Efforts Toward a Synthesis of Crotogoudin and Crotobarin

Article abstract of DOI:10.1055/s-0036-1588560

Two synthesis designs for the diterpenoid crotogoudin are discussed, and efforts to achieve each are described. First, a Cope rearrangement/intramolecular Diels-Alder cascade reaction was investigated. Second, a bioinspired sequence of cationic bicyclization and A-ring oxidative fragmentation set-up for a lactonization induced by a phenolic oxidation, ultimately providing a tricyclic intermediate that required only installation of the bridging ring of the salient bicyclo[2.2.2]octane system. This last endeavor was fraught with difficulty, but did lead to the development of conditions for cyclization of related keto-alkenes via manganese(III)-based radical chemistry.

Full text of DOI:10.1055/s-0036-1588560

SYNLETT
0
9
3
6
-
5
2
1
4
1
4
3
7
-
2
0
9
6
© Georg Thieme Verlag Stuttgart · New York
2017, 28, A–E
letter
A
en
D. N. Mai et al.
Letter
Synlett
Efforts Toward a Synthesis of Crotogoudin and Crotobarin
Duy N. Mai◊
Dmitriy Uchenik◊
O
OMe
X
6 steps
(20%)
Christopher D. Vanderwal*
0
0
0
0
0-
0
0
1
7-
2
1
8
4-
5
2
1
O
O
O
HO
Department of Chemistry, 1102 Natural Sciences II,
University of California, Irvine, CA 92697-2025, USA
cdv@uci.edu
H
O
O
3 steps
X = H, crotogoudin
X = OAc, crotobarin
◊ These authors contributed equally
Published as part of the ISHC Conference Special Section
Received: 11.07.2017
Accepted after revision: 07.08.2017
Published online: 16.08.2017
herein some efforts along these lines that permitted the as-
sembly of advanced tricyclic intermediates.
Our first strategy involved a high-risk/high-reward cas-
cade reaction of polyunsaturated precursors of type 5
(Scheme 1). We assumed that installation of the exocyclic
alkene in 1 from 3, the product of an intramolecular Diels–
Alder reaction of 4, would be straightforward.4 Although
the proposed cycloaddition involved electronically poorly
matched components, the topologically similar substrate
used in Peese and Gin’s nominine synthesis gave credence
to this idea.9 The ene-lactone dienophile in 4/4′ was to arise
from a (likely unfavorable) Cope rearrangement of 5 via
conformation 5′. The idea of heating this polyunsaturated,
achiral intermediate to induce the establishment of a Cope
equilibrium, which might be siphoned off via a productive
intramolecular Diels–Alder cycloaddition, was provocative.
We made and evaluated one such system, as described in
Scheme 2.
DOI: 10.1055/s-0036-1588560; Art ID: st-2017-b0554-l
Abstract Two synthesis designs for the diterpenoid crotogoudin are
discussed, and efforts to achieve each are described. First, a Cope rear-
rangement/intramolecular Diels–Alder cascade reaction was investigat-
ed. Second, a bioinspired sequence of cationic bicyclization and A-ring
oxidative fragmentation set-up for a lactonization induced by a pheno-
lic oxidation, ultimately providing a tricyclic intermediate that required
only installation of the bridging ring of the salient bicyclo[2.2.2]octane
system. This last endeavor was fraught with difficulty, but did lead to
the development of conditions for cyclization of related keto-alkenes via
manganese(III)-based radical chemistry.
