Progress toward the Enantioselective Synthesis of Curcusones A-D via a Divinylcyclopropane Rearrangement Strategy

Article abstract of DOI:10.1021/acs.orglett.9b03829

We report our iterative efforts toward the divergent total syntheses of curcusones A-D via Suzuki coupling, intramolecular cyclopropanation, and a key divinylcyclopropane rearrangement. Progress of our synthesis was repeatedly challenged by the highly sub

Full text of DOI:10.1021/acs.orglett.9b03829

Letter
pubs.acs.org/OrgLett
Cite This: Org. Lett. XXXX, XXX, XXX−XXX
Progress toward the Enantioselective Synthesis of Curcusones A−D
via a Divinylcyclopropane Rearrangement Strategy
Austin C. Wright,‡ Chung Whan Lee,‡ and Brian M. Stoltz*
The Warren and Katharine Schlinger Laboratory for Chemistry and Chemical Engineering, Division of Chemistry and Chemical
Engineering, California Institute of Technology, MC 101-20, Pasadena, California 91125, United States
S
* Supporting Information
ABSTRACT: We report our iterative efforts toward the divergent total syntheses of curcusones A−D via Suzuki coupling,
intramolecular cyclopropanation, and a key divinylcyclopropane rearrangement. Progress of our synthesis was repeatedly
challenged by the highly substrate-dependent cyclopropanation step, which we could ultimately overcome by judicious choice of
substituents on the six-membered ring fragment.
After attempting several strategically related routes, our final
retrosynthesis proposed that a late-stage α-functionalization of
enone 7 would permit a divergent approach to the
enantiomeric series of curcusones A−D (Scheme 1). As
he curcusone family comprises various synthetically
Tchallenging and biologically active rhamnofolane diterpe-
noid natural products, several of which have known
bioactivity.1 Among them, curcusone C (3, Figure 1)
demonstrates the most potent and varied anticancer properties,
including antiproliferative activity against human hepatoma,
ovarian carcinoma, and promyeolycytic leukemia.2 Despite
these enticing biological features, 3 and all of its structural
relatives have yet to surrender to any total synthesis campaigns.
It is worthwhile to note that the Dai lab recently reported a
synthesis of racemic oxo-bridged 5 and 6 over 21 steps, but
found that the putative natural product structures had been
incorrectly assigned by NMR.3,4 We do not anticipate these
issues with 3, as its structure has been unambiguously
determined by X-ray crystallography.2 Herein, we report our
early synthetic forays into the enantioselective construction of
curcusones A−D via Suzuki coupling and sequential
divinylcyclopropane rearrangement.
Scheme 1. Retrosynthetic Analysis of ent-Curcusone C (ent-
3)
such, we envisioned performing an α-functionalization,
oxidation, and olefination of silyl ether 7 (highlighted in
red). The ene-dione moiety of 7 may be derived from
ketoalcohol 8 by means of alcohol oxidation and acid- or base-
promoted olefin migration. The central seven-membered ring
present in 8 could be assembled by a stereospecific
divinylcyclopropane rearrangement and subsequent oxidative
Received: October 27, 2019
Figure 1. Proposed structures of curcusones A−J.
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.9b03829
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
cleavage of hydroxymethylated cyclopropane 9. We predicted
that the cyclopropane moiety in 9 might be installed via
intramolecular π-bond-cyclopropanation of a metallocarbenoid
derived from diazo ketoester 10 followed by methylenation
and reductive lactone opening. The diazo oxobutanoyl
functionality in 10 (highlighted in red) could be incorporated
by acylation and diazo transfer of alcohol 11. Bicycle 11 may
be accessed by a Suzuki cross-coupling of monocyclic
fragments (+)-12 and 13.
In order to validate our synthetic approach, we focused on
developing a suitable model system with which we could
investigate the crucial divinylcyclopropane rearrangement. To
this end, we were able to join coupling partners rac-12 and 165
via Suzuki coupling to expediently provide 17 (Scheme 2).
