Synthesis of ent-nanolobatolide

Article abstract of DOI:10.1002/anie.201100926

Efficient and adaptable: The key steps in the total synthesis of ent-nanolobatolide, the enantiomer of the novel and potent neuroprotective agent, involve an oxidative ring expansion of (-)-menthone, a Nazarov cyclization, an intermolecular Diels-Alder re

Full text of DOI:10.1002/anie.201100926

DOI: 10.1002/anie.201100926
Natural Products
Synthesis of ent-Nanolobatolide**
Hau Man Cheng, Weiwei Tian, Philippe A. Peixoto, Bhartesh Dhudshia, and David Y.-K. Chen*
Neurodegeneration is a physiological phenomenon charac-
terized by the progressive loss either of neurons or their
function. The prevalence of neurodegenerative disorders,
such as Alzheimerꢀs disease and Parkinsonꢀs disease, has
placed an ever-increasing burden on the healthcare system as
a result of the debilitating social and economical impact of
these diseases.[1] Indeed, the search for effective therapeutic
intervention for these detrimental diseases continues to be of
high priority on the global health agenda, and the identifica-
tion of novel chemical entities that possess neuroprotective
properties has an undisputed role in this context. In 2009,
Sheu and co-workers reported the isolation and structural
elucidation of nanolobatolide,
a novel C18 terpenoid
obtained from the extracts of Formosan soft coral Sinularia
nanolobata.[2] Besides its structural intricacy, the biological
effects of nanolobatolide on the neurological system were
particularly fascinating. At a concentration of 10 mm, nano-
lobatolide not only displayed a cytotoxic effect against
microglial cells, but was also found to exhibit anti-neuro-
inflammatory activity with a 45.5% reduction in the INF-g
stimulated expression of proinflammatory protein iNOS
relative to the control cells, which were treated only with
INF-g.[2] Furthermore, nanolobatolide showed a neuropro-
tective effect in the 6-OHDA (6-hydroxydopamine) induced
neurotoxicity of neuroblastoma SH-SY5Y, with neuroprotec-
tive activities ranging between 41.4% and 83.3% across the
0.01–10 mm concentration range.[2] The novel molecular
architecture of nanolobatolide, coupled with its impressive
biological properties, prompted us to undertake its synthesis.
Herein we report the total synthesis of ent-nanolobatolide
((ꢀ)-1; Scheme 1), through a strategy that also provided
evidence in support of its biogenetic origin.
Scheme 1. Molecular structure of ent-nanolobatolide ((ꢀ)-1) and pro-
posed biosynthesis via diene acid 2 and triene 3. Retrosynthetic
analysis of ent-nanolobatolide ((ꢀ)-1) leading to epoxide 4, triene 3,
trienone 6 and (ꢀ)-menthone (7).
late-stage synthetic maneuver to complete the oxo-bridged
tetracyclic framework of the nanolobatolide structure. The
thus obtained retrosynthetic intermediate, tetracyclic epoxide
4, was expected to be derived from triene 3 through an
intermolecular Diels–Alder reaction[3] with ethyl acrylate (5)
and a subsequent chemo- and stereoselective epoxidation. By
recognizing the positions of the olefinic functionalities within
the [5,7]-bicyclic system of 3, a Nazarov reaction[4] involving
trienone 6 was envisaged for its construction. Finally, the
asymmetric information, that was present throughout the
synthesis and ultimately resulted in ent-nanolobatolide ((ꢀ)-
1), originates from the readily available chiral building block,
(ꢀ)-menthone (7).
Inspired by the speculated biosynthetic pathway that
suggested a diene acid precursor (2) and a guaiane-type triene
precursor (3),[2] a logical synthetic strategy was formulated,
the retrosynthetic analysis of which is shown in Scheme 1.