Key words diterpenoid, cascade, cationic cyclization, phenolic oxida-
tion, dearomatization, Birch reduction, Diels–Alder cycloaddition, radi-
cal cyclization
The unusual 3,4-seco-atisanes crotogoudin and croto-
barin (1 and 2, Scheme 1) were recently reported by
Rasoanaivo and co-workers.1 These compounds were cyto-
toxic to the four human tumor cell lines tested (KB, HT29,
A549, and HL60, IC50values: 0.5–2.5 μM) with nearly equal
potencies to docetaxel, the positive control, and induced
cell-cycle arrest at the G2/M transition with apparent asso-
ciated apoptosis. The underlying mechanism of activity is
not known; however, as the isolation chemists suggest, it
might well involve covalent modification of a biomolecule
involved in the cell cycle, given the presence of the unsatu-
rated ketone conjugate acceptor.2,3 The Carreira group re-
ported the first synthesis of crotogoudin via an elegant but
lengthy route (>25 steps LLS)4 that permitted the revision
of the absolute configuration of the natural product, such
that it fits into the large ent-atisane family. More recently, a
concise synthesis of racemic crotogoudin was reported by
Liu and co-workers.5 Attractive approaches have also been
published by the groups of Maier,6 Singh,7 and Jia.8 Our goal
was to obtain enough of these secondary metabolites to en-
able studies of their mechanism of action. We describe
X
IMDA
O
O
OR
O
O
O
O
O
O
3
4
1: X = H, crotogoudin
2: X = OAc, crotobarin
bond
rotation
3
5
1
6
4
2
Cope
H
O
O
O
R
R
OR
O
O
O
5
5'
4'
retro-
Diels–
Alder
Cope
IMDA
O
O
1
O
O
–
O
O
O
O
O
7
6
Scheme 1 A cascade strategy toward crotogoudin; IMDA: intramolec-
ular Diels–Alder reaction
© Georg Thieme Verlag Stuttgart · New York — Synlett 2017, 28, A–E

B
D. N. Mai et al.
Letter
Synlett
1. TsOH·H2O
2. LDA, TMSCl
then IBX, MPO
1. I2, NaOH
2. PdCl2(dppf)
t-BuLi, ZnBr2, then
PdCl2(PPh3)2
I
OTBDPS
OTBDPS
OTBDPS
OTBDPS
OTBDPS
OTBDPS
BPin
9
(66%)
11
3. LDA, PhNTf2
O
O
(37%
3 steps)
Br
OTf
OTBDPS
OTBDPS
BrZn
13
12
10
8
BPin
O
O
(77%
2 steps)
1. TMSCl, MeOH
2. DMSO, (COCl)2, Et3N
3. NaClO2, NaH2PO4
(30%
3 steps)
TFAA
(10 equiv)
Reagents and conditions:
combinations of microwave (100–250 °C),
additives (e.g., K2CO3, Et3N, AcOH),
conditions
X
O
O
OH
OTf
(34–50%)
O
and solvents (e.g., toluene, o-DCB, H2O, DMF)
OTf
O
O
OTf
O
15
14
16
Scheme 2 Approach to crotogoudin based upon a pericyclic cascade sequence. .
Alkyl iodide 8 was prepared in six steps from δ-valero-
lactone.10 It was coupled via a Negishi reaction11 with bi-
functional reagent 9,12 affording boronic ester 10. After ste-
reospecific iododeborylation,12 a second Negishi coupling
with 11 (precursor halide made in four steps from cyclo-
hexanone10) provided 12. The careful orchestration of steps
shown was needed to generate compounds of type 5 be-
cause of the reactivity of the ene-lactone; further, we found
that most 2-oxygenated 1,3-cyclohexadienes were unsta-
ble, and we could only handle the corresponding triflate.
As a result, the generation of triflate 15 from 12 was
challenging, and 15 was the only compound of general type
5 that we were able to access. The successful sequence from
12 involved ketal hydrolysis, selective ketone dehydrogena-
tion,13 and O-triflation of the cross-conjugated enolate to
afford 13. Desilylation under acidic conditions and two-step
oxidation to keto-acid 14 preceded dehydration to the sen-
sitive ene-lactone 15. Although lengthy, this sequence pro-
vided tens of milligrams of 15 for a range of attempts to
thermally isomerize it into crotogoudin-like compound 16
(conditions shown in Scheme 2).10 Unfortunately, in all cas-
es wherein reactions occurred, extensive decomposition
was observed. Wary of the many possible side reactions, in-
cluding Cope reaction with the other 1,5-diene system, and
other cycloaddition possibilities, we turned to computation
to assess the likelihood of success. Computed barriers for
both key reactions were prohibitively high, and those com-
puted for undesired pathways were lower. Had this cascade
to convert polyunsaturated, achiral intermediate 15 (or
more stable derivatives) into the polycyclic architecture of
crotogoudin succeeded, we would have attempted the reac-
tion of 6 (Scheme 1), a precursor with all of the carbon at-
oms of crotogoudin; after the Cope/IMDA cascade, a cyclor-
eversion to expel acetone might directly give crotogoudin.
