A biomimetic total synthesis of (+)-ainsliadimer A

Article abstract of DOI:10.1021/ol101705j

Figure Presented. A protecting group free and biomimetic total synthesis of (+)-ainsliadimer A has been accomplished in 14 steps from α-santonin. The synthesis relies on a hydrogen bonding promoted [4 + 2]-hetero-Diels-Alder dimerization to afford the key homodimer intermediate, which demonstrates the feasibility of using nonenzymatic conditions to achieve the proposed biosynthesis.

Full text of DOI:10.1021/ol101705j

ORGANIC
LETTERS
2010
Vol. 12, No. 19
4284-4287
A Biomimetic Total Synthesis of
(+)-Ainsliadimer A
Chao Li,† Xueliang Yu,† and Xiaoguang Lei*,†,‡
School of Pharmaceutical Science and Technology, Tianjin UniVersity,
Tianjin 300072, China, and National Institute of Biological Sciences (NIBS),
Beijing 102206, China
leixiaoguang@nibs.ac.cn
Received July 21, 2010
ABSTRACT
A protecting group free and biomimetic total synthesis of (+)-ainsliadimer A has been accomplished in 14 steps from r-santonin. The synthesis
relies on a hydrogen bonding promoted [4 + 2]-hetero-Diels-Alder dimerization to afford the key homodimer intermediate, which demonstrates
the feasibility of using nonenzymatic conditions to achieve the proposed biosynthesis.
Structurally unique and complex natural products have histori-
cally presented challenges to the art and science of organic
synthesis.1 Biomimetic synthesis toward these natural products
can provide numerous advantages,2 such as high yield, good
stereo- and chemoselectivity, and exclusion of protecting
groups.3 Recently, Zhang and co-workers discovered a new
sesquiterpene lactone dimer, (+)-ainsliadimer A (1) (Figure 1),
with an unprecedented carbon skeleton from Ainsliaea macro-
cephala, which is used in traditional Chinese medicine for the
treatment of various diseases, including angina and rheumatoid
arthritis.4 The structure of 1 was confirmed by spectroscopic
analysis as well as X-ray crystallography. Its unique heptacyclic
ring system includes 11 contiguous stereogenic centers and a
highly functionalized cyclopentane ring with three quaternary
carbons connecting the two monomeric sesquiterpene lactone
Figure 1. (+)-Ainsliadimer A.
units. These unique architectural features combined with potent
inhibitory effect against the production of NO in RAW264.7
stimulated by lipopolysaccharide LPS (IC50 ) 2.41 µg/mL)
render 1 an ideal target for total synthesis. Herein, we report
the first total synthesis of (+)-ainsliadimer A based on a concise
and biomimetic approach.
† Tianjin University.
‡ National Institute of Biological Sciences (NIBS).
(1) For an excellent review, see: Nicolaou, K. C.; Vourloumis, D.;
Winssinger, N.; Baran, P. S. Angew. Chem., Int. Ed. 2000, 39, 44.
(2) For recent reviews of biomimetic synthesis of natural products, see:
(a) Bulger, P. G.; Bagal, S. K.; Marquez, R. Nat. Prod. Rep. 2008, 25, 254.
(b) Beaudry, C. M.; Malerich, J. P.; Trauner, D. Chem. ReV. 2005, 105,
4757. (c) Yoder, R. A.; Johnston, J. N. Chem. ReV. 2005, 105, 4730.
(3) For a recent review of protecting group free synthesis, see: Young,
I. S.; Baran, P. S. Nat. Chem. 2009, 1, 193.
According to the biosynthetic hypothesis,4 a key [4 +
2]-hetero Diels-Alder cycloaddition of the naturally occurring
monomeric sesquiterpene lactone dehydrozaluzanin C (2) may
be responsible for the formation of the homodimer 3 (Scheme
1).5 Hydrolysis of the 3,4-dihydro-2H-pyran moiety under acidic
conditions should afford compound 4. Subsequent intramo-
lecular aldol reaction would eventually produce the natural
(4) Wu, Z.-J.; Xu, X.-K.; Shen, Y.-H.; Su, J.; Tian, J.-M.; Liang, S.; Li,
H.-L.; Liu, R.-H.; Zhang, W.-D. Org. Lett. 2008, 10, 2397.