Thus, an intramolecular epoxide-opening reaction involving
the C11 carboxylate and the C7–C8 epoxide was reserved as a
As shown in Scheme 2, our initial foray towards the
synthesis of the nanolobatolide structure featured an oxida-
tive ring-expansion of (ꢀ)-menthone (7) followed by an
intramolecular Dieckmann condensation to construct the
[5,7]-bicyclic motif of enedione 11. This first-generation
strategy, as we shall see, enabled the validation of the
proposed intermolecular Diels–Alder reaction and the intra-
molecular epoxide opening, which led to the preparation of
the highly advanced tetracyclic intermediates 16 and 17.
In this case, kinetic TMS silyl enol ether formation of (ꢀ)-
menthone (7) followed by Simmons–Smith cyclopropana-
tion[5] afforded cyclopropane 8 in 85% yield over the two
steps. Oxidative ring-enlargement of cyclopropane 8 took
place smoothly in the presence of CAN and NaI,[6a] and the
initially formed b-iodo cycloheptanone underwent further
elimination in the presence of NaOAc to furnish cyclo-
heptenone 9 (86% yield over the two steps).[6b] Conversion of
enone 9 into keto ester 10 required a three-carbon homol-
ogation reaction, which was readily accomplished through
[*] Dr. H. M. Cheng, Dr. W. Tian, Dr. P. A. Peixoto, Dr. B. Dhudshia,
Prof. Dr. D. Y.-K. Chen
Chemical Synthesis Laboratory@Biopolis
Institute of Chemical and Engineering Sciences (ICES)
Agency for Science, Technology and Research (A*STAR)
11 Biopolis Way, The Helios Block, #03-08
Singapore 138667 (Singapore)
Fax: (+65)6874-5869
E-mail: david_chen@ices.a-star.edu.sg
[**] We thank Doris Tan (ICES) for assistance with high-resolution mass
spectrometry (HRMS). Financial support for this work was provided
by A*STAR, Singapore.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201100926.
Angew. Chem. Int. Ed. 2011, 50, 4165 –4168
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4165

Communications
took place smoothly to afford the corresponding bicyclic
diketone in 89% yield; this compound was treated with
PhSeCl and then H2O2 to deliver enedione 11 in 60% yield.
Gratifyingly, the proposed biomimetic Diels–Alder reaction[2]
using the lithium enolate of enedione 11 (formed upon
treatment of 11 with LDA) and ethyl acrylate (5) proceeded
smoothly to give tricycle 12 in 86% yield as a single
diastereoisomer, which exists in its enol form. Conversion of
enol ketone 12 into enone 14 required extensive screening of
reaction protocols, and ultimately a [Fe(acac)3]-mediated[7]
coupling between enol phosphate 13 (prepared from 12 upon
its treatment with ClP(O)OEt2, Et3N, and DMAP; 78%
yield) and MeMgBr was successful and gave 14 in 65% yield.
In preparation for the intramolecular epoxide-opening reac-
tion, for assembly of the tetracyclic core of the nanolobatolide
structure, enone 14 was treated with mCPBA to afford
epoxide 15 as a single diastereoisomer in 92% yield. In the
presence of TiCl4, epoxide 15 underwent lactone formation to
construct the final ring required in the nanolobatolide
structure in 55% yield. With the tetracyclic keto lactone 16
secured, its conversion into the nanolobatolide structure
merely required the final introduction of the C5–C6 trisub-
stituted olefin. Much to our disappointment, this seemly
simple transformation has remained elusive to date, primarily
because of the inability of ketone 16 to undergo enolization
under a variety of reaction conditions.[8] Reduction of the
ketone moiety in 16 afforded the corresponding secondary
alcohol 17 as a single stereoisomer, and this alcohol also
resisted further elimination through transformations involv-
ing either E1 or E2 mechanistic pathways.[9] Attempts to form
the C5–C6 trisubstituted olefin from tricyclic enone 14 or
epoxide 15, prior to the bridged lactone formation also failed.
We rationalized that the severe steric congestion experienced
at C6, which is shielded by the neighboring seven-membered
ring that contains an isopropyl substituent, and the [2,2,1]-
bicyclic system, is likely to be responsible for this late-stage
obstacle. For this reason we opted for a revised strategy that
involved an early introduction of the C5–C6 trisubstituted
olefin.
Recognizing the shortfall in our initial attempts at the
synthesis of the nanolobatolide structure, a revised strategy,
which ultimately brought our synthetic route to fruition, is
outlined in Scheme 3. Conveniently, (ꢀ)-menthone (7) once
again served as the chiral starting material, and cyclohepte-
none 20, which contains the target C5–C6 trisubstituted
olefin, was prepared by a synthetic sequence analogous to
that developed for its regioisomeric congener 9. The only
deviation in the synthesis was the initial TMS silyl enol ether
formation, which was carried out under thermodynamic
conditions. A Simmons–Smith cyclopropanation[5] of this
TMS silyl enol ether afforded cyclopropane 18 (72% yield
over the two steps from 7), which was readily transformed to
cycloheptenone 20 through the action of FeCl3 and then
NaOAc in 56% overall yield.[6b] In preparation for the
proposed Nazarov cyclization to construct the [5,7]-bicyclic
framework of dienone 22, enone 20 was converted into the
intermediate triflate 21; the cross-coupling of 21 with
tetravinyl tin in the presence of CO under palladium catalysis
([Pd(PPh3)4]) led to the trienone 6 (60% yield over the two
Scheme 2. Synthesis of advanced tetracyclic intermediates 16 and 17.
Reagents and conditions: a) LDA (0.5m in THF, 1.3 equiv), THF, ꢀ78
! 08C, 1 h; then TMSCl (1.3 equiv), 08C, 1 h, 92%; b) CH2I2
(2.0 equiv), Et2Zn (1.0m in hexanes, 2.0 equiv), hexanes, 0!258C,
12 h, 92%; c) CAN (2.0 equiv), NaI (1.0 equiv), CH3CN/H2O (8:1),
08C, 1 h; d) NaOAc (6.0 equiv), MeOH, 808C, 12 h, 86% over two
steps; e) vinylmagnesium bromide (1.4m in THF, 1.7 equiv), CuI
(0.1 equiv), THF, ꢀ788C, 1 h, 86%; f) O3, CH2Cl2, ꢀ788C, 30 min;
then PPh3 (1.0 equiv), 258C, 16 h, 100%; g) Ph3PCHCO2Et (1.2 equiv),
CH2Cl2, 408C, 12 h, 93%; h) H2, Pd/C (10 wt.% , 0.1 equiv), EtOH,
258C, 12 h, 95%; i) KOtBu (3.0 equiv), THF, 08C, 45 min, 89%;
j) PhSeCl (1.05 equiv), py (1.1 equiv), CH2Cl2, 08C, 30 min; then H2O2
(70% aq., 3.8 equiv), CH2Cl2, 08C, 30 min, 60%; k) LDA (0.2m in THF,
1.5 equiv), THF, ꢀ788C, 1 h; then ethyl acrylate (5, 2.0 equiv), ꢀ788C,
2 h, 86%; l) ClPO(OEt)2 (1.5 equiv), HMPA (1.5 equiv), Et3N
(1.5 equiv), DMAP (0.15 equiv), Et2O, 08C, 3 h, 78%; m) [Fe(acac)3]
(5.0 equiv), MeMgBr (3.0m in Et2O, 20 equiv), THF/NMP (10:1),
ꢀ308C, 15 min, 65%; n) mCPBA (1.0 equiv), CH2Cl2, 258C, 6 h, 92%;
o) TiCl4 (1.5 equiv), CH2Cl2, ꢀ78!258C, 12 h, 55%; p) NaBH4
(1.5 equiv), MeOH, 08C, 3 h, 92%. acac=acetoacetonate, CAN=ceric
ammonium nitrate, DMAP=4-methylaminopyridine, HMPA=hexa-
methylphosphoryl amide, LDA=lithium diisopropylamide, mCPBA=
meta-chloroperoxybenzoic acid, NMP=N-methylpyrrolidone, py=pyr-
idine, THF=tetrahydrofuran, TMS=trimethylsilyl.