Still, our synthesis would have remained lengthy, because
of the difficulties in installing the orthogonal functional-
group arrangements in these types of cascade precursor,
and preliminary computational and experimental results
were not encouraging. Enamored with the attractive struc-
ture and properties of the targets, we sought a much more
direct route.
Our second approach (Scheme 3) features steps that are
likely relevant to the biogenesis of atisane and seco-atisane
diterpenoids: a cation-π cyclization to forge the B-ring and
an A-ring oxidative cleavage, which would establish the C5–
C10 stereorelationship with perfect stereochemical control
(see 19).14,15 Formation of the δ-lactone would arise from
carboxylate addition to an electron-deficient carbon center
generated in the course of a phenolic oxidation. The end-
game, with introduction of the bridging ring, would call for
intermolecular Diels–Alder reactivity using a suitable ketene
or cumulene equivalent diene.16 Recently, a related strategy
was successfully implemented by Liu and co-workers.5
Cation-π cyclization precursor 23 was made by copper-
catalyzed coupling17 of allylic acetate and benzylic Grignard
precursors. Lewis acid catalyzed bicyclization provided 24
(along with some ortho product), which was subjected to a
bioinspired A-ring cleavage18 followed by liberation of the
phenol in 19. Oxidative dearomatization19 provided gram
quantities of tricyclic dienone 26. At this stage, monoreduc-
tion was required to afford enone 27. This reaction proved
incredibly challenging owing to competing overreduction
or rearomatization to regenerate 19. Our best results were
obtained using the procedure of Lipshutz and co-workers,20
which were still rather capricious. We uncovered a possible
dependence on oxygen for this reduction, and investigated
the relevance of dppb-monooxide21 as the key ligand for
conjugate reduction, but this approach never led to repro-
ducibility.10 Owing to these problems, we could at best ac-
cumulate ca. 50 mg of 27, which was still enough to learn of
the tremendous challenges associated with generating re-
active Diels–Alder dienes of types 28–30. All attempts to
make enol/enamine derivatives of 27 led either to ejection
of the bridging carboxylate to reform phenol 19, or to de-
composition. These outcomes were unchanged even when
the operations were performed in the presence of model di-
enophiles. The poor material throughput associated with
© Georg Thieme Verlag Stuttgart · New York — Synlett 2017, 28, A–E

C
D. N. Mai et al.
Letter
Synlett
OP
"biomimetic"
A-ring oxidative
cleavage
OP
CO2H
OR
OH
phenol oxidative
dearomatization
Diels–Alder
with ketene
FGI
1
and
2
10
5
A
B
dienone
or cumulene
equivalent
O
O
O
HO
reduction
H
H
O
O
20: from arene-terminated
cation-π cyclization
17
18
OTBDPS
19
OTBDPS
OTBDPS
OTBDPS
1. PCC
Li2CuCl4
(89%)
EtAlCl2
2. m-CPBA
+
OAc
10
O
(48%)
(70%, 2 steps)
5
O
O
HO
BrMg
H
O
H
21
23
22
24
25
(60%
1.TsOH·H2O, Δ
2 steps)
2. TBAF
decomposition
or regeneration
of phenol 19
X
O
Cu(OAc)2·H2O,
dppb, TMDS,
PhME, 0 °C
O
OH
PhI(OAc)2
10
5
CF3CH2OH
(variable)
O
O
O
(66%)
OH
O
O
O
O
28: X = OSiR3
29: X = NR2
30: X = H
27
26
19
Scheme 3 Approach to crotogoudin based upon a bioinspired polyene cyclization and A-ring oxidative cleavage; TBDPS: tert-butyldiphenylsilyl; PCC:
pyridinium dichromate; m-CPBA: meta-chloroperbenzoic acid; TBAF: tetra-n-butylammonium fluoride; TMDS: tetramethyldisilazane
this sequence limited our studies of this and related end-
games, and therefore led us to consider a revised approach
that might be more scalable.
enone conjugation, and (3) iodolactonization with sponta-
neous β-elimination of hydroiodic acid. With compound 27
in hand in larger quantities, further explorations of Diels–
Alder reactivity (as shown in Scheme 3) as well as conjugate
addition chemistry were undertaken, without success. Con-
jugate allylation (Hosomi–Sakurai and copper-mediated),
vinylation, propargylation, allenylation, and more were uni-
formly met with failure. Closely related systems have been
shown to be particularly poor participants in functionaliza-
tions of the enone β-position.23 No doubt, the developing
1,3-diaxial interaction in such reactions (see desired prod-
uct 33) is at least partly responsible for this lack of reactivity.