10.1021/ol101705j  2010 American Chemical Society
Published on Web 09/02/2010

be prepared from the structurally related natural product
3-epizaluzanin C (5) through direct oxidation.8 3-Epizalu-
zanin C (5) could be derived from estafiatin (6).9 Accord-
ingly, both 5 and 6 have been previously synthesized from
another sesquiterpene natural product R-santonin (7). Ando
and co-workers have reported a long synthetic route to access
5 from 7.9 The key intermediate A1 was prepared from 7 in
14 steps. A base-mediated rearrangement of A1 afforded A2,
which was further transformed to 5 in seven steps. In another
case, Greene and co-workers have reported an interesting
and much more concise route to generate 6 starting from
the direct rearrangement of 7 by photoirradiation to afford
O-acetylisophotosantonic lactone (8).10 However, this route
included several problematic steps, especially the challenge
to install the trisubstituted alkene moiety on the key inter-
mediate 9. In order to efficiently and rapidly access monomer
dehydrozaluzanin C (2) from R-santonin (7), we decided to
further optimize the synthetic sequence as well as several
reaction conditions based on these two previous studies.11
Scheme 1. Proposed Biosynthesis of (+)-Ainsliadimer A
product (+)-ainsliadimer A. It was suggested that the central
hetero-Diels-Alder cycloaddition might be catalyzed by a
Diels-Alderase.6 Recently, the enzymatic catalysis of the
Diels-Alder reaction in the biosynthesis of natural products
has been attracting increasing attention.7 Therefore, it is
appealing to investigate this topic in the biosynthesis pathway
of 1. Inspired by this biosynthesis proposal, we set out to
synthesize monomer 2 and to study its dimerization to 3.
Our retrosynthetic analysis for the synthesis of monomer
dehydrozaluzanin C (2) is depicted in Figure 2. According
to the previously reported method, dehydrozaluzanin C could
The synthesis of dehydrozaluzanin C (2) began with
R-santonin (7), a commercially available material (Scheme
2). Photolysis of 7 with a high-pressure Hg lamp (500 W)
Scheme 2. Synthesis of Dehydrozaluzanin C
in acetic acid afforded O-acetylisophotosantonic lactone (8).12
We have modified the reaction procedure for the large-scale
preparation of 8 with consistent yield (38-40%).11 Hydro-
genation of the double bond smoothly provided compound
10 in quantitative yield and excellent diastereoselectivity.
Selective reduction of ketone 10, followed by mesylation of
the corresponding hydroxyl group and in situ anti-elimina-
Figure 2. Retrosynthetic analysis of dehydrozaluzanin C.
Org. Lett., Vol. 12, No. 19, 2010
4285

tion, afforded alkene 12 (52% over two steps).13 We also
observed a small amount of undesired disubstituted alkene
byproduct. Saponification of 12 with 5% aq. KOH in ethanol
cleanly generated alcohol 13. The kinetic dehydration of 13
turned out to be a little challenging. The normal condition
using SOCl2 and Et3N as base at -78 °C generated an
inseparable mixture of di- and tetrasubstituted alkenes (4:
1), which impeded the further synthesis. Through careful
reaction optimization by screening different bases, we were
pleased to find that, by using DABCO as base, we could
reproducibly access the kinetically controlled dehydrated
product disubstituted alkene 9 in greater than 12 to 1 ratio.