Upon assembling the model bicycle, we turned our attention
toward constructing the cyclopropane functionality. Thus,
esterification of 17 with diketene (18) afforded ketoester 19,
which was subjected to α-diazo transfer to furnish cyclo-
propanation precursor 20 in excellent yield.6
we instead opted to forge the second vinyl group from
common intermediate 21 via carbonyl methylenation (Scheme
3). Thus, olefination of the ketone functionality in 21 using
Wilkinson’s catalyst and TMSCHN2 afforded rearrangement
precursor 22.8 Gratifyingly, reductive rupture of the butyr-
olactone ring in 22 resulted in spontaneous rearrangement to
provide tricyclic alcohol 24, presumably via intermediate bis-
alkoxide 23. Due to the known conformational restrictions
imposed by the cyclopropane moiety, this rearrangement likely
proceeds through an endo-boat9 transition state to stereo-
specifically afford the rearranged product following proto-
demetalation of the alkoxide.
Scheme 3. Validation of Model Studies
Scheme 2. Model Studies for the Crucial Rearrangement
Upon corroborating our synthetic approach via model
studies, we attempted to deploy this technology on the actual
system. We initially intended to install the exo-methylene in
the eastern ring of 1−4 prior to cross-coupling. To this end,
limonene oxide (25, Scheme 4) was exposed to eliminative
epoxide opening to provide an allylic alcohol, which could be
subjected to DMP oxidation10 to afford enone 26 as well as an
undesired hetero-Diels−Alder adduct.11 Following optimiza-
tion, we eventually found that 26 could be elaborated to enol
triflate 27 in satisfactory yields using KHMDS and Comins
reagent.12 Meanwhile, boronate coupling partner (−)-12 was
assembled from known vinyl bromide 2813 via CBS reduction,
alcohol protection, and O-silylation. Pleasingly, Suzuki
coupling of fragments (−)-12 and 27 delivered requisite
bicyclic diene 30, which upon silyl deprotection, acetoacety-
lation, and diazo transfer delivered annulation precursor 32 via
β-ketoester 31.
With model precursor 20 in hand, we attempted the key
cyclopropanation step. Both rhodium and copper catalysts
were employed to effect this transformation (Figure 2).
Although dirhodium catalysts (entries 1−4) either offered no
reactivity or resulted in decomposition, Cu(TBSal)2 (entry 5)
gratifyingly furnished desired annulated product 21 in
moderate yield and with only minor levels of decomposition.
Scheme 4. 1st Generation Synthesis of Diazo 32
Figure 2. Optimization screen for the cyclopropanation step.
Having successfully forged model cyclopropane 21, we next
aimed to prepare and rearrange the envisioned divinylcyclo-
propane system. At the outset of our synthetic explorations, we
hoped to derive the second necessary vinyl group from the
ketone moiety in 21 via silyl enol ether formation. During
these studies, we encountered a mechanistically unusual
cascade reaction consisting of five successive pericyclic
rearrangements.7 In order to obviate this unexpected pathway,
B
DOI: 10.1021/acs.orglett.9b03829
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
With diazo 32 in hand, we next investigated the critical
cyclopropanation step (Figure 3). Unfortunately, all attempts
to advance 32 to 33 resulted in decomposition (entries 1−3), a
complex mixture of byproducts (entry 4), or simply no
reactivity (entry 5). These combined issues necessitated a
revision of our synthetic route. Given the amenability of our
model substrate (20, Figure 2, vide supra) toward similar
cyclopropanation conditions, we speculated that our inability
to execute this transformation on diazo 32 might be due to
subtle stereoelectronic influences14 imparted by the presence
of the proximal exo-methylene group (Figure 3, highlighted in
red). In light of this, we instead chose to install a ketone
functionality, which we suspected could be elaborated to the
corresponding exo-methylene upon olefination.
Figure 4. Synthetic evolution of cyclohexene coupling fragments.
ether 13, Scheme 6). This route commenced with bromination
of 35 to afford bromoenone 41, which could be nonselectively
reduced in a 1,2-fashion and O-silylated to produce desired cis
epimer 13 along with the undesired trans epimer in roughly
equal portions. It is worthwhile to note that the thermody-
namically17 and kinetically18 favored trans epimer was also
investigated as a potential synthetic intermediate but was
eventually found to be totally uncooperative toward cyclo-
propanation conditions due to exclusive decomposition.