This impasse was unfortunate, because in the model
system shown in Equation 1, we have set a reasonable prec-
edent for completion of the tetracyclic ring system of croto-
goudin via oxidative radical cyclization. This example of a
manganese(III)-mediated ketone alkene cyclization24 is re-
markable, because the vast majority of such reaction types
appear to require doubly activated (very acidic) carbonyl
systems, such as malonates and β-keto-esters.25 Although
we have not evaluated this reaction type extensively, we
believe that these conditions could prove general for the cy-
clization of alkenes onto otherwise unactivated ketones.
The rapid synthesis of 35 from inexpensive materials (Rob-
inson annulation of cyclohexanone, Hosomi–Sakurai conju-
gate allylation26 to give 34,27 radical cyclization to 35)
shows a rather rapid increase in complexity, and it is unfor-
tunate that we were never able to effect conjugate allylation
of enone 27 to study this radical cyclization as a means to
complete a synthesis of crotogoudin.
1. PCC
CO2H
OMe
OMe
2. m-CPBA
3. Dowex 50W X8
(76%, 3 steps)
HO
H
H
31
32
1. Li, NH3
3. I2, NaHCO3
(22%, 3 steps or
26% from 31)
2. SiO2, (CO2H)2
CH2Cl2, MeOH
O
O
many
nucleophiles
x
Nu
O
O
O
O
33
27
Scheme 4 A more robust approach to tricyclic enone intermediate 27
An adjustment in strategy, wherein the order of oxida-
tion/reduction operations on the arene was changed, led to
the sequence shown in Scheme 4. Known cationic cycliza-
tion product 3122 was made in a very similar manner to 24,
but in much higher overall throughput.10 Oxidative A-ring
cleavage of 31 proceeded uneventfully to afford 32. The key
conceptual reversal of the oxidation and reduction steps
from Scheme 3 led to the successful formation of 27 via (1)
Birch reduction, (2) careful enol ether hydrolysis to avoid
© Georg Thieme Verlag Stuttgart · New York — Synlett 2017, 28, A–E

D
D. N. Mai et al.
Letter
Synlett
(9) Peese, K. M.; Gin, D. Y. J. Am. Chem. Soc. 2006, 128, 8734.
(10) Please see the Supporting Information for details.
(11) Negishi, E.-I. Acc. Chem. Res. 1982, 15, 340.
(12) Wang, C.; Tobrman, T.; Xu, Z.; Negishi, E.-I. Org. Lett. 2009, 11,
4092.
Mn(OAc)3·2H2O (2.2 equiv)
Cu(OTf)2 (1.0 equiv), DMSO
80 °C, 36 h
H
O
O
(69%)
(13) Nicolaou, K. C.; Gray, D. L. F.; Montagnon, T.; Harrison, S. T.
Angew. Chem. Int. Ed. 2002, 41, 996.
(14) Suri, O. P.; Satti, N. K.; Sury, K. A.; Dhar, K. L.; Kachroo, P. L.;
Kawasaki, T.; Miyahara, K.; Tsunehiro, T.; Fumiko, Y. J. Nat. Prod.
1990, 53, 470.
34
35
Equation 1
We have examined two distinct approaches to the syn-
thesis of crotogoudin and crotobarin. The first pericyclic
cascade strategy was certainly undermined by the instabili-
ty of key intermediates, but computational studies suggest-
ed that poor selectivity for the desired pathway would also
have plagued this approach. A bioinspired approach led to
advanced tricyclic intermediates; unfortunately, the inabili-
ty to manipulate enone 27 prevented further forward prog-
ress toward these targets. In this study, we have shown the
strategic equivalence of dearomatizing oxidative heterocy-
clizations of phenols/dienone reduction and Birch reduc-
tion/halocyclization. Furthermore, we provide a rare exam-
ple of a manganese(III)-mediated keto-alkene cyclization
wherein the ketone is not further activated.28
(15) Lal, A.; Cambie, R. C.; Rutledge, P. S.; Woodgate, P. D. Phytochem-
istry 1990, 29, 1925.