Selenenylation of 9, followed by oxidation of the resulting
selenide and subsequent selenoxide elimination, led to the
formation of the desired R-alkylidene-γ-butyrolactone 14.9
Epoxidation of 14 on the convex face afforded estafiatin (6)
as the major diastereomer.10 Treatment of 6 with aluminum
isopropoxide in toluene under microwave condition gave the
R-allylic alcohol 3-epizaluzanin C (5), which was further
oxidized by Dess-Martin periodinane oxidation to afford
dehydrozaluzanin C (2).
dinarily important advancement for organic synthesis.15
Inspired by the elegant work of Rawal et al. for hydrogen-
bond-promoted hetero-Diels-Alder reactions,16 we began to
investigate the potential of hydrogen bond donor catalysis.
Initially, the Diels-Alder dimerization was carried out in a
number of hydrogen-bonding solvents (CH3OH, H2O, DMF,
CHCl3, iPrOH, tBuOH, 0.1 M) at 35 °C for 12 h. We only
observed a small amount of the desired dimer 3 in CHCl3.16d
We also observed that the addition of triethylamine or a
catalytic amount of HCl did not accelerate the dimerization,
which ruled out the possibility of involvement of HCl as a
catalyst. However, prolonged reaction time or elevated
reaction temperature only led to more decomposition.
Encouraged by this result, we further examined several
commonly used hydrogen bond donor catalysts (Table 1).
Table 1. Hydrogen Bonding Promoted HDA Dimerization
With monomer 2 in hand, we initiated studies on hetero-
Diels-Alder dimerization to test the proposed biosynthesis
(Scheme 1). Although spontaneous hetero-Diels-Alder cy-
cloadditions of R-alkylidene ketones have been reported in
the literature,14 we did not observe any desired dimer 3 when
monomer 2 was allowed to stand at 20 °C without solvent
for 2 weeks.6b If the reaction temperature was elevated to
60 °C, we identified a significant amount of decomposed
material as well as partially recovered monomer 2. Next,
we examined the feasibility of using Lewis acid catalysis
for the hetero-Diels-Alder cycloaddition. Exposure of
monomer 2 to several Lewis acids such as Et2AlCl, BF3·OEt2,
SnCl4, TiCl4, Yb(OTf)3, and Sc(OTf)3 in THF or CH2Cl2 or
under neat condition at various temperatures from -78 to
40 °C led to either no reaction or decomposition.
conversion
(%)a
yield
(%)b
entry
1
catalysts
CSA
decomposed
1,1′-binaphthyl-2,2′-
diylhydrogenphosphate
N,N′-dimethylurea
N,N′-dimethylthiourea
N,N′-diphenylguanidinium
BArF salt
2
3
4
trace
trace
trace
5
trace
trace
trace
trace
80
6
TADDOL
7
phenol
8
4-bromophenol
ꢀ-naphthol
9
18
64
71
70
10
11
12
(()-BINOL
79
(+)-BINOL
86
(-)-BINOL
85
Hydrogen bonding catalysis that can mimic the action of
enzymes or antibodies has recently emerged as an extraor-
a 1.0 equiv of catalyst (BINOL 0.5 equiv), neat, 50 °C, 60 h. The reaction
mixture was analyzed by HPLC-MS, and the conversion was calculated by
the integration of UV peaks. b All yields are isolated yields.
(5) For examples of biomimetic synthesis of natural products through
Diels-Alder dimerization, see: (a) Lumb, J.; Trauner, D. J. Am. Chem.
Soc. 2005, 127, 2870. (b) Li, C.; Johnson, R. P.; Porco, J. A., Jr. J. Am.
Chem. Soc. 2003, 125, 5095. (c) Lei, X.; Johnson, R. P.; Porco, J. A., Jr.
Angew. Chem., Int. Ed. 2003, 42, 3913.
As the data show, the cycloadditions are accelerated to a
much greater extent by using ꢀ-naphthol or (()-BINOL
(entries 9 and 10, Table 1), in which the OH group is
expected to form hydrogen bond to the ketone moiety. We
observed that (()-BINOL represented better activation effect
to afford homodimer 3 as a single stereoisomer in 64% yield
along with 21% of recovered monomer 2. The relative
stereochemistry of dimer 3 was confirmed by 2D-NMR
analysis.11 The high facial selectivity could be explained by
the fact that the bulky seven-membered ring of dienophile
blocks the ꢀ-face, and as a result, the diene approaches to
(6) (a) Ref 4. (b) Huang, S.-X.; Xiao, W.-L.; Li, L.-M.; Li, S.-H.; Zhou,
Y.; Ding, L.-S.; Lou, L.-G.; Sun, H.-D. Org. Lett. 2006, 8, 1157.
