A Formal Enantiospecific Synthesis of 7,20-Diisocyanoadociane

Article abstract of DOI:10.1002/anie.201603581

7,20-Diisocyanoadociane (DICA) is a potent antimalarial isocyanoterpene endowed with a fascinating tetracyclic structure composed of fused chair cyclohexanes. We report a highly stereocontrolled synthesis of a late-stage intermediate, the “Corey dione”, from which DICA has been made previously. This formal synthesis features a rapid buildup of much of the complexity of the target through a sequence of enone tandem vicinal difunctionalization, Friedel–Crafts cyclodehydration, and sequential stereocontrolled reductions. Most importantly, this success establishes the broader feasibility of our previously developed general synthesis approach to the isocyanoterpene family and provides a blueprint for a very direct synthesis of DICA and related natural products.

Full text of DOI:10.1002/anie.201603581

Angewandte
Communications
Chemie
International Edition: DOI: 10.1002/anie.201603581
German Edition: DOI: 10.1002/ange.201603581
Isocyanoterpenes
A Formal Enantiospecific Synthesis of 7,20-Diisocyanoadociane
Philipp C. Roosen and Christopher D. Vanderwal*
Abstract: 7,20-Diisocyanoadociane (DICA) is a potent anti-
malarial isocyanoterpene endowed with a fascinating tetracy-
clic structure composed of fused chair cyclohexanes. We report
a highly stereocontrolled synthesis of a late-stage intermediate,
the “Corey dione”, from which DICA has been made
previously. This formal synthesis features a rapid buildup of
much of the complexity of the target through a sequence of
enone tandem vicinal difunctionalization, Friedel–Crafts
cyclodehydration, and sequential stereocontrolled reductions.
Most importantly, this success establishes the broader feasi-
bility of our previously developed general synthesis approach
to the isocyanoterpene family and provides a blueprint for
a very direct synthesis of DICA and related natural products.
P
otent antimalarial activity among a fascinating array of
polycyclic structures, unexplained structure–activity relation-
ships, and poorly understood mechanism(s) of action combine
to elevate the importance of the isocyanoterpenes (ICTs, 1–4,
Figure 1) for detailed study.[1] A synthesis design that is
applicable to the broad range of structures in this family
would serve admirably in the development of the potential of
these natural products. In that context, we report a formal
enantiospecific synthesis of 7,20-diisocyanoadociane (DICA,
1) that builds on a general approach to many ICTs that was
first showcased in our recent synthesis of kalihinol B (2).[2]
The ICTs have been known to science since the mid-
1970s;[3] however, the first indication of antimalarial activity
in this family of natural products came from studies by the
Angerhofer and the Kçnig groups in the 1990s[4] and, since
that time, many studies on the synthesis, antimalarial activity,
and mechanism of action of the ICTs have accumulated.[1]
The focus of this report is synthesis, and in that respect,
several prior accomplishments are particularly relevant.
Corey and Magriotis reported a first synthesis of DICA in
1987 with a route of about 27 steps, with asymmetry arising
from an auxiliary-controlled enolate Michael addition; how-
ever, stereocontrol in the introduction of the two isonitrile
substituents was not accomplished in this inaugural syn-
thesis.[5] Rather, tetracyclic dione 5 (“Corey dione”) was
subjected to double nucleophilic methylation and activation
of the corresponding tertiary carbinols as trifluoroacetate
esters, the displacement of which under the conditions shown
was, unsurprisingly, not selective. Nearly two decades later,
Fairweather and Mander reported a longer synthesis that
Figure 1. Potent antimalarial isocyanoterpenes and the Corey dione,
a late-stage intermediate in the Corey–Magriotis synthesis of DICA. W2
and Dd2 are drug-resistant strains of Plasmodium falciparum,
a malaria-causing parasite.
