Total Synthesis of the Meroterpenoid Manginoid A as Fueled by a Challenging Pinacol Coupling and Bicycle-forming Etherification

Article abstract of DOI:10.1002/anie.202016178

The manginoids are a unique collection of bioactive natural products whose structures fuse an oxa-bridged spirocyclohexanedione with a heavily substituted trans-hydrindane framework. Herein, we show that such architectures can be accessed through a strategy combining a challenging pinacol coupling and bicycle-forming etherification with several additional chemo- and regioselective reactions. The success of these key events proved to be highly substrate and condition specific, affording insights for their application to other targets. As a result, not only has a 19-step total synthesis of manginoid A been achieved, but a potential roadmap to access other members of the family and related natural products has also been identified.

Full text of DOI:10.1002/anie.202016178

Angewandte
Communications
Chemie
How to cite: Angew. Chem. Int. Ed. 2021, 60, 11127–11132
International Edition: doi.org/10.1002/anie.202016178
German Edition: doi.org/10.1002/ange.202016178
Total Synthesis
Total Synthesis of the Meroterpenoid Manginoid A as Fueled by
a Challenging Pinacol Coupling and Bicycle-forming Etherification
Yu-An Zhang, Amanda Milkovits, Valay Agarawal, Cooper A. Taylor, and Scott A. Snyder*
Abstract: The manginoids are a unique collection of bioactive
natural products whose structures fuse an oxa-bridged spiro-
cyclohexanedione with a heavily substituted trans-hydrindane
framework. Herein, we show that such architectures can be
accessed through a strategy combining a challenging pinacol
coupling and bicycle-forming etherification with several addi-
tional chemo- and regioselective reactions. The success of these
key events proved to be highly substrate and condition specific,
affording insights for their application to other targets. As
a result, not only has a 19-step total synthesis of manginoid A
been achieved, but a potential roadmap to access other
members of the family and related natural products has also
been identified.
Our overall approach to tackling the six stereocenters,
functional group array, and overall level of steric congestion
found within the manginoids is shown retrosynthetically at
the bottom of Scheme 1, using 1 as our inaugural target.
Seeking to take advantage of its steric encumbrance, our final
operation sought to add a lone methyl group with both facial
T
he manginoids (1–7, Scheme 1) are a recently isolated
family of monoterpene shikimic acid-conjugated meroterpe-
noids obtained from the plant pathogen Guignardia mangi-
ferae.[1] Of note, the lead member of this collection, man-
ginoid A (1), can affect potent inhibition (IC50 = 0.84 Æ
0.07 mm) of 11b-hydroxysteroid dehydrogenase type 1 (11b-
HSD1), an enzyