Total synthesis of incarvilleatone and incarviditone: Insight into their biosynthetic pathways and structure determination

Article abstract of DOI:10.1021/ol302205w

A concise biomimetic total synthesis of incarvilleatone and incarviditone is achieved in one pot via the highly stereoselective hetero-and homodimerization of (±)-rengyolone, respectively. The structure of incarviditone is revised on the basis of spectroscopic and computational evidence.

Full text of DOI:10.1021/ol302205w

ORGANIC
LETTERS
2012
Vol. 14, No. 18
4878–4881
Total Synthesis of Incarvilleatone and
Incarviditone: Insight into Their Biosynthetic
Pathways and Structure Determination
Kun Zhao,† Gui-Juan Cheng,‡ Hongzhi Yang,† Hai Shang,† Xinhao Zhang,‡
Yun-Dong Wu,*,‡ and Yefeng Tang*,†
The Comprehensive AIDS Research Center, Department of Pharmacology &
Pharmaceutical Sciences, School of Medicine, Tsinghua University, Beijing 100084,
China, and Lab of Computational Chemistry and Drug Design, Laboratory of Chemical
Genomics, Peking University Shenzhen Graduate School, Shenzhen 518055, China
chydwu@ust.hk; yefengtang@tsinghua.edu.cn
Received August 8, 2012
ABSTRACT
A concise biomimetic total synthesis of incarvilleatone and incarviditone is achieved in one pot via the highly stereoselective hetero- and
homodimerization of (()-rengyolone, respectively. The structure of incarviditone is revised on the basis of spectroscopic and computational evidence.
Dimeric natural products have been recognized as a
group of attractive targets for synthetic chemists, because
of their fascinating molecular architectures and prominent
biological profiles.1 Not surprisingly, significant efforts
have been devoted to their synthesis. Among various
approaches to access the dimeric targets, bioinspired
strategies which usually exploit a biomimetic dimerization
reaction as the key step are of particular interest, since they
take advantage of the symmetrical elements of the targets
and, thus, dramatically simplify their synthesis.2
† Tsinghua University
‡ Peking University
(1) For leading reviews on the topic of dimeric natural products,
see: (a) Lian, G. Y.; Yu, B. Chem. Biodiversity 2010, 7, 2660–2691.
(b) Vrettou, M.; Gray, A. A.; Brewer, A. R. E.; Barrett, A. G. M.
Tetrahedron 2007, 63, 1487–1536. (c) Voloshchuk, T.; Farina, N. S.;
Wauchope, O. R.; Kiprowska, M.; Haberfield, P.; Greer, A. J. Nat.
Prod. 2004, 67, 1141–1146.
(2) For selected examples, see: (a) Li, C.; Lobkovsky, E.; Porco, J. A.,
Jr. J. Am. Chem. Soc. 2000, 122, 10484–10485. (b) Lei, X.; Johnson,
R. P.; Porco, J. A., Jr. Angew. Chem., Int. Ed. 2003, 42, 3913–3917.
(c) Myers, A. G.;Herzon, S. B. J. Am. Chem. Soc. 2003, 125, 12080–12081.
(d) Zhang, W.; Luo, S.; Chen, Q.; Hu, H.; Jia, X.; Zhai, H. J. Am. Chem.
Soc. 2005, 127, 18–19. (e) Nicolaou, K. C.; Lim, Y. H.; Papageorgiou,
C. D.; Piper, J. L. Angew. Chem., Int. Ed. 2005, 44, 7917–7921. (f) Baran,
P. S.; Hafensteiner, B. D.; Ambhaikar, N. B.; Guerrero, C. A.; Gallagher,
J. D. J. Am. Chem. Soc. 2006, 128, 8678–8693. (g) Movassaghi, M.;
Schmidt, M. A. Angew. Chem., Int. Ed. 2007, 46, 3725–3728. (h) Li, Z. T.;
Gao, Y. X.; Tang, Y. F.; Dai, M. J.; Wang, G. X.; Wang, Z. G.; Yang, Z.
Org. Lett. 2008, 10, 3017–3020. (i) Snyder, S. A.; Kontes, F. J. Am. Chem.
Soc. 2009, 131, 1745–1752. (j) Li, C.; Dian, L. Y.; Zhang, W. D.; Lei, X. G.
J. Am. Chem. Soc. 2012, 134, 12414–12417.
Figure 1. Structure of incarvilleatone (1) and incarviditone (2).
Incarvilleatone (1, Figure 1), a cyclohexylethanoid di-
mer, was recently isolated by Zhang et al. from incarvillea
younghuibandii, a plant used as a Chinese folk medicine
to treat dizziness and anemia.3 Preliminary biological
(3) Gao, Y. P.; Shen, Y. H.; Zhang, S. D.; Tian, J. M.; Zeng, H. W.;
Ye, J.; Li, H. L.; Shan, L.; Zhang, W. D. Org. Lett. 2012, 14, 1954–1957.
r
10.1021/ol302205w
Published on Web 09/04/2012
2012 American Chemical Society

evaluation showed that incarvilleatone displayed a potent
inhibitory effect against NO production in LPS-induced
RAW264.7 macrophages, rendering it an attractive target
for further biomedical studies. Structurally, incarvillea-
tone (1) features a highly congested, cage-shaped hexacyc-
lic skeleton, which poses a synthetic challenge to chemists.
Interestingly, incarviditone (2), another dimeric natural
product biogenetically relevant to incarvilleatone (1), was
also discovered by the same group from incarvillea delavayi.4
Biogenetically, both 1 and 2 are assumed to be derived from
(()-rengyolone (3), a coexisting substance found in the
same species as 1 and 2.3,4
As suggested by Zhang et al., incarvilleatone (1) might
be generated from the heterodimerization of (()-3 through
a sequential oxa-Michael addition and DielsꢀAlder reac-
tion (path a, Scheme 1).3 Alternatively, we envisioned that
Scheme 1. Proposed Biosynthetic Pathway of Incarvilleatone (1)
and Incarviditone (2)
Figure 2. Transition structures (with relative free energy) of the
stereodetermining step.
homodimerization of (()-3(e.g.,(þ)-3þ (þ)-3) would afford
incarviditone (2) through either a cascade oxa-Michael/
DielsꢀAlder/retro-aldol reactionsequence(pathc, Scheme1)
or a tandem oxa-Michael/Michael addition reaction (path d,
Scheme 1).
Aspart of ourcontinued interestsinthe total synthesis of
bioactive natural products with unique molecular archi-
tectures and intriguing biogenetic pathways, we launched a
program toward the synthesis of 1 and 2. Strategically, we
envisioned that the most elegant and efficient approach to
make these targets would rely on a biomimetic strategy,
through which both 1 and 2 could be generated from (()-3
in one pot via hetero- and homodimerization, respectively.
However, further analysis indicated that the biomimetic
strategy may be challenging, or even impossible. One of the
major concerns stems from the exceptional stereoselectiv-
ity involved in the initial step of the putative biosynthetic
pathways. As shown in Figure 2, there are four possible
transition structures (TS-1 to TS-4) for the oxa-Michael
addition, which could probably generate at most four
stereoisomers. To gain deeper insight into the stereo-
chemical outcomes of the dimerization processes, compu-
tational studies were performed.7 B3LYP/6-31þG* calcu-
lations revealed that, for the heterodimerization, TS-1,
which leads to the formation of incarvilleatone (1), is
6.0 kcal/mol lower in energy than TS-2. For the case of
homodimerization, TS-3, which leads to the generation of
incarviditone (2), is 7.8 kcal/mol less favorable than TS-4.