featured completely stereocontrolled introduction of the
tertiary carbinolamine precursors to the C7 and C20 isoni-
triles by using Curtius degradation of carboxylic acid pre-
cursors.[6] In 2011, Miyaoka and co-workers disclosed a formal
synthesis of DICA, reaching the Corey dione through
a sequence that is strategically aligned with the Corey and
Magriotis accomplishment but somewhat lengthier.[7] Piers
and co-workers completed syntheses of amphilectane and
cycloamphilectane ICTs with isonitriles located at ring
junctions,[8] and Miyaoka and Okubo made an amphilectane
natural product related to 3.[9] Finally, the stunning 2014
synthesis of (Æ)-amphilectene 3 by Pronin and Shenvi in
approximately 10 steps raised the bar for synthesis in this
area;[10] moreover, they introduced an important tool for
stereocontrolled introduction of the tertiary isonitriles,
namely a largely invertive displacement of tertiary trifluor-
oacetates.[11]
[*] P. C. Roosen, Prof. C. D. Vanderwal
Department of Chemistry, 1102 Natural Sciences II
University of California, Irvine Irvine, CA 92697-2025 (USA)
E-mail: cdv@uci.edu
A key element of our recent synthesis of kalihinol B (2)
was the stereocontrolled tandem vicinal difunctionalization of
cyclohexenone intermediate 6 (Figure 2), which we accessed
in a direct way. The challenge of the C7-C6-C1 stereotriad was
thus succinctly met through substrate-controlled selectivity.
In a retrospective analysis, we recognized the potential power
Supporting information and the ORCID identification number(s) for
the author(s) of this article can be found under http://dx.doi.org/10.
1002/anie.201603581.
7180
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 7180 –7183

Angewandte
Communications
Chemie
Figure 2. A potentially general synthesis plan for ICTs.
of this strategy to address a conserved structural motif among
many of the ICTs. Herein, we describe our first successful
foray along these lines, which led to a formal synthesis of
DICA by intersection with the Corey dione (5, Figure 1). This
accomplishment serves as a proof-of-principle for the gen-
erality of our synthesis, and it is noteworthy for the high levels
of stereocontrol in accessing the all-trans-perhydropyrene of
5, for the enabling use of carbonyl-based reactivity, and for
the ability to telescope multiple reaction sequences. Ulti-
mately, a highly stereoselective synthesis of the Corey dione
was achieved via only ten purified intermediates from
commercially available materials.
Scheme 1. Synthesis of tetracyclic enone 10 through a Birch reduction
strategy. Reagents and conditions: a) ArMgBr, cat. CuI·2LiCl, THF
(À788C!RT) then HMPA, BrCH2CO2Et; b) LiAlH4, Et2O (08C!RT),
56% yield over 2 steps; c) TsCl, pyridine (408C); d) cat. OsO4, NMO,
aq. acetone; NaIO4, aq. THF; e) cat. TsOH, Dean–Stark, toluene
(reflux), 80% over 3 steps; e) Li, MeOH, NH3, THF (À408C); HCl, aq.
MeOH, 65%. The diastereoselectivity for each reaction is shown in the
scheme. THF=tetrahydrofuran, HMPA=hexamethylphosphoramide,
Ts =para-toluenesulfonyl.
Our synthesis of the Corey dione begins with cyclo-
hexenone (Æ)-7 and is shown in Schemes 1 and 2, in which
only the structures of chromatographically purified inter-
mediates are provided; all other intermediates were used in
crude form. We accessed racemic enone 7 on multigram scale
from known (Æ)-3-methyl-4-pentenal[12] and methyl vinyl
ketone, surmising that application of a chiral organocatalyst
to the Michael addition step,[13] as in our kalihinol B synthesis,
would render the route highly enantioselective (see below).
Key to this strategy is the use of the single C4 asymmetric
center (DICA numbering) in 7 to control all others in the
target molecule; the center at C3 will be removed and
stereoselectively reinstalled at a later stage.