Moving forward, the redundancy of silyl protecting groups in
both coupling partners motivated us to attempt deprotecting
the boronate fragment prior to cross-coupling. Unfortunately,
after an exhaustive screening we found no silyl deprotection
conditions that could accommodate the presence of the base-
and acid-sensitive boronate functionality. Other protecting
groups were also investigated with minimal success.19
Interestingly, over the course of our efforts we eventually
discovered that a cyclopentenol protection strategy could be
avoided altogether by performing an unconventional double
lithiation/boronate trapping procedure on (+)-29 with
pinacolborane to provide allyl alcohol 42. With revised
boronate 42 and bromide 13 in our possession, we next
explored the crucial Suzuki coupling. Gratifyingly, standard
coupling conditions afforded bicyclic alcohol 11 in good yield.
Figure 3. Unsuccessful cyclopropanation of 32.
Now targeting the enantiomeric series of curcusones A−D
via coupling of enone fragment 36, we assembled cyclo-
hexenone 35 over three steps from (S)-perillaldehyde (34,
Scheme 5) according to known methods.15 An ensuing α-
iodination of 35 using I2 and pyridine provided desired
iodoenone 36, which itself could undergo the anticipated
Suzuki coupling with boronate (+)-1216 to furnish correspond-
ing bicycle 37 in good yield. Silyl ether 37 could be advanced
to β-ketoester 38 without event following deprotection and
transacylation. Although the vital base-mediated diazo transfer
and cyclopropanation sequence gratifyingly furnished desired
enone 39, significant amounts of unwanted olefin isomer 40
were also observed. Unfortunately, all efforts to perform a
subsequent double olefination on the two ketones of 39 failed,
prompting us to again reassess our route.
Scheme 6. Third Generation Assembly of Revised Bicycle
11
Considering the undesired isomerization pathways facilitated
by the presence of the gamma proton in enone 36 (Figure 4,
highlighted in red), we finally decided to instead target a 1,2-
reduced and protected analogue of enone 36 (i.e., silyl allyl
Scheme 5. Second Generation Route toward ent-1−4
As expected, bicycle 11 could be readily advanced to
cyclopropanation precursor 10 following esterification and
diazo transfer (Scheme 7). We were further pleased to find that
exposure of diazo 10 to previously optimized cyclopropanation
conditions did indeed afford annulated product 44, albeit only
on small scales (i.e., 20 mg or less) and with substantial
portions of an unwanted byproduct.20,21 With provision of
cyclopropane 44, we next hoped to append the second vinyl
tether. After extensive studies, we discovered that methylena-
tion of the ketone in 44 to olefin 45 could be achieved under
Kauffmann olefination conditions,22 thereby finally establishing
the essential divinylcyclopropane system.
C
DOI: 10.1021/acs.orglett.9b03829
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
possessing the 5/7/6 carbocyclic framework of curcusones A−
J. Several end-game strategies to complete the first total
syntheses of curcusone A−D are currently under careful
scrutiny in our group.
Scheme 7. Third Generation Synthesis of Cyclopropane 45
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acs.orglett.9b03829.
NMR and IR spectra for all unknown compounds
(PDF)
Upon accessing divinylated intermediate 45, we set out to
induce the pivotal rearrangement. To this end, reductive
opening of the butyrolactone moiety of 45 provided diol 9
along with minor amounts of desired rearrangement product
46 (Scheme 8). Fortuitously, we found that this crude product
mixture smoothly underwent the envisioned divinylcyclopro-
pane rearrangement upon gentle heating to provide tricycle 46,
possessing the carbocyclic skeleton embedded in each of the
curcusones.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: stoltz@caltech.edu.
ORCID
Brian M. Stoltz: 0000-0001-9837-1528
Author Contributions
‡A.C.W. and C.W.L. contributed equally.
Notes
Scheme 8. Construction of Tricycle 46 by Lactone Opening
and Thermal Rearrangement
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
We thank NSF (CHE-1800511), Amgen, and Caltech for
funding this research.
■
REFERENCES
■
(1) (a) Chianese, G.; Fattorusso, E.; Aiyelaagbe, O. O.; Luciano, P.;
̈
̈
Schroder, H. C.; Muller, W. E. G.; Taglialatela-Scafati, O.
Spirocurcasone, a Diterpenoid with a Novel Carbon Skeleton from
Jatropha curcas. Org. Lett. 2011, 13, 316−319. (b) Liu, J.-Q.; Yang, Y.-
F.; Li, X.-Y.; Liu, E.-Q.; Li, Z.-R.; Zhou, L.; Li, Y.; Qiu, M.-H.