Both convex/convex facial arrangement and hydrogen
bondingwerefoundtocontributetothe energypreferences
of TS-1 over TS-2 and TS-4 over TS-3. The computational
findings rationalize the stereochemical outcome of the
heterodimerization of (()-3 but cast doubt on that of the
homodimerization process. Even though TS-3 cannot be
a cascade oxa-Michael/Michael/aldol reaction sequence
(path b, Scheme 1) could also account for the transforma-
tions from (()-3 to 1.5 Meanwhile, we postulated that a
(4) Chen, Y. Q.; Shen, Y. H.; Su, Y. Q.; Kong, L. Y.; Zhang, W. D.
Chem. Biodiversity. 2009, 6, 779–783.
(5) Although the proposed intermediate 5 has to date not been
reported as a natural product, computational studies suggested the
stepwise mechanism via a tandem Michael/aldol reaction is more likely
involved in the transformation from 4 to 1 than the concerted one via a
DielsꢀAlder reaction (for details, see SI).
~
ꢀ
ꢀ
(7) Carreno, M. C.; Gonzalez-Lopez, M.; Urbano, A. Angew. Chem.,
(6) Calculations were performed on B3LYP/6-31þG(d) with Gaussian
09. For details of discussion, see SI.
Int. Ed. 2006, 45, 2737–2741.
Org. Lett., Vol. 14, No. 18, 2012
4879

excluded in an enzyme catalyzed reaction, it would be hard
to envision its involvement in a practice of synthesis.
Instead, we hypothesized that incarviditone (2) might be
generated from the homodimerization of (()-3 via TS-4
instead of TS-3. If this hypothesis is correct, the structure
of incarviditone (2) may be incorrectly assigned in the
original isolation paper4 and require revision.
Table 1. Condition Screening of the Dimerization of
(()-Rengyolone (3)
Keeping this hypothesis in mind, we pursue the biomi-
metic total synthesis of 1 and 2. The monomer (()-3 was
initially prepared from commercially available 11 by em-
ploying the known method through a sequence of oxidative
dearomatization, reduction of the peroxy functionality,
and intramolecular oxa-Michael addition (Scheme 2).7
However, scale-up of this sequence proved to be problem-
atic, and the overall yield of (()-3 was only 20ꢀ25%.
Therefore, a more practical and scalable procedure was
developed, which involved three synthetic operations: (1)
selective TBS-protection of the primary alcohol of 11 to
afford 12, (2) dearomatization of 12 utilizing PhI(OAc)2 to
provide 13,8 and (3) deprotection of 13 with TBAF fol-
lowed by intramolecular conjugate addition to yield (()-3.
This procedure enabled the multigram-scale synthesis of
(()-3 in 50% overall yield for three steps.
yield of yield of
entry
base
solvent
1 (%)
2 (%)
1
2
3
4
5
6
7
8
DABCO (excess)
DBU (excess)
dioxane/H2O
DCM
NR
NR
15a
0a
NR
NR
17a
7a
t-BuOK (2.0 equiv)
THF
NaHMDS (2.0 equiv) THF
NaH (2.0 equiv)
NaH (2.0 equiv)
NaH (1.0 equiv)
NaH (0.5 equiv)
THF
DCM
DCM
DCM
0a
0a
25b
38b
30b
30b
40b
35b
a Refers to 1H NMR yield with 1,3,5-trimethylbenzene as the internal
standard. b Refers to isolated yield. DABCO = 1,4-diazabicyclo[2.2.2]-
octane. DBU =1,8-diazabicyclo[5.4.0]undec-7-ene. NaHMDS = sodium
bis(trimethylsilyl)amide.
the workofCarreno,11 inwhich the homodimerization and
Scheme 2. Synthesis of the Monomer (()-Rengyolone (3)
ꢀ
trimerization of a p-quinol type of substrates via successive
conjugate additions were well investigated, we tried the
NaH/DCM combination in this reaction. To our delight,
treatment of (()-3 with NaH (2.0 equiv) in DCM for
12 h afforded two major products in yields of 25% and
30%, respectively (entry 6). One of them was shown to
be carvilleatone (1) based on the good agreement of its
spectroscopic data (1H NMR, 13C NMR, and HRMS)
with those of the natural product3 as well as the X-ray
crystallographic study.12 The other, as will be discussed in
more detail below, was incarviditone.
