Ivorenolide B, an immunosuppressive 17-membered macrolide from Khaya ivorensis: Structural determination and total synthesis

Article abstract of DOI:10.1021/ol500667d

Ivorenolide B (1), an unprecedented 17-membered macrolide featuring conjugated acetylenic bonds and four chiral centers, was isolated from Khaya ivorensis. The structure of 1 was fully determined by spectroscopic analysis and total syntheses of its four m

Full text of DOI:10.1021/ol500667d

Letter
pubs.acs.org/OrgLett
Ivorenolide B, an Immunosuppressive 17-Membered Macrolide from
Khaya ivorensis: Structural Determination and Total Synthesis
Yao Wang,†,§ Qun-Fang Liu,‡,§ Ji-Jun Xue,† Yu Zhou,‡ Huang-Chao Yu,† Sheng-Ping Yang,‡ Bo Zhang,‡
Jian-Ping Zuo,‡ Ying Li,*,† and Jian-Min Yue*,†,‡
†State Key Laboratory of Applied Organic Chemistry, College of Chemistry and Chemical Engineering, Lanzhou University, Lanzhou
730000, China
‡State Key Laboratory of Drug Research, Shanghai Institute of Materia Medica, Chinese Academy of Sciences, 555 Zuchongzhi Road,
Shanghai 201203, China
S
* Supporting Information
ABSTRACT: Ivorenolide B (1), an unprecedented 17-
membered macrolide featuring conjugated acetylenic bonds
and four chiral centers, was isolated from Khaya ivorensis. The
structure of 1 was fully determined by spectroscopic analysis
and total syntheses of its four most possible stereoisomers.
Compound 1 showed significant immunosuppressive activity.
Khaya ivorensis A. Chev. (Meliaceae), an African ethnobotanical
also cultivated in China, contains the limonoids as the major
metabolites1 and has been demonstrated to exhibit anti-
inflammatory and antimalarial activities.2 Our recent study on
K. ivorensis led to the isolation of an immunosuppressive 18-
membered macrolide, ivorenolide A.3 As a continuation of our
studies of immunosuppressive agents from natural sources, an
unprecedented 17-membered macrolide, ivorenolide B (1)
(Figure 1), featuring conjugated acetylenic bonds and four
total syntheses of its stereoisomers as well as the evaluation of
their immunosuppressive activity.
Ivorenolide B (1) was obtained as a colorless, gumlike
material with [α]25 −13.0 (c 0.126, MeOH). The molecular
D
formula, C18H24O4, was established by an HREIMS ion at m/z
304.1666 [M+] (calcd for C18H24O4, 304.1675) and indicated
seven indices of hydrogen deficiency. The IR spectrum revealed
the presence of hydroxy (3419 cm−1) and carbonyl (1739
cm−1) groups. In accordance with the molecular formula, 18
carbon signals were resolved in the 13C NMR spectrum (Table
S1, Supporting Information), which included a carbonyl (δ
173.2), a methyl (δ 9.8), eight sp3 methylenes, four oxygenated
sp3 methines (δ 56.8, 62.0, 62.8, and 65.7), and four sp
1
quaternary carbons (δ 70.0, 70.0, 79.0, and 80.5). In the H
NMR spectrum (Table S1, Supporting Information), a broad
singlet at δ 8.48, which showed no correlation with any carbon
in the HSQC spectrum, was assigned to the hydroxy group.
The aforementioned functionalities accounted for 5 degrees of
unsaturation, indicating that compound 1 is bicyclic.
Figure 1. Structures of ivorenolides A and B (1).
1
The H−1H COSY data established the presence of two
proton-bearing and spin-coupled structural units, A (C-2 to C-
11) and B (C-16 to C-18), as drawn in bold (Figure 2).
Analysis of the HMBC data (Figure 2) showed the
connectivity between units A and B, the four sp quaternary
carbons, and the ester group, which form the 17-membered
macrolide backbone of 1, indicating that it belongs to the ring-
constrained series of ivorenolide A.3 As shown in Figure 2, the
ester carbonyl C-1 (δ 173.2) was shown to be attached to C-2
by the HMBC correlations of H2-2 and H2-3 to C-1. The
presence of a 9,10-epoxy moiety was evident by the diagnostic
chemical shifts of CH-9 (δC 56.8, δH 3.11) and CH-10 (δC 62.0,
δH 3.60) and corroborated by the multiple HMBC correlations
oxygenated chiral centers, has been further isolated from the
same material. The presence of the conjugated acetylenic bonds
and the 9,10-epoxy group on the northern hemisphere of the
molecule results in a planar loop conformation for the 17-
membered macrolide of 1, which made the stereochemical
assignment, particularly the configuration of C-16, very
challenging. In addition, the gumlike physical characteristic of
1 prevented the use of single-crystal X-ray diffraction. To
determine the absolute configuration of 1, a concise and
efficient total synthetic strategy, utilizing ring-closing metathesis
(RCM) as the key step, was carried out to produce four of its
most possible stereoisomers 1a−d, which allowed the
unambiguous assignment of its absolute configuration. We
report herein the isolation and structural elucidation via the
Received: March 4, 2014
Published: March 25, 2014
© 2014 American Chemical Society
2062
dx.doi.org/10.1021/ol500667d | Org. Lett. 2014, 16, 2062−2065