Cytotoxicity of naturally occurring rhamnofolane diterpenes from
Jatropha curcus. Phytochemistry 2013, 96, 265−272. (c) Picha, P.;
Naengchomnong, W.; Promratanapongse, P.; Kano, E.; Hayashi, S.;
Ohtsubo, T.; Zhang, S. W.; Shioura, H.; Kitai, R.; Matsumoto, H.;
Kawahara, K.; Puribhat, S.; Phanthumachinda, P. Effect of natural
pure compounds curcusones A and C from tropical herbal plant
Jatropha curcas on thermosensitivity and development of thermotol-
erance in Chinese hamster V-79 cells in vitro. J. Exp. Clin. Cancer Res.
1996, 15, 177−183.
With the 5−7−6 carbon skeleton finally in hand, we were
eager to elaborate diol 46 to ketoalcohol 48 via chemoselective
oxidation of the primary alcohol to deliver acid 47 (Scheme 9).
We suspected that intermediate 47 may itself undergo the
crucial oxidative cleavage upon acid chloride formation and
subsequent carboxy-inversion.23 We were however disap-
pointed to find that all attempts to advance 46 to 47 have
thus far led to rapid decomposition, likely due to the
exceptional instability of 46 toward oxidative conditions.24
Eventually, the combined difficulties posed by this substrate
instability and by the prohibitively scale-dependent cyclo-
propanation step motivated us to embark on an entirely new
route, which was since found to hinge on an RCM approach to
assemble the central seven-membered ring.25
(2) Naengchomnong, W.; Thebtaranonth, Y.; Wiriyachitra, P.;
Okamoto, K. T.; Clardy, J. Isolation and structure determination of
four novel diterpenes from jatropha curcus. Tetrahedron Lett. 1986,
27, 2439−2442.
(3) Li, Y.; Dai, M. Total Syntheses of the Reported Structures of
Curcusones I and J Through Tandem Gold Catalysis. Angew. Chem.,
Int. Ed. 2017, 56, 11624−11627.
Scheme 9. Envisioned Oxidative Cleavage Sequence
(4) Revised structures of curcusones I and J were recently proposed,
see: Sarotti, A. M. Structural revision of two unusual rhamnofolane
diterpenes, curcusones I and J, by means of DFT calculations of NMR
shifts and coupling constants. Org. Biomol. Chem. 2018, 16, 944−950.
(5) Stang, P. J.; Treptow, W. Single-Step Improved Synthesis of
Primary and Other Vinyl Trifluoromethanesulfonates. Synthesis 1980,
1980, 283−284.
(6) Davies, H. M. L.; Cantrell, W. R., Jr.; Romines, K. R.; Baum, J. S.
Synthesis of Furans via Rhodium(II) Acetate-Catalyzed Reaction of
Acetylenes with α-Diazocarbonyls: Ethyl 2-Methyl-5-Phenyl-3-Fur-
ancarboxylate. Org. Synth. 1992, 70, 93−97.
(7) Lee, C. W.; Taylor, B. L. H.; Petrova, G. P.; Patel, A.;
Morokuma, K.; Houk, K. N.; Stoltz, B. M. An Unexpected Ireland−
Claisen Rearrangement Cascade During the Synthesis of the Tricyclic
Core of Curcusone C: Mechanistic Elucidation by Trial-and-Error
In summary, we have presented our introductory synthetic
efforts toward the construction of curcusones A−D via a
divinylcyclopropane rearrangement disconnection. During our
studies, we invoked several unusual reaction procedures,
including an uncommon Kauffmann methylenation and a
novel one-step borylation of an unprotected α-bromo allyl
alcohol. We also notably accessed a late-stage intermediate
D
DOI: 10.1021/acs.orglett.9b03829
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
and Automatic Artificial Force-Induced Reaction (AFIR) Computa-
tions. J. Am. Chem. Soc. 2019, 141, 6995−7004.
synthesis of the ineleganolide core by a tandem annulation cascade.
Chem. Sci. 2017, 8, 507−514.
(8) Lebel, H.; Paquet, V. Rhodium-Catalyzed Methylenation of
Aldehydes. J. Am. Chem. Soc. 2004, 126, 320−328.
(22) (a) Oesterreich, K.; Klein, I.; Spitzner, D. “One-pot” Reactions:
Total Synthesis of the Spirocyclic Marine Sesquiterpene, (+)-Axenol.