(16) Aggarwal, V. K.; Ali, A.; Coogan, M. P. Tetrahedron 1999, 55, 293.
(17) Gansäuer, A.; Justicia, J.; Rosales, A.; Worgull, D.; Rinker, B.;
Cuerva, J. M.; Oltra, J. E. Eur. J. Org. Chem. 2006, 4115.
(18) Ushakov, D. B.; Raja, A.; Franke, R.; Sasse, F.; Maier, M. E. Synlett
2012, 23, 1358.
(19) Bauer, R. A.; Wenderski, T. A.; Tan, D. S. Nat. Chem. Bio. 2013, 9,
21.
(20) Baker, B. A.; Bošković, Z. V.; Lipshutz, B. H. Org. Lett. 2008, 10,
289.
(21) Boezio, A. A.; Pytkowicz, J.; Côté, A.; Charette, A. B. J. Am. Chem.
Soc. 2003, 125, 14260.
(22) (a) Nasipuri, D.; Chaudhuri, S. R. R. J. Chem. Soc., Perkin Trans. 1
1975, 262. (b) Zhao, J.-F.; Zhao, Y.-J.; Loh, T.-P. Chem. Commun.
2008, 1353.
(23) For a discussion of the difficulties of conjugate additions on this
type of scaffold, see: Cherney, E. C. PhD Thesis; The Scripps
Research Institute: La Jolla, CA, 2014, 95.
Funding Information
We thank the UC Cancer Research Coordinating Committee for fund-
ing initial phases of this research with grant CRC-15-380750 and the
(24) Snider, B. B. Chem. Rev. 1996, 96, 339.
(25) For one of the few examples of simple ketones functioning in
these reactions, see: O’Neil, S. V.; Quickley, C. A.; Snider, B. B.
J. Org. Chem. 1997, 62, 1970.
NSF for continued funding (CHE-1564340)
N
oaitn
a
l
Secince
F
o
u
n
d
oaitn
1(
5
6
4
3
4
0)
(26) Hosomi, A.; Sakurai, H. J. Am. Chem. Soc. 1977, 99, 1673.
(27) Sakurai, H.; Hosomi, A.; Hayashi, J. Org. Synth. 1984, 62, 86.
(28) Representative Experimental Procedure and Characteriza-
tion Data
Acknowledgment
We thank Hung Pham and Professor Ken Houk (UCLA) for assistance
with the computational evaluation of the cascade reaction in Scheme
1. Philipp Roosen is acknowledged for providing compound 34.
Compound 26: Phenol 19 (1.48 g, 5.39 mmol) was dissolved in
CF3CH2OH (50 mL) assisted by sonication. The open flask was
cooled to 0 °C. A solution of diacetoxyiodobenzene (1.82 g, 5.66
mmol) in CF3CH2OH (4 mL) was added dropwise. The reaction
mixture was stirred at 0 °C for 2 h, diluted with water (100 mL),
and extracted with CH2Cl2 (3 × 100 mL). The combined organic
extracts were dried with MgSO4, filtered, and concentrated in
vacuo. The resultant crude dark red oil was purified by column
chromatography (20% EtOAc in hexanes) to give 26 as a white
solid (0.96 g, 66% yield; mp 105–107 °C. 1H NMR (500 MHz,
CDCl3): δ = 6.86 (d, J = 10.2 Hz, 1 H), 6.33 (d, J = 10.2 Hz, 1 H),
6.16 (s, 1 H), 5.02 (s, 1 H), 4.82 (s, 1 H), 2.88–2.71 (m, 4 H), 2.42
(d, J = 13.3 Hz, 1 H), 2.02–1.83 (m, 4 H), 1.77 (s, 3 H), 0.89 (s, 3
H).13C NMR (125 MHz, CDCl3): δ = 184.8, 169.9, 156.7, 144.1,
143.6, 131.1, 126.2, 116.3, 82.8, 44.0, 41.1, 31.8, 28.6, 28.0, 26.0,
23.3, 17.5. IR (thin film): 3077, 3053, 2969, 2949, 1741, 1674,
1640, 1615 cm–1. ESI-HRMS (MeOH): m/z calcd for C17H20O3Na
[M + Na]+: 295.1310; found: 295.1311.
