaction from 4 was performed
by the action of a Grignard reagent prepared from 4-pen-
tenyl bromide to provide alcohol 10 in good yield. Jones
oxidation of 10 was accompanied by hydrolysis of the
dimethylacetal moiety, and upon ethylation the resulting
carboxylic acid led to ethyl ester 3. An intramolecular
olefin metathesis of 3 was affected by the action of a first
generation Grubbs catalyst (20 mol %) in dichloro-
methane under diluted conditions (4 mM) at rt to give
oxecine 1110 in 47% yield along with open-chain dimers
(21%).12 Although more diluted conditions were utilized,
(6) (a) Grubbs, R. H. Handbook of Metathesis, Vols. 1ꢀ3; Wiley-
VCH: 2003. (b) Synthesis of (()-helianane with a 8-membered ring via
olefin metathesis was reported. See: Stefinovic, M.; Snieckus, V. J. Org.
Chem. 1998, 63, 2808–2809.
(7) Langer, P.; Kracke, B. Tetrahedron Lett. 2000, 41, 4545–4547.
(8) Mitsunobu, O. Synthesis 1981, 1–28.
(9) (a) Watanabe, M.; Date, M.; Furukawa, S. Chem. Pharm. Bull.
1989, 37, 292–297. (b) Kim, S.; Fan, G.-J.; Lee, J.; Lee, J. J.; Kim, D. J.
Org. Chem. 2002, 67, 3127–3130. (b) Bauta, W. E.; Lovett, D. P.;
Cantrell, W. R.; Burke, B. D. J. Org. Chem. 2003, 68, 5967–5973.
(10) The Z-geometry was determined by the coupling constant value
(J = 11 Hz) of two olefinic protons in the 1H NMR spectra in C6D6 not
in CDCl3.11
Org. Lett., Vol. 13, No. 13, 2011
3361

the yield of the desired product was not improved.13 The
use of a second generation Grubbs catalyst provided 11
and a cyclic dimer12 in 25 and 23% yield, respectively. The
structure of 11 was confirmed by NMR and MS analyses.
The oxecine 11 was hydrogenated over 10% Pd(OH)2
under a hydrogen atmosphere to give the proposed struc-
ture 2 in good yield.
of the natural product are likely to be those of an acyclic
compound. The singlet signal at 3.79 ppm derived from
H-14 of the natural product also reflects that the circum-
stance around the isolated methylene protons is equal,
suggesting the side chain including the ketone group is not
cyclic. Based on the results and the fact that phomopsin C1
with an acyclic chain was isolated from the same origin, we
estimated that the real structure of natural phomopsin B
may be phenol 18 with an acyclic chain. Although the
molecular formula of the natural product was determined
to be C18H24O5 by HR-EIMS at m/z 320.1613 (calcd
320.1618), the data would be explained by those of the
corresponding dehydration product. By using the CAS
database, we found that our proposed structure was a
known compound named dothiorelone A. Su et al.15,16
had isolated it17 from a similar mangrove endophytic
fungus, Dothiorella sp. HTF3 in 2004. Subsequently,
Lin et al. and Lin, Wu, and Proksch et al. reported its isola-
tion from Phomopsis sp. ZSU-H7616ꢀ18 and Rhizophora
mucronata,17,19 respectively. Recently, this natural product
was also isolated from Cytospora sp. by Abreu et al.16,17,20
An exact comparison of the data, however, was impossible
since the NMR spectra of dothiorelone A were measured
in DMSO-d6 or acetone-d6. For confirmation of our
hypothesis, we synthesized 18 as follows (Scheme 3). The
phenol 5 was benzylated under the Mitsunobu conditions
togive 12.9b According tothe method described above, this
compound was transformed into aldehyde 14 via 13.
Reaction of 14 with pentenylmagnesium bromide gave
15, which was oxidized by a Jones reagent, affording 16
after esterfication. Cross metathesis of 16 with 3-buten-2-
ol in the presence of a second generation Grubbs catalyst
afforded 17 in good yield. Reductive debenzylation of 17
withPd(OH)2 inethylacetategave18ingood yield. The 1H
and 13C NMR datainthe several solvents reportedandMS
data of 18 were well matched with those of the natural
product cited in the literature.21 Therefore, the proposed
structure of phomopsin B was revised to be 18, showing
that phomopsin B is identical to dothiorelone A.22
Scheme 3
As shown in Table 1S,14 1H and 13C NMR data of 2 did
not match with those of the natural product in the
literature.1 In particular, the splitting pattern of protons
adjacent to carbonyl carbons was quite different. Thus,
H-7 of natural phomopsin B was reported to be δ 2.84 as a
triplet (J = 7.5 Hz) whereas that of a synthetic sample was
observed at 2.59 and 2.88 ppm as dt (J = 14.2 and 5.5 Hz)
and ddd (J = 14.2, 10.0, and 5.9 Hz), respectively. In
general, such protons on the ring of a simple cycloalkenone
with a chiral center did not seem to be equivalent, thus
resulting in a pair of complicated signals with different
chemical shifts in the 1H NMR spectra. Therefore, the data
(15) Xu, Q.; Wang, J.; Huang, Y.; Zheng, Z.; Song, S.; Zhang, Y.; Su,
W. Acta Oceanolog. Sin. 2004, 23, 541–547.
(16) No [R]D value was denoted.