Organic Letters
Letter
Scheme 1. Retrosynthetic Analysis of Ivorenolide B (1)
Figure 2. 1H−1H COSY (), selected HMBC (H→C), and key
ROESY (↔) correlations of 1.
of H2-8 to C-9 and C-10, H-10 to C-9 and C-11, and H-11 to
C-10. The hydroxy group was assigned to CH-11 (δC 62.8, δH
4.87) by the downfield chemical shifts and the mutual HMBC
correlations of H-11 to C-10, and H-10 to C-11. The HMBC
correlations of H-11 to C-12 (δC 80.5) and C-13 (δC 70.0) and
of H-16 to C-14 (δC 70.0) and C-15 (δC 79.0) indicated that a
conjugated alkynyl moiety (C-12 to C-15, four sp quaternary
carbons) was bridged between C-11 and C-16.3,4 The presence
of a 1,16-lactone was indicated by the strong HMBC
correlation from H-16 to C-1 and was supported by the
downfield chemical shift of H-16 (δ 5.49). The planar structure
of 1 was thus determined. The ROESY (Figure 2) correlation
Bromoalkyne 4a was produced from known starting material
7a5 using N-bromosuccinimide (NBS) and AgNO3 (Scheme
2).
Scheme 2. Preparation of 4a
between H-9 and H-10 and the small coupling constant (J9,10
=
4.4 Hz) revealed that H-9 and H-10 are cis-configured, and they
were arbitrarily assigned the β-orientation. Consequently, the
coupling constant (J10,11 = 7.6 Hz) between H-10 and H-11 and
the ROESY correlation between H-11 and H-8b (δ 1.38)
indicated an α-configuration for H-11. The relative stereo-
chemistry of C-16 could not be assigned with the available data.
Compared with ivorenolide A,3 the highly similar NMR data
from C-9 to C-11 in two compounds supported a trans-
orientation for the 9,10-epoxy and 11-OH in 1, which thus
indicated either a 9R,10S,11R configuration or a 9S,10R,11S
configuration. The structure of ivorenolide B (1) features an
unprecedented 17-membered macrolide that incorporates
conjugated acetylenic bonds and four chiral centers.
To fully determine the stereochemistry of 1, a synthetic
strategy (Scheme 1) involving a ring-closing metathesis (RCM)
as the key step was implemented. This would allow the rapid
synthesis of the four most likely stereoisomers, including 1a
(9R,10S,11R,16S), 1b (9R,10S,11R,16R), 1c (9S,10R,11S,16R),
and 1d (9S,10R,11S,16S). This strategy will also facilitate the
modular construction of a small library of stereo analogues to
study the immunosuppressive activity. With respect to the
retrosynthetic analysis of 1a and 1b, it was envisioned that 1a
(1b) could be constructed via the asymmetric epoxidation of 2a
(2b), which could be accessible from 3a (3b) through ring-
closing metathesis (RCM). The cyclization precursor 3a (3b)
could be constructed through the assembly of bromoalkyne 4a
(4b) and alkyne 5a (5b) via the Cadiot−Chodkiewicz coupling
reaction. Finally, 5a (5b) could be obtained by acylation of
alcohol 6a (6b) with commercially available 9-decenoic acid.
The synthesis of 8a commenced with acylation of known
alcohol 6a6 with commercially available 9-decenoic acid in the
presence of DCC and DMAP. Compound 8a was transformed
to 5a by removal of the trimethylsilyl (TMS) group. With 4a
and 5a in hand, we next focused on the preparation of 3a,
which was obtained in good yield in the presence of Cu(I) via
the Cadiot−Chodkiewicz coupling.7 In an attempt to facilitate
the ring-closing metathesis (RCM)8 of 3a, we first employed
the Grubbs second-generation catalyst in CH2Cl2 at room
temperature; however, no product was formed, even when the
reaction was refluxed for 24 h. When we utilized toluene as the
solvent and heated the reaction to 80 °C for 24 h, the desired
RCM product 2a (Z/E = 1/1.6) was obtained in 90% yield.
Interestingly, when using the first-generation Grubbs catalyst
and refluxing in CH2Cl2, the desired product 2a was obtained
in better yield (Table 1). Oxidation of 2a (Z-form) with m-
chloroperoxybenzoic acid (m-CPBA) afforded 9a as the sole
product; it is assumed that the m-CPBA approached the double
bond from the sterically less hindered face of the molecule.9
Finally, removal of the TBDPS ether afforded 1a (Scheme 3).
Similarly, by using 6b (R-form) as the starting alcohol,
compound 1b was synthesized.
The conjugated acetylenic bonds and the 9,10-epoxy moiety
cause the molecule to adopt a planar loop conformation. Thus,
the NMR data (Table S1, Supporting Information) of the four
stereoisomers were predicted to possess high similarities. To
determine the absolute configuration of 1 and extend the scope
of our medicinal chemistry studies, compounds 1c and 1d
(Figure 3), the enantiomers of 1a and 1b, respectively, were
2063
dx.doi.org/10.1021/ol500667d | Org. Lett. 2014, 16, 2062−2065