(7) For excellent reviews, see: (a) Oikawa, H.; Tokiwano, T. Nat. Prod.
Rep. 2004, 21, 321. (b) Stocking, E. M.; Williams, R. M. Angew. Chem.,
Int. Ed. 2003, 42, 3078.
(8) Macias, F. A.; Galindo, J. C.; Molinillo, G. J. M. G.; Castellano, D.
Phytochemistry 2000, 54, 165.
(9) (a) Ando, M.; Ibayashi, K.; Minami, N.; Nakamura, T.; Isogai, K.
J. Nat. Prod. 1994, 57, 433. (b) Ando, M.; Kusaka, H.; Ohara, H.; Takase,
K.; Ymaoka, H.; Yanagi, Y. J. Org. Chem. 1989, 54, 1952.
(10) Edgar, M. T.; Greene, A. E.; Crabbe, P. J. Org. Chem. 1979, 44,
159.
(11) For details, see Supporting Information.
(12) (a) Zhang, W.; Luo, S.; Fang, F.; Chen, Q.; Hu, H.; Jia, X.; Zhai,
H. J. Am. Chem. Soc. 2005, 127, 18. (b) Blay, G.; Cardona, L.; Garcia, B.;
Lahoz, L.; Pedro, J. R. J. Org. Chem. 2001, 66, 7700. (c) Barton, D. H. R.;
De Mayo, P.; Shafiq, M. J. Chem. Soc. 1957, 929.
(15) For a recent review, see: Doyle, A. G.; Jacobsen, E. N. Chem. ReV.
2007, 107, 5713.
(16) (a) Unni, A. K.; Takenaka, N.; Yamamoto, H.; Rawal, V. H. J. Am.
Chem. Soc. 2005, 127, 1336. (b) Thadani, A. N.; Stankovic, A. R.; Rawal,
V. H. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 5846. (c) Huang, Y.; Unni,
A. K.; Thadani, A. N.; Rawal, V. H. Nature 2003, 424, 146. (d) Huang,
Y.; Rawal, V. H. J. Am. Chem. Soc. 2002, 124, 9662.
(13) Piers, E. B.; Cheng, K. F. Chem. Commun. 1969, 562.
(14) (a) Esmieu, W. R.; Worden, S. M.; Catterick, D.; Wilson, C.; Hayes,
C. J. Org. Lett. 2008, 10, 3045. (b) Gossinger, A. C. E.; Kalb, R.;
Orglmeister, E.; Schwaiger, J. Tetrahedron 2007, 63, 8326.
4286
Org. Lett., Vol. 12, No. 19, 2010

the dienophile from the less hindered R-face (Scheme 1).
The higher yield using BINOL rather than ꢀ-naphthol
indicates that the bidentate nature of BINOL may help orient
two monomers to facilitate the [4 + 2]-hetero-Diels-Alder
cycloaddition. Because monomer 2 is a chiral substrate, it
would be of interest to probe if one of the enantiomers of
BINOL is a more efficient catalyst. Interestingly, we found
either enantiomer of BINOL showed a slightly enhanced
activation effect compared with the racemate (entries 10-12,
Table 1). This result might be rationalized by the fact that
on the solid state racemic BINOL could be composed of the
centrosymmetric structure characterized by intermolecular
hydrogen bonds formed by a pair of (+)- and (-)- enanti-
omers, which reduces the hydrogen bonding effect on
monomer 2.17 Further reaction condition studies showed that
prolonged reaction time or elevated reaction temperature only
led to more decomposition. These results implicated that this
type of dimerization might be catalyzed by an enzyme, such
as Diels-Alderase.
reaction temperature only led to monomer 2 through retro-
Diels-Alder reaction, as well as decomposition. These
thermolysis experiments rule out the possible occurrence of
degenerate [3,3]-sigmatropic rearrangement.