The synthesis began with the vicinal difunctionalization of
(Æ)-7 (Scheme 1).[14] Conjugate arylation/enolate alkylation
was efficient and highly diastereoselective, setting the three
required stereocenters of C4, C13 and C8 as all-trans. Facile
epimerization a to the ketone necessitated a quick reduction
to diol 8 (d.r. 7:1 at the new carbinol center at C7, which is not
relevant to the target). Our observations of the propensity of
the diol to cyclize led us to purposefully generate the
tetrahydrofuran ring as a means of internally protecting
both groups; two-stage oxidative alkene cleavage and acid-
mediated Friedel–Crafts-type cyclodehydration afforded
dihydronaphthalene 9, thereby erasing the configuration at
C3. Birch reduction of the styrenyl system followed by acidic
hydrolysis/conjugation efficiently provided enone 10 with
exquisite stereocontrol at both C1 and C3, as confirmed by X-
ray crystallographic analysis. Notably, this efficient and
readily executed sequence provides reliable access to grams
of tetracyclic enone 10. Cyclohexenone 10 can also be
prepared as a single diastereomer through purification at
separate stages;[15] however, the irrelevance of the C7
stereocenter to the target did not justify the extra effort.
Cyclohexenone 10 has features reminiscent of the A/B
ring system of many steroidal systems which, when treated
under dissolving-metal reduction conditions, typically afford
excellent selectivity for the trans ring junction.[16] Likely
owing to the syn-pentane interaction present in key reduction
intermediates (Figure 3), selecting for the trans ring fusion
proved challenging in this case (Table 1). When enone 10 was
treated with Li/NH3, the major product was cis- protonated 12
with 2:1 selectivity (entry 1). A screen of alkali-metal
reductants and the use of tBuOH as a bulky proton source
had little impact on the selectivity (entries 2–4). Homoge-
neous hydride reagents such as Karstedtꢀs catalyst[17] and
tBuCu/DIBAl[18] only enhanced the cis-selectivity (entries 5–
6). Fortunately, hetereogenous reducing agents provided
mixtures significantly enriched in the trans product
(entries 7–9). A small screen of hydrogenation catalysts led
Figure 3. Comparison of protonation events to explain the preference
for cis-reduced 12 over trans-reduced 11 with dissolving metals.
Angew. Chem. Int. Ed. 2016, 55, 7180 –7183
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7181

Angewandte
Communications
Chemie
Table 1: Selectivity in the reduction of enone 10.
tated by the failure of a planned Lewis acid mediated
tetrahydrofuran ring opening that had worked on simpler
model systems. This oxidation needed to be heavily opti-
mized[15] before we arrived at the conditions shown (Fieserꢀs
reagent in MeNO2), which provided 13 in 38% yield over the
two steps.[20] An overall efficient five-step sequence was
carried out to open the lactone to differentiated diol 14, the
primary alcohol of which was oxidized to permit aldol
condensation to close the final ring in 15. The cross-
conjugated dienolate derived from enone 15 was methylated
at C15 with a moderate axial preference; after enone
hydrogenation/benzyl ether hydrogenolysis and oxidation of
the C7 alcohol, base-mediated equilibration afforded the
Corey dione (5). Access to this target, which completes
a formal synthesis of DICA, was achieved in a total of 18 steps
from 7,[21] but with only eight chromatographic purifications
and with excellent relative stereochemical control at all eight
centers of the perhydropyrene scaffold.
Having successfully procured the Corey dione in racemic
form, we aimed to render the synthesis asymmetric. The
choice of cyclohexenone 7 as a starting material was inten-
tional to take advantage of our groupꢀs experience with the
organocatalytic asymmetric Robinson annulation, which had
served well in our kalihinol B synthesis.[2,13] Unfortunately, we
could not uncover conditions for efficient ring-closing aldol
condensations without significant erosion of enantiopurity,
and an alternative entry to the asymmetric manifold was
sought (Scheme 3).