Organic Letters
Letter
a
The specific rotations [α]25 = −13.0 (for 1, c 0.126, MeOH),
Table 1. Optimization of the RCM Conditions with 3a
D
−8.3 (for 1a, c 0.15, MeOH), and +9.7 (for 1c, c 0.3, MeOH)
further indicated that the absolute configuration of ivorenolide
B (1) was identical to that of the synthetic 1a
(9R,10S,11R,16S). This conclusion was confirmed by CD
analysis (Figure 4), in which the CD curve of 1 matched that of
entry
catalyst
solvent temp (°C) time (h)
yield (%)
1
2
3
4
Grubbs II
Grubbs II
Grubbs II
Grubbs I
CH2Cl2
CH2Cl2
toluene
CH2Cl2
rt
43
80
43
24
24
24
24
0
0
90 (Z:E = 1:1.6)
93 (Z:E = 1:1.5)
a
Reactions were performed in the indicated solvents (0.001 M) under
argon, and the catalysts were added in one portion.
Scheme 3. Synthesis of 1a and 1b
Figure 4. CD spectra of compounds 1 and 1a−d.
1a and was opposite to 1c that is the enantiomer of 1a, and the
CD curves of two enantiomers 1b and 1d were opposite, which
was consistent with their specific rotations (Supporting
Information). Thus, the absolute configuration of 1 was
determined to be 9R,10S,11R,16S.
Ivorenolide B (1) and the synthetic compounds 1a−d were
then evaluated for immunosuppressive activity. Compound 1
and its synthetic equivalent 1a exhibited significant inhibition of
LPS-induced B-cell proliferation, while the synthetic enan-
tiomer 1c and stereoisomers 1b and 1d were inactive (Table 2).
Table 2. Immunosuppressive Effects of 1 and 1a−d on
Murine Lymphocyte Proliferation Induced by ConA (5 μg/
a,b
mL) or LPS (10 μg/mL)
ConA-induced T-cell
proliferation
LPS-induced B-cell
proliferation
compd
CC50 (μM)
IC50 (μM)
SI
IC50 (μM)
SI
synthesized from 4b (R-form of 4a) using the same strategy
(Supporting Information).
1
20.7
21.7
N
>10
>10
N
−
−
−
−
−
7.2
8.5
N
2.9
2.6
1a
1b
1c
1d
−
−
−
N
N
N
N
N
N
a
The selectivity index (SI) is defined as the ratio of the concentration
of the compound that reduced cell viability to 50% (CC50) to the
concentration of the compound needed to inhibit the proliferation by
b
50% relative to the control value (IC50). N indicates inactive, which
was defined as inhibition rates lower than 30% at 10 μM.
Figure 3. Structures of 1c and 1d.
The NMR spectra of compounds 1 and 1a−d showed high
similarities with slight variations of the proton and/or carbon
chemical shifts of C-1, CH2-3, CH2-8, CH-9, CH-16, and CH3-
17 (Table S1, Supporting Information). Extensive comparison
showed that the NMR spectra of compounds 1, 1a, and 1c
completely matched, indicating that their structures were the
same and/or enantiomers, and the NMR spectra of 1b also well
matched those of 1d, confirming that they were enantiomers.
In conclusion, this study identified an immunosuppressive
agent, ivorenolide B (1), isolated from K. ivorensis. Compound
1 featured an unprecedented 17-membered macrolide with
conjugated acetylenic bonds and four chiral centers. The
absolute configuration of 1 was determined through asym-
metric total synthesis, which was successfully achieved in six
steps with a 26% overall yield from the known compounds.
Compared to our recent 12-step synthesis of ivorenolide A,3
2064
dx.doi.org/10.1021/ol500667d | Org. Lett. 2014, 16, 2062−2065