Although only modest yields were obtained, the results
demonstrated the feasibility of the biomimetic strategy.
Encouraged by these results, we examined the effect of
other reaction parameters on the transformation. Although
substantial amounts of starting material were recovered in
the reaction listed in entry 6, efforts to improve the conver-
sion by extending the reaction time, elevating the reaction
temperature, or using a large excess of NaH only led to the
formation of some unidentified products. In sharp con-
trast, when the dimerization reaction was performed with a
reduced amount of NaH (1.0 equiv), 1 and 2 were obtained
in a satisfactory yield of 38% and 40%, respectively,
accompanied with the recovery of 10% of the starting
material (entry 7). Furthermore, we found that even sub-
stoichiometric amounts (0.5 equiv) of NaH could effectively
promote the transformation, with only a lightly decreased
yield of 1 and 2 obtained (entry 8).13 No stereoisomers of 1
With (()-rengyolone (3) in hand, we started to explore
conditions that could promote the proposed dimerization
reaction. Commonly used base/solvent combinations9 which
have been applied to promote the intermolecular oxa-
Michael addition,10 such as DABCO/dioxane/H2O, DBU/
DCM, t-BuOK/THF, NaHMDS/THF, and NaH/THF,
were examined, but none of them provided satisfying re-
sults (Table 1, entries 1ꢀ5) (for more details of condition
screening, see Supporting Information (SI)). Inspired by
(8) You, Z.; Hoveyda, A. H.; Snapper, M. L. Angew. Chem., Int. Ed.
2009, 48, 547–550.
(9) For selected examples of base-promoted intermolecular oxa-
ꢁ
Michael addition, see: (a) Dumez, E.; Rodriguez, J.; Dulcere, J.-P.
Chem. Commun. 1997, 1831–1832. (b) Enders, D.; Haertwig, A.; Raabe,
€
G.; Runsink, J. Eur. J. Org. Chem. 1998, 1771–1792. (c) Lesch, B.; Brase,
S. A. Angew. Chem., Int. Ed. 2004, 43, 115–118. (d) Nising, C. F.;
€
€
Ohnemuller, U. K.; Brase, S. A. Angew. Chem., Int. Ed. 2006, 45, 307–
309. (e) Xiong, X.; Ovens, C.; Pilling, A. W.; Ward, J. W.; Dixon, D. J.
Org. Lett. 2008, 10, 565–567.
~
(11) Carreno, M. C.; Ribagorda, M. Org. Lett. 2003, 5, 2425–2428.
(12) For X-ray crystallagraphic data of 1, see SI.
(13) For a detailed discussion on the dimerization of (()-(3) with
substoichiometric amounts of NaH, see SI.
(10) For reviews on the topic of oxa-Michael addition, see: (a)
€
Nising, C. F.; Brase, S. Chem. Soc. Rev. 2008, 37, 1218–1228. (b) Nising,
€
C. F.; Brase, S. Chem. Soc. Rev. 2012, 41, 988–999.
4880
Org. Lett., Vol. 14, No. 18, 2012

Figure 4. X-ray crystal structure of 14.
differentiated from each other and the observed NOESY
cross-peaks of HꢀC(50) with HꢀC(9), in conjunction with
Figure 3. Key NOE correlations and DFT optimized geometries
of proposed structure (2) and revised structure (8) (distances
˚
are in A).
˚
the corresponding short H--H distances (2.21 A) in the
DFT optimized structures (Figure 3), clearly suggested
that HꢀC(50) and HꢀC(60) should be oriented in a
β-configuration. Fortunately, as more direct and conclu-
sive evidence, the X-ray crystal structure of 14 (for its
preparation, see SI), a derivative of incarviditone, was
obtained, which unambiguously confirmed the above
stereochemical assignments (Figure 4).16 As such, we con-
cluded that the structure of incarviditone should be cor-
rected to 8 (Figure 3).