Compound 35: Mn(OAc)3·2H2O (0.316 g, 1.18 mmol) and
Cu(OTf)2 (0.195 g, 0.539 mmol) were added to a flame-dried
vial, degassed in triplicate (backfilling with argon), and sus-
pended in DMSO (5.4 mL). Ketone 34 (0.103 g, 0.533 mmol) was
added neat, via syringe, and the reaction mixture was heated to
80 °C in a preheated aluminum block for 36 h. The brown sus-
pension was cooled to r.t., diluted with water (2 mL), extracted
with CH2Cl2 and Et2O (2 × 2 mL). The combined organic extracts
Supporting Information
Supporting information for this article (Experimental procedures for
the synthesis of new compounds from Schemes 2, 3, and 4, and from
Equation 1, and their characterization data, are provided) is available
online at https://doi.org/10.1055/s-0036-1588560.
S
u
p
p
ortiInfogrmoaitn
S
u
p
p
ortioInfgrmoaitn
References and Notes
(1) Rakotonandrasana, O. L.; Raharinjato, F. H.; Rajaonarivelo, M.;
Dumontet, V.; Martin, M.-T.; Bignon, J.; Rasoanaivo, P. J. Nat.
Prod. 2010, 73, 1730.
(2) Gersch, M.; Kreuzer, J.; Sieber, S. A. Nat. Prod. Rep. 2012, 29, 659.
(3) Drahl, C.; Cravatt, B. F.; Sorensen, E. J. Angew. Chem. Int. Ed.
2005, 44, 5788.
(4) Breitler, S.; Carreira, E. M. Angew. Chem. Int. Ed. 2013, 52, 11168.
(5) Song, L.; Zhu, G.; Liu, Y.; Liu, B.; Qin, S. J. Am. Chem. Soc. 2015,
137, 13706.
(6) Ushakov, D. B.; Maier, M. E. Synlett 2013, 24, 705.
(7) Behera, T. K.; Singh, V. Tetrahedron 2014, 70, 7983.
(8) Guo, Y.; Liu, Q.; Jia, Y. Chem. Commun. 2015, 51, 889.
© Georg Thieme Verlag Stuttgart · New York — Synlett 2017, 28, A–E

E
D. N. Mai et al.
Letter
Synlett
were washed with NaHCO3, water, brine, dried over Na2SO4,
filtered through SiO2, and concentrated in vacuo. The resultant
crude material was purified by column chromatography (1–3%
EtOAc in hexanes) to give 35 as a yellow oil (0.070 g, 69% yield).
1H NMR (600 MHz, CDCl3): δ = 4.92 (d, J = 0.7 Hz, 1 H), 4.81 (d,
J = 1.0 Hz, 1 H), 2.85 (dd, J = 3.6, 2.2 Hz, 1 H), 2.69 (dd, J = 19.1,
3.2 Hz, 1 H), 2.23 (dd, J = 17.1, 2.3 Hz, 1 H), 2.18–2.12 (m, 2 H),
1.74 (dd, J = 19.2, 1.4 Hz, 1 H), 1.73–1.62 (m, 2 H), 1.60–1.48 (m,
3 H), 1.37–1.16 (m, 4 H), 1.08 (ddd, J = 35.7, 13.0, 3.5 Hz, 1 H).
13C NMR (125 MHz, CDCl3): δ = 212.7, 143.3, 110.6, 55.0, 44.0,
42.9, 37.5, 36.3, 36.1, 32.9, 31.2, 26.2, 21.8. HRMS (CI/CH2Cl2):
m/z calcd for
208.1703.
C13H18ONH4 [M +
NH4]+: 208.1701; found:
© Georg Thieme Verlag Stuttgart · New York — Synlett 2017, 28, A–E