conjugate addition, oxidative olefin cleavage, Wittig olefina-
tion, and hydrogenation, with a 76% overall yield for this
four-step transformation. An intramolecular Dieckmann
condensation of keto ester 10 in the presence of KtOBu
4166 www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 4165 –4168

(LDA/TMSCl or Et3N/TMSI), failed to give the desired
tricycle. We considered whether an alternative diene system
that could be readily synthesized from dienone 22 would
undergo a related intermolecular Diels–Alder reaction,
thereby furnishing the carbocyclic backbone of nanolobato-
lide while retaining the C5–C6 olefin. At this point, prompted
by biosynthetic considerations, triene system 3 became a
natural and enticing option. As such, we envisaged a methyl
addition/dehydration sequence to access the targeted triene
system 3 (Scheme 3). In this context, while the intermediate
tertiary alcohol 23 could be readily prepared from dienone 22
through a 1,2-addition (MeLi, ca. 1:1 d.r.), the subsequent
dehydration of 23 proved capricious and generally afforded a
complex mixture containing varying amounts (determined by
1H NMR spectroscopy) of the desired triene 3 accompanied
with regioisomeric triene systems and unidentified by-prod-
ucts. In view of the transiency and liability of the triene system
3, the proposed Diels–Alder reaction was sucessfully ach-
ieved by intercepting the in situ generated triene 3 with ethyl
acrylate (5), without isolation of 3, in the presence of
BF3·OEt2 as the Lewis acid promoter (both for the generation
of triene 3 from tertiary alcohol 23, and for the intermolecular
Diels–Alder reaction). This reaction gave tricyclic diene ester
24 in 38% overall yield from dienone 22. In accordance with
the biosynthetic pathway leading to the nanolobatolide
structure (Scheme 1), the conversion of diene ethyl ester 24
into the diene acid 2 was required. The reluctance of ethyl
ester 24 to undergo saponification necessitated a three-step
reduction/oxidation sequence to afford diene acid 2. Pleas-
ingly, we serendipitously discovered that during the NaClO2-
mediated oxidation of the aldehyde precursor to the diene
acid 2, the nanolobatolide structure could be detected (by
1H NMR spectroscopy) and was subsequently isolated in
48% yield (from 24). This unexpected one-pot transformation
suggested a facile oxidation of the initially formed diene acid
2 to epoxy acid 4 under the NaClO2 conditions, followed by
intramolecular epoxide opening/lactonization. Synthetic
nanolobatolide structure exhibited identical 1H and
13C NMR spectra and mass spectra to those reported for the
natural product,[2] and the optical rotation measurement of
Scheme 3. Total synthesis of ent-nanolobatolide ((ꢀ)-1). Reagents and
conditions: a) iPr2NH (1.2 equiv), MeMgBr (3.0m in Et2O, 1.2 equiv),
Et2O, 258C, 12 h; then 7 (1.0 equiv), TMSCl (2.0 equiv), Et3N
(2.5 equiv), 258C, 8 h; b) CH2I2 (2.4 equiv), Et2Zn (1.0m in hexanes,
2.4 equiv), hexanes, 0!258C, 12 h, 72% over the two steps; c) FeCl3
(3.0 equiv), DMF, 08C, 4 h; d) NaOAc (5.0 equiv), MeOH, 808C, 10 h,
56% over two steps; e) LHMDS (1.0m in THF, 1.2 equiv), PhNTf2
(1.2 equiv), THF, ꢀ78!08C, 2 h, 85%; f) [Pd(PPh3)4] (0.05 equiv), LiCl
(3.0 equiv), tetravinyl tin (1.5 equiv), CO, DMF, 608C, 1.5 h, 70%;
g) Sc(OTf)3 (0.2 equiv), CH2Cl2, 0!258C, 3 h, 75%; h) MeLi (1.6m in
Et2O, 5.0 equiv), THF, ꢀ78!08C, 1 h; i) BF3·OEt2 (5.0 equiv), ethyl
acrylate (5)/toluene (1:5), ꢀ20!08C, 3 h, 34% (38% brsm from 22);
j) DIBAL-H (1.0m in toluene, 5.0 equiv), CH2Cl2, ꢀ788C, 1.5 h; then
DMP (3.0 equiv), NaHCO3 (10.0 equiv), CH2Cl2, 2 h; then NaClO2
(3.0 equiv), 2-methyl-2-butene (30 equiv), tBuOH/pH 7 buffer (1:1),
18 h, 48% over three steps. brsm=based on recovered starting
material, DIBAL-H=diisobutylaluminium hydride, DMF=N,N’-di-
methylformamide DMP=Dess–Martin periodinane, LHMDS=lithium
hexamethyldisilazide, Tf=trifluoromethanesulfonyl.