Synlett 2002, 1712−1714. (b) Kauffmann, T.; Fiegenbaum, P.;
(9) Kruger, S.; Gaich, T. Recent Applications of the Divinylcyclo-
̈
propane−Cycloheptadiene Rearrangement in Organic Synthesis.
Beilstein J. Org. Chem. 2014, 10, 163−193.
̈
Papenberg, M.; Wieschollek, R.; Sander, J. Organomolybdan- und
Organowolfram-Reagenzien, II. uber den carbonylolefinierenden μ-
̈
̈
Methylenkomplex aus Mo2Cl10 und vier aquivalenten Methyllithium.
Chem. Ber. 1992, 125, 143−148.
(10) Wang, Q.; Fan, S. Y.; Wong, H. N. C.; Li, Z.; Fung, B. M.;
Twieg, R. J.; Nguyen, H. T. Enantioselective synthesis of chiral liquid
crystalline compounds from monoterpenes. Tetrahedron 1993, 49,
619−638.
(23) Denney, D. B.; Sherman, N. Degradation of Acids to Alcohols
by the Carboxy-Inversion Reaction. J. Org. Chem. 1965, 30, 3760−
3761.
(11) The following hetero-Diels−Alder adduct (highlighted in red)
was observed as the major diastereomeric byproduct:
(24) The following oxidizing conditions led to the exclusive
decomposition of 46:
(12) Comins, D. L.; Dehghani, A.; Foti, C. J.; Joseph, S. P. Pyridine-
Derived Triflating Reagents: N-(2-Pyridyl)-triflimide and N-(5-
Chloro-2-pyridyl)triflimide. Org. Synth. 1997, 74, 77−80.
(13) Karaarslan, M.; Gokturk, E.; Demircan, A. Thermal Intra-
molecular Diels−Alder Reaction of Furan; Synthesis of Nitrogen
Tetracycles, Isobenzofuran and Isobenzothiophene. J. Chem. Res.
2007, 2007, 117−120.
(14) The exo-methylene group may pose moderate steric constraints
on the system, preventing facile metallocarbenoid formation and
subsequent cyclopropanation. It might also electronically deactivate
the highly conjugated π system toward cyclopropanation, encouraging
an unwanted decomposition pathway.
(25) Wright, A. C.; Stoltz, B. M. Enantioselective construction of the
tricyclic core of curcusones A−D via a cross-electrophile coupling
approach. Chem. Sci. Advance Article. 2019 DOI: 10.1039/
C9SC04127C.
(15) Siegel, C.; Gordon, P. M.; Razdan, R. K. An Optically Active
Terpenic Synthon for Δ9-Cannabinoids: Synthesis of (−)-11-
Hydroxy-Δ9-tetrahydrocannabinol (THC) and its 1’,1’-Dimethylhep-
tyl Analog. J. Org. Chem. 1989, 54, 5428−5430.
(16) Coupling partner (+)-12 was prepared using (S)-Me-CBS
catalyst.
(17) Reversible Meerwein−Ponnderf−Verley reduction consistently
favored the unwanted trans isomer, indicating that it is the
thermodynamically favored epimer.
(18) All nonasymmetric and irreversible reactions resulted in the
trans isomer as the predominant diastereomeric product, suggesting
that its formation is kinetically favored.
(19) The fragile boronate functionality was incompatible toward
most alcohol protection/deprotection strategies. The sole exception
to this was found with PMB-based protecting groups, which could be
delicately excised under pH-buffered conditions using DDQ.
(20) The sole byproduct of cyclopropanation encountered during
our synthetic studies was identified as ketone 49. A plausible
mechanism for its formation involves intramolecular C−H insertion
of the carbenoid to afford a β-lactone followed by base-induced
lactone opening and fragmentation; see: Xu, X.; Deng, Y.; Yim, D. N.;
Zavalij, P. Y.; Doyle, M. P. Enantioselective cis-β-lactam synthesis by
intramolecular C−H functionalization from enoldiazoacetamides and
derivative donor−acceptor cyclopropenes. Chem. Sci. 2015, 6, 2196−
2201. However, we cannot rule out the following radical-based
fragmentation pathway:
(21) Similar decomposition pathways were observed during our
campaign toward ineleganolide; see: Craig, R. A., II; Roizen, J. L.;
Smith, R. C.; Jones, A. C.; Stoltz, B. M. Enantioselective, convergent
E
DOI: 10.1021/acs.orglett.9b03829
Org. Lett. XXXX, XXX, XXX−XXX