In summary, we have achieved a concise, biomimetic,
and one-pot total synthesis of incarvilleatone (1) and
incarviditone (8) via the highly stereoselective hetero-
and homodimerization of (()-rengyolone (3), respectively.
Furthermore, the structure of incarviditone is revised
based on computational studies and spectroscopic evi-
dence. This work showcases how a proposed biogenetic
pathway could be implemented in the laboratory environ-
ment by judicious choice of reaction conditions and, in
turn, how the practice of biomimetic synthesis can help
elucidate the underlying biosynthetic pathways of the
targets.
or 2 or the intermediates 5 and 7 (Scheme 1) were identified
under the optimal conditions, indicating that both the hetero-
and homodimerization of (()-(3) proceeded through a
cascade reaction with exceptionally high stereoselectivity.14
With enough synthetic material in hand, we beganto test
our hypothesis about the structure assignment of incarvi-
ditone. We noticed that in the original isolation paper
the relative stereochemistries of C-50 and C-60 of 2 were
assigned as an R-configuration, mainly because of the
observed cross-peaks of HꢀC(60) with HꢀC(5) and
HꢀC(50) with HꢀC(90) in the NOESY correlation. How-
ever, such assignments might be misleading due to two
facts: (1) HꢀC(60) and HꢀC(5) are located at adjacent
carbons of a five-membered ring, and thus the observed
NOESY signal between them cannot unequivocally prove
their cis-relationship; (2) the 1H NMR signal of HꢀC(90)
largely overlaps with that of HꢀC(9) in the documented
NOESY spectrum (in CDCl3), and thus the observed
cross-peak of HꢀC(50) with HꢀC(90) could be instead
attributed to that of HꢀC(50) with HꢀC(9).15 In fact,
both H--H distances of HꢀC(50)/HꢀC(90) and HꢀC(50)/
Acknowledgment. We gratefully acknowledge the fi-
nancial support from the National Science of Foundation
of China (Y.T., 21102081; Y.W., 21133002), China’s 985
project (Y.T., 411307141), New Teachers’ Fund for Doctor
Stations, Ministry of Education (Y.T., 20110002120011),
and the Shenzhen Peacock program (Y.W.). We also thank
Prof. Wei-dong Zhang (Second Military Medical Univer-
sity, China) for providing us with the spectrum of natural
incarviditone as well as helpful suggestions and Prof. Olaf
Wiest (University of Notre Dame) for insightful discussion.
˚
˚
HꢀC(9) (4.12 A or 4.48 A) in the B3LYP/6-31þG(d)
optimized geometry of 2 are too long to show the NOESY
correlation (Figure 3). Additional evidence came from
the following NMR experiments: when the NOESY spec-
trum of synthetic incarviditone was collected in CD3OD,
the signals of HꢀC(9) and HꢀC(90) could be well
(14) Following the reviewer’s suggestion, the purified incarvilleatone
and incarviditone were respectively treated with 1.0 equiv of NaH in
DCM to see whether there exists an equilibrium between them. It turned
out that no new spot could be observed on the TLC, which suggested
such an equilibrium could be precluded under the current conditions.
(15) Careful analysis of the authentic NMR spectrum of incarvidi-
tone, which was kindly provided by Prof. Zhang, revealed that the
signals of H-C(9) and H-C(90) were misassigned to each other in the
original isolation paper (for details, see SI).
Supporting Information Available. Computational de-
tails, experimental procedure, and full spectroscopic data
for all new compounds. This material is available free of
charge via the Internet at http://pubs.acs.org.
(16) CCDC 897020 contains the supplementary crystallographic data
for 14. These data can be obtained free of charge from the Cambridge
Crystallographic Data Centre via www. ccdc.cam.ac.uk/data_request/cif.
The authors declare no competing financial interest.
Org. Lett., Vol. 14, No. 18, 2012
4881