the
synthetic
material
(synthetic:
25
½aꢁD ¼ꢀ10.8 degcm3 gꢀ1 dmꢀ1 (c = 0.13 gcmꢀ3, CHCl3) and
25
Ref. [2]
½aꢁD ¼ + 11.4 degcm3 gꢀ1 dmꢀ1
(c = 0.72 gcmꢀ3,
CHCl3)) implied that the naturally occurring substance (1)
should be represented as the enantiomer of the structures
shown in Scheme 1–3.[10]
In summary, the total synthesis of ent-nanolobatolide
((ꢀ)-1) has been accomplished by using an efficient strategy
that obviates the use of protecting and blocking groups. The
developed synthetic sequence from (ꢀ)-menthone (7) fea-
tured an oxidative ring-expansion, a Nazarov cyclization,[4] an
intermolecular Diels–Alder reaction,[3] and an intramolecular
epoxide-opening reaction. The successful execution of the
two latter transformations also provided support to the
speculated biosynthetic pathway of nanolobatolide.[2] In
view of the neuroprotective properties of nanolobatolide,
the work described herein (Scheme 2 and Scheme 3) should
pave way for the synthesis of a diverse array of rationally
designed nanolobatolide analogues for chemical and biolog-
steps). Pleasingly, the Nazarov cyclization[4] of trienone 6 took
place smoothly when using Sc(OTf)3, to give dienone 22 in
75% yield. Unfortunately, our previously developed reaction
conditions for the biomimetic intermolecular Diels–Alder
reaction (i.e., LDA, ethyl acrylate, 11!12) proved unsuc-
cessful with this newly prepared dienone substrate 22.
Extensive screening of reaction conditions, which included
(but were not limited to) base (KHMDS and LHMDS), and a
route through the silyl enol ether derivative of dienone 22
Angew. Chem. Int. Ed. 2011, 50, 4165 –4168
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
4167

Communications
[4] a) C. Santelli-Rouvier, M. Santelli, Synthesis 1983, 429 – 442;
b) S. E. Denmark, Comp. Org. Synth., Vol. 5, Pergamon Press,
Oxford, 1991, pp. 751 – 784; c) K. L. Habermas, S. E. Denmark,
T. K. Jones, Org. React. 1994, 45, 1 – 158.
[5] a) H. E. Simmons, T. L. Cairns, S. A. Vladuchick, C. M. Holness,
Org. React. 1973, 20, 1 – 131; b) M. J. Conia, Pure Appl. Chem.
1975, 43, 317 – 326; c) A. B. Charette, Organozinc Reagents 1999,
263 – 285; d) A. B. Charette, A. Beauchemin, Org. React. 2001,
58, 1 – 415; e) W. A. Donaldson, Tetrahedron 2001, 57, 8589 –
8627.