Entry
Reduction Conditions
Yield [%][a]
11/12
1
2
3
4
5
6
7
8
9
Li, NH3, THF (À408C)
Na, NH3, THF (À788C)
K, NH3, THF (À788C)
64
85
82
80
92
86
94
93
93
1:2
1:1
1:1
1:3
1:5
<1:20
6:1
8:1
15:1
K, t-BuOH, NH3, THF (À788C)
Karstedt, Et3SiH (708C)[b]
t-BuCu, DIBAl, HMPA, THF (À508C)
H2, Pd/C, EtOAc
H2, Rh/alumina, EtOAc[c]
H2, Rh/C, EtOAc[c]
[a] Yields of isolated product after column chromatography. [b] Followed
by TBAF, THF. [c] i. 400–500 psi H2 ii. PCC, Celite, CH2Cl2 (to reoxidize
any undesired alcohol formed). Karstedt=platinum(0)-1,3-divinyl-
1,1,3,3-tetramethyldisiloxane complex, DIBAl=diisobutylaluminum hy-
dride, TBAF=tetrabutylammonium fluoride, PCC=pyridinium chloro-
chromate.
to the discovery that reductions with Rh/C were highly trans-
selective (d.r. 15:1).
The crude trans-fused hydrogenation product, which
contained some over-reduced C20 alcohol, was oxidized to
the lactone/ketone 13 (Scheme 2).[19] The moderate yield of
the oxidation accounts for virtually all of the losses in the two-
stage process. This ether-to-lactone oxidation was necessi-
Scheme 2. Synthesis of the Corey dione (5). Reagents and conditions:
g) 500 psi H2, cat. Rh/C, EtOAc; h) CrO3, aq. AcOH, MeNO2, 38%
over 2 steps; i) cat. TsOH, ethylene glycol, Dean–Stark, benzene
(reflux); j) LiAlH4, THF (08C!RT); k) TrCl, NEt3, cat. DMAP, DCE
(708C); l) KH, BnCl, cat. TBAI, THF (508C); m) cat. TsOH, cat. PPTS,
aq. acetone (708C), 75% over 5 steps; n) DMP, NaHCO3, CH2Cl2;
o) cat. TsOH, Hickman still, benzene (reflux), 72% over 2 steps;
p) LDA, MeI, HMPA, THF (À788C!RT), 85%; q) 900 psi H2, cat. Pd/
C, EtOAc; r) PCC, Celite, CH2Cl2; NaOH, MeOH (508C), 65% yield
over 2 steps. Tr =triphenylmethyl, DMAP=4-dimethylaminopyridine,
DCE=1,2-dichloroethane, Bn=benzyl, TBAI=tetrabutylammonium
iodide, PPTS=pyridinium para-toluenesulfonate, DMP=Dess–Martin
periodinane, LDA=lithium diisopropylamide.
Scheme 3. Chiral-pool synthesis of dihydronaphthalene (À)-9.
Reagents and conditions: a) TBSOTf, NEt3, CH2Cl2 (08C); b) m-CPBA,
aq. NaHCO3, Et2O; c) NaIO4, aq. HF, MeCN, 62% over 3 steps;
d) ArMgBr, cat. CuI·2LiCl, THF (À788C!RT) then HMPA,
BrCH2CO2Et; e) LiAlH4, Et2O (08C!RT), 37% yield over 2 steps;
f) TsCl, pyridine (408C), 83%; g) mCPBA, NaHCO3, CH2Cl2; h) cat.
TsOH, Hickman still, benzene (reflux), 60% yield over 2 steps.
TBSOTf=tert-butyldimethylsilyl triflate, mCPBA=meta-chloroperoxy-
benzoic acid.