(17) Its absolute configuration is not shown.
(18) Huang, Z.; Guo, Z.; Yang, R.; Yin, X.; Li, X.; Luo, W.; She, Z.;
Lin, Y. Chem. Nat. Compd. 2009, 45, 625–628.
(19) Xu, J.; Kjer, J.; Sendker, J.; Wray, V.; Guan, H.; Edrada, R.;
Mueller, W. E. G.; Bayer, M.; Lin, W.; Wu, J.; Proksch, P. Bioorg. Med.
Chem. 2009, 17, 7362–7367. In this paper, the erroneous name “dothior-
elone B” is used for dothiorelone A.
(20) Abreu, L. M.; Phipps, R. K.; Pfenning, L. H.; Gotfredsen, C. H.;
Takahashi, J. A.; Larsen, T. O. Tetrahedron Lett. 2010, 51, 1803–1805.
(21) We found that some signals around aromatic and carbonyl
groups in 18 are sensitive to dissolved solvent or additive trace solvent
in CDCl3 solution. Complete NMR assignments in CDCl3, CDCl3 with
water or MeOH, DMSO-d6, and acetone-d6 are summarized in Tables 2S
and 3S; see Supporting Information.
(22) The sign of [R]D of both natural products was reported to be
negative although there was a discrepancy with their magnitude between
refs 1 and 19.
(23) (a) Laughton, C. A.; Bradshaw, T. D.; Gescher, A. Int. J. Cancer
1989, 44, 320–324. (b) Evans, P. A.; Leahy, D. K. J. Am. Chem. Soc.
2002, 124, 7882–7883. (c) Reddy, G. V.; Kumar, R. S. C.; Sreedhar, E.;
Babu, K. S.; Rao, J. M. Tetrahedron Lett. 2010, 51, 1723–1726.
(11) Takahashi, S.; Satoh, H.; Hongo, Y.; Koshino, H. J. Org. Chem.
2007, 72, 4578–4581.
(12) For the structures of open-chain and cyclic dimers and their
structural determination, see Supporting Information.
(13) GrubbsꢀHoveyda catalysts first and second generation were
also employed for the metathesis. However, the desired product was
obtained in a low yield (4ꢀ10%).
(14) See Supporting Information.
3362
Org. Lett., Vol. 13, No. 13, 2011

natural product. We found that the methylene protons at
C-6 and C-14 in the 1H NMR spectra of 1 showed a triplet
at 2.85 ppm and singlet at δ 3.76, respectively, and
concluded that the structure of natural phomopsin A also
would not be the oxonine derivative reported but an ethyl
phenylacetate derivative carrying a long side chain with a
secondary alcohol. The compound was identified as
dothiorelone B15,18 by the CAS database. Taking into
account the origin and the spectral data, it is likely that
the structure of phomopsin A should be revised to be
dothiorelone B. The 13C NMR data of phomopsin A,1
however, are not completely identical to those of dothior-
elone B.15,18 We suggest that reinvestigation of the assign-
ments of dothiorelone B is necessary; see Supporting
Information.
Scheme 4
In summary, we achieved the first total synthesis of the
proposed structure 2 for phomopsin B and ent-dothior-
elone A (S)-18, thus leading to the conclusion that pho-
mopsin B is identical to dothiolerone A. The R
configuration of natural dothiolerone A was determined
by chiral synthesis.
In order to determine the absolute configuration of
dothiorelone A, a cross metathesis with chiral olefin 1923
was also performed, giving 20 (Scheme 4). Reduction of 20
under the same conditions that were used for 17 caused a
partial cleavage of an allylic benzyl ether moiety. Hence,
the double bond in 20 was initially reduced with diimide
and then the resulting compound underwent hydrogeno-
Acknowledgment. This work was supported by the
Chemical Genomics Project (RIKEN) and in part by a
Grant-in-Aid for Scientific Research from the Ministry of
Education, Science, Sports, and Culture of Japan.
22
lysis, giving (S)-18 {[R]D þ3.4 (c = 0.50, methanol)}. The
sign of its [R]D was opposite to those of the natural
phomopsin B1 and dothiorelone A {lit.19 [R]D ꢀ6 (c =
0.1, methanol)}. Therefore, the absolute configuration of
dothiorelone A was determined to be R.
Supporting Information Available. Experimental pro-
cedures, NMR spectra of 2ꢀ4, 6ꢀ18, 20, and dimers. This
material is available free of charge via the Internet at
http://pubs.acs.org.
Structural revision of phomopsin B then focused our
attention on the proposed structure of phomopsin A and
1
made us re-examine the H and 13C NMR data of the
Org. Lett., Vol. 13, No. 13, 2011
3363