Organic Letters
Letter
Castells, J. Angew. Chem., Int. Ed. 2006, 45, 6086−6101. (d) Prunet, J.
Angew. Chem., Int. Ed. 2003, 42, 2826−2830.
(9) (a) Salamci, E. Tetrahedron 2010, 66, 4010−4015. (b) Kumar, V.
P.; Kavitha, N.; Chandrasekhar, S. Eur. J. Org. Chem. 2013, 6325−
6334.
this approach is more concise and effective involving a Cadiot−
Chodkiewicz coupling and a ring-closing metathesis reaction as
the key steps to construct the macrolide scaffold. With the same
synthetic strategy, the enantiomer 1c and two diastereoisomers
1b and 1d were also synthesized, thereby allowing the
assignment of the absolute stereochemistry of 1. Both the
natural product 1 and its synthetic equivalent 1a showed
remarkable immunosuppressive activities. This finding and the
formally reported immunosuppressive ivorenolide A have
provided novel macrolide scaffolds for the development of
immunosuppressive agents.
ASSOCIATED CONTENT
* Supporting Information
■
S
Detailed synthetic information as well as spectroscopic
characterization of all compounds. This material is available
free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Authors
■
*E-mail: liying@lzu.edu.cn.
*E-mail: jmyue@mail.shcnc.ac.cn.
Author Contributions
§These authors contributed equally.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We are grateful for financial support from the National Natural
Science Foundation of China (NSFC; nos. 21072084 and
21272099). We thank Prof. Y. K. Xu of Xishuangbanna
Tropical Botanical Garden, CAS, for identification of the plant
material.
REFERENCES
■
(1) (a) Adesida, G. A.; Adesogan, E. K.; Okorie, D. A.; Taylor, D. A.
H.; Styles, B. T. Phytochemistry 1971, 10, 1845−1853. (b) Taylor, D.
A. H. Phytochemistry 1977, 16, 1847−1849. (c) Vanucci, C.; Lange, C.;
Lhommet, G.; Dupont, B.; Davoust, D.; Vauchot, B.; Clement, J. L.;
Brunk, F. Phytochemistry 1992, 31, 3003−3004. (d) Zhang, B.; Yang, S.
P.; Yin, S.; Zhang, C. R.; Wu, Y.; Yue, J. M. Phytochemistry 2009, 70,
1305−1308.
(2) Agbedahunsi, J. M.; Fakoya, F. A.; Adesanya, S. A. Phytomedicine
2004, 11, 504−508.
(3) Zhang, B.; Wang, Y.; Yang, S. P.; Zhou, Y.; Wu, W. B.; Tang, W.;
Zuo, J. P.; Li, Y.; Yue, J. M. J. Am. Chem. Soc. 2012, 134, 20605−
20608.
(4) Pretsch, E.; Buhlmann, P.; Affolter, C. Structure Determination of
̈
Organic Compounds: Tables of Spectral Data, 3rd ed.; Springer: New
York, 2000; p 88. Rong, G. B.; Zhu, S. Z. Trans., East China Univ. Sci.
Technol.: Shanghai 2002, 88.
(5) Ghosh, S.; Pradhan, T. K. Synlett 2007, 15, 2433−2435.
(6) (a) Denmark, S. E.; Yang, S. M. J. Am. Chem. Soc. 2004, 126,
12432−12440. (b) Lian, X. D.; Ma, S. M. Angew. Chem., Int. Ed. 2008,
47, 8255−8258. (c) Mori, M.; Nakai, T. Tetrahedron Lett. 1997, 38,
6233−6236.
(7) (a) Cho, E. J.; Kim, M.; Lee, D. Org. Lett. 2006, 8, 5413−5416.
(b) Ko, E.; Liu, J.; Perez, L. M.; Lu, G.; Schaefer, A.; Burgess, K. J. Am.
Chem. Soc. 2011, 133, 462−477. (c) Li, Z.; Fowler, F. W.; Lauher, J.
W. J. Am. Chem. Soc. 2009, 131, 634−643. (d) Shi Shun, A. L. K.;
Tykwinski, R. R. Angew. Chem., Int. Ed. 2006, 45, 1034−1057.
(8) (a) Xie, J.; Ma, Y. L.; Horne, D. A. Chem. Commun. 2010, 46,
4770−4772. (b) Nicolaou, K. C.; Roschangar, F.; Vourloumis, D.
́
Angew. Chem., Int. Ed. 1998, 37, 2014−2045. (c) Gradillas, A.; Perez-
2065
dx.doi.org/10.1021/ol500667d | Org. Lett. 2014, 16, 2062−2065