7182
www.angewandte.org
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 7180 –7183

Angewandte
Communications
Chemie
Inspiration for a new starting point came from analysis of
How to cite: Angew. Chem. Int. Ed. 2016, 55, 7180–7183
Angew. Chem. 2016, 128, 7296–7299
the synthesis in hand. While the preparation of dihydronaph-
thalene 9 requires oxidative cleavage of an alkene precursor
to generate an aldehyde for Friedel–Crafts cyclodehydration,
the intermediate aldehyde could instead arise from a deho-
mologated alkene via net oxidation. This insight identified the
known cyclohexenone (À)-19, which is available in enantio-
pure form,[22] as an attractive chiral-pool starting material.
Conversion of the inexpensive terpene (À)-perillaldehyde
into cyclohexenone (À)-19 on a gram scale has been
previously reported;[22b] however, to improve throughput,
we found that purification by distillation was preferable
(Scheme 3). In an unoptimized sequence, conjugate addition
and alkylation selectively installed the three stereogenic
centers with desired diastereoselectivity as seen in (+)-20, in
a similar manner as we previously showed to make 8.
Condensation of diol (+)-20 via the tosylate furnished trans-
fused tetrahydrofuran (+)-21. Alkene epoxidation is followed
by heating at reflux with acid, which presumably triggered
epoxide rearrangement to the aldehyde followed by in situ
cyclodehydration to dihydronaphthalene (À)-9. This
sequence intersects our racemic synthesis of DICA
(Schemes 1 and 2), thereby permitting access to all later
intermediates in enantiopure form. Because of our desire to
improve several aspects of the overall synthesis, we elected
not to revisit our formal DICA synthesis using this optically
active material, though it is clear that this chiral-pool
approach is suitable for doing so with respect both to control
of absolute configuration and material throughput. In its
current form, this route ends up at 21 steps from peril-
laldehyde with only 10 purifications.
Our formal synthesis of DICA that intersects with the
Corey dione features the highly diastereoselective introduc-
tion of the eight stereogenic centers of this perhydropyrene
intermediate. A shift in strategy away from poorly enantio-
selective Robinson annulation toward the adoption of
a chiral-pool starting material assures the production of
enantiopure material through an enantiospecific route.[23] As
we move toward an improved complete synthesis of DICA,
the challenges that remain include differentiating the C7 and
C20 carbonyl groups so that the salient isonitriles can be
installed with stereochemical control, and further stream-
lining the route by obviating some of the less attractive
functional-group interconversions and protecting-group
manipulations. These issues notwithstanding, with this work
we have clearly demonstrated that our strategy beginning
with chiral cyclohexenones is applicable not just to the
kalihinanes, but also more broadly within the ICT family.
[1] An excellent review: M. J. Schnermann, R. A. Schenvi, Nat.
Prod. Rep. 2015, 32, 543 – 577.
[2] a) M. E. Daub, J. Prudhomme, K. Le Roch, C. D. Vanderwal, J.
Am. Chem. Soc. 2015, 137, 4912 – 4915; for an earlier approach
to DICA from our group, see: b) P. C. Roosen, C. D. Vanderwal,
Org. Lett. 2014, 16, 4368 – 4371.
[3] a) F. Cafieri, E. Fattorusso, S. Magno, S. Santacroce, D. Sica,
Tetrahedron 1973, 29, 4259 – 4262; b) B. J. Burreson, C. Christo-
pherson, P. J. Scheuer, J. Am. Chem. Soc. 1975, 97, 201 – 202;
c) B. J. Burreson, C. Christopherson, P. J. Scheuer, Tetrahedron
1975, 31, 2015 – 2018; d) B. J. Burreson, P. J. Scheuer, J. Finer, J.
Clardy, J. Am. Chem. Soc. 1975, 97, 4763 – 4764; e) J. T. Baker,
R. J. Wells, W. E. Oberhänsli, G. B. Hawes, J. Am. Chem. Soc.
1976, 98, 4010 – 4012.