Completion of the total synthesis of 1 is illustrated in
Scheme 3. Hydrolysis of compound 3 under mild acidic
Scheme 3. Total synthesis of (+)-Ainsliadimer A
condition afforded compound 4. In the final step of the
synthesis, the intramolecular aldol reaction of 4 was best
accomplished by treatment with a large excess of DBU in
dilute CH2Cl2 (0.0008 M) at 20 °C, which efficiently
generated the final natural product 1 in 89% yield. Synthetic
1 was confirmed to be identical with data reported for natural
ainsliadimer A by 1H and 13C NMR spectra,and high-
resolution mass spectrum (HRMS). The optical rotation of
synthetic 1 was [R]D +42 (c ) 0.002, CHCl3), while natural
1 was [R]D +47 (c ) 0.05, CHCl3), therefore confirming
that the absolute configuration of synthetic 1 was consistent
with natural 1.
It was anticipated that under the thermo conditions for [4
+ 2]-hetero-Diels-Alder cycloaddition there might be a
degenerate [3, 3]-sigmatropic rearrangement18 that could
occur within dimer 3 between the ketone carbonyl oxygen
and the enol ether distal alkene carbon (Figure 3). The
In conclusion, a protecting group free and biomimetic total
synthesis of (+)-ainsliadimer A has been accomplished in
14 steps. The synthesis demonstrates the feasibility of using
nonenzymatic conditions to achieve the proposed biosyn-
thesis. The dimerization represents the first example of
hydrogen bonding mediated [4 + 2]-hetero-Diels-Alder
cycloaddition of both electron-deficient diene and dienophile.
The mechanistic basis for this transformation and its synthetic
utility are under investigation. Further studies toward the
synthesis of additional dimeric sesquiterpene lactones as well
as biological evaluation of ainsliadimer A and congeners are
in progress and will be reported in due course.
Figure 3. Possible degenerate [3,3]-sigmatropic rearrangement of
dimer 3.
Acknowledgment. We thank Prof. Richard Hsung (Uni-
versity of Wisconsin at Madison), Dr. Shun Su (Scripps
Research Institute), and Mr. Houhua Li (NIBS) for helpful
discussions, Ms. Mingyan Zhao and Ms. Rui Liu (NIBS)
for NMR and HPLC-MS analysis, and Dr. Jiang Zhou
(Peking University) for HRMS analysis. Financial support
fromtheNationalHighTechnologyProjects863(2008AA022317)
and NSFC (20802050) is gratefully acknowledged.
occurrence of degenerate rearrangement may be detectable
by isotopic labeling.19 To gain evidence for this rearrange-
ment and its ease, we prepared a deuterium-labeled com-
pound 15 under basic conditions.20 When dimer 15 was
heated (neat, 50 °C, 60 h), starting material was recovered,
and the rearranged product 3′ was not observed. Elevated
(17) Tamura, R.; Takahashi, H.; Ushio, T.; Nakajima, Y.; Hirotsu, K.;
Toda, F. Enantiomer 1998, 3, 149.
Supporting Information Available: Experimental pro-
cedures and characterization data for all new compounds.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(18) Valiulin, R. A.; Dressen, D. G.; Riggs, J. R.; Habte, F. M.;
Kutateladze, A. G. J. Org. Chem. 2009, 74, 3484.
(19) Baldwin, J. E.; Broline, B. M. J. Am. Chem. Soc. 1978, 100, 4599.
(20) (a) Coumbarides, G. S.; Eames, J.; Suggate, M. J.; Weerasooriya,
N. J. Labelled Compd. Radiopharm. 2006, 49, 641. (b) For the detailed
preparation procedure and characterization for 15, see: Supporting Informa-
tion.
OL101705J
Org. Lett., Vol. 12, No. 19, 2010
4287