[6] a) J. Jiao, L. X. Nguyen, D. R. Patterson, R. A. Flowers II, Org.
Lett. 2007, 9, 1323 – 1326; b) K. Kakiuchi, M. Ue, H. Tsukahara,
T. Shimizu, T. Miyao, Y. Tobe, Y. Odaira, M. Yasuda, K. Shima, J.
Am. Chem. Soc. 1989, 111, 3707 – 3712.
[7] a) N. Hayashi, M. Nakada, Tetrahedron Lett. 2009, 50, 232 – 235;
b) J. R. Scheerer, J. F. Lawrence, G. C. Wang, D. A. Evans, J.
Am. Chem. Soc. 2007, 129, 8968 – 8969.
[8] Attempts to install the C5–C6 olefin from ketone 16 through the
corresponding enol triflate, enol phosphate, vinyl halide, and
hydrazone all proved unsuccessful, leading to either recovery or
extensive decomposition of the starting material 16.
[9] Attempts to eliminate alcohol 17 through its corresponding
triflate, selenide, halide, xanthate, or under Mitsunobu, Martinꢀs
sulfurane, Burgess reagent, and POCl3/SOCl2–pyridine condi-
tions all proved ineffective for this process.
[10] Notably, although (+)-menthone, which would lead to the
naturally occurring form of (+)-nanolobatolide (1) in accord-
ance to our synthetic strategy, is less readily available and more
expensive than (ꢀ)-menthone, both enantiomeric forms of
menthol are readily available, and this compound can be easily
oxidized to give (+) and (ꢀ)-menthone.
ical investigations. These activities are currently underway in
our laboratory.
Received: February 6, 2011
Revised: March 3, 2011
Published online: April 6, 2011
Keywords: biomimetic synthesis · cycloaddition · cyclization ·
.
natural products · neurological agents
[1] a) E. Esposito, S. Cuzzocrea, Curr. Med. Chem. 2010, 17, 2764 –
2774; b) M. M. Husain, K. Trevino, H. Siddique, S. M. McClin-
tock, Neuropsychiatr. Dis. Treat. 2008, 4, 765 – 777; c) M. R.
Loizzo, R. Tundis, F. Menichini, F. Menichini, Curr. Med. Chem.
2008, 15, 1209 – 1228; A. Lleo, Curr. Genomics 2007, 8, 550 – 558;
d) P. I. Moreira, X. Zhu, A. Nunomura, M. A. Smith, G. Perry,
Expert Rev. Neurother. 2006, 6, 897 – 910; e) C. H. Dowding,
C. L. Shenton, S. S. Salek, Drugs Aging 2006, 23, 693 – 721;
f) R. M. Wilson, S. J. Danishefsky, Acc. Chem. Res. 2006, 39,
539 – 549.
[2] Y.-J. Tseng, Z.-H. Wen, C.-F. Dai, M. Y. Chiang, J.-H. Sheu, Org.
Lett. 2009, 11, 5030 – 5032.
[3] a) K. C. Nicolaou, S. A. Snyder, T. Montagnon, G. Vassiliko-
giannakis, Angew. Chem. 2002, 114, 1742 – 1773; Angew. Chem.
Int. Ed. 2002, 41, 1668 – 1698; for the Diels–Alder reaction of
cyclopentadiene containing guaiane sesquiterpenes, see: b) G.
Lian, B. Yu, Chem. Biodiversity 2010, 7, 2660 – 2691; c) T. A.
Geissman, T. E. Winters, Tetrahedron Lett. 1968, 9, 3145 – 3147;
d) K. Vokꢁc, Z. Samek, V. Herout, F. Sorm, Tetrahedron Lett.
1968, 9, 3855 – 3857; e) J. Beauhaire, J. L. Fourrey, M. Vuil-
horgne, J. Y. Lallemand, Tetrahedron Lett. 1980, 21, 3191 – 3194.
4168
www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 4165 –4168