[4] a) C. K. Angerhofer, J. M. Pezzuto, G. M. Kçnig, A. D. Wright,
O. Sticher, J. Nat. Prod. 1992, 55, 1787 – 1789; b) G. M. Kçnig,
A. D. Wright, C. K. Angerhofer, J. Org. Chem. 1996, 61, 3259 –
3267; For a review, see: c) A. D. Wright, G. M. Kçnig, C. K.
Angerhofer, P. Greenidge, A. Linden, R. Desqueyroux-Fafflndez,
J. Nat. Prod. 1996, 59, 710 – 716.
[5] E. J. Corey, P. A. Magriotis, J. Am. Chem. Soc. 1987, 109, 287 –
288.
[6] K. A. Fairweather, L. N. Mander, Org. Lett. 2006, 8, 3395 – 3398.
[7] H. Miyaoka, Y. Okubo, M. Muroi, H. Mitome, E. Kawashima,
Chem. Lett. 2011, 40, 246 – 247.
[8] a) E. Piers, M. Llinas-Brunet, J. Org. Chem. 1989, 54, 1483 –
1484; b) E. Piers, M. A. Romero, Tetrahedron 1993, 49, 5791 –
5800; c) E. Piers, M. Llinas-Brunet, R. M. Oballa, Can. J. Chem.
1993, 71, 1484 – 1494; for an otherwise unpublished synthesis of
8-isocyano-10-cycloamphilectene from the Piers laboratory, see:
d) T. W. Schindeler, Ph.D. thesis, University of British Columbia,
1998.
[9] H. Miyaoka, Y. Okubo, Synlett 2011, 547 – 550.
[10] S. V. Pronin, R. A. Shenvi, J. Am. Chem. Soc. 2012, 134, 19604 –
19606.
[11] S. V. Pronin, C. A. Reiher, R. A. Shenvi, Nature 2013, 501, 195 –
199.
[12] 3-Methyl-4-pentenal has been prepared many times, and we used
elements of several known preparations. See the the Supporting
Information for details.
[13] Y. Chi, S. H. Gellman, Org. Lett. 2005, 7, 4253 – 4256.
[14] M. T. Reetz, A. Kindler, J. Organomet. Chem. 1995, 502, C5 – C7.
[15] Please see the Supporting Information for more details.
[16] D. Caine, Org. React. 1976, 23, 1 – 258.
[17] C. R. Johnson, R. K. Raheja, J. Org. Chem. 1994, 59, 2287 – 2288.
[18] A. R. Daniewski, J. Kiegiel, Synth. Commun. 1988, 18, 115 – 118.
À
[19] A recent review of C H oxidations: T. Newhouse, P. S. Baran,
Angew. Chem. Int. Ed. 2011, 50, 3362 – 3374; Angew. Chem.
2011, 123, 3422 – 3435.
[20] The successful oxidation was patterned after a similar reaction:
A. Wettstein, K. Mischer, Helv. Chim. Acta 1942, 25, 718 – 731.
[21] Cyclohexenone 7 was made in three steps with two purifications,
each by distillation. See the the Supporting Infroamtion for
details.
[22] a) R. V. Stevens, K. F. Albizati, J. Org. Chem. 1985, 50, 632 – 640;
b) C. Siegel, P. M. Gordon, D. B. Uliss, G. R. Handrick, H. C.
Dalzell, R. K. Razdan, J. Org. Chem. 1991, 56, 6865 – 6872.
[23] A excellent review, which includes perillaldehyde as a starting
material: T. Money, M. K. C. Wong, Stud. Nat. Prod. Chem.
1995, 16, 123 – 288.
Acknowledgements
P. C.R. was supported in part by a UC Irvine Regentsꢀ
Dissertation Fellowship. We thank UC Irvine for support of
this research, and Dr. Joseph Ziller and Jason Jones for
assistance with X-ray crystallography.
Keywords: chiral pool · stereocontrol · isocyanoterpenes ·
Received: April 12, 2016
natural products · total synthesis
Published online: May 10, 2016
Angew. Chem. Int. Ed. 2016, 55, 7180 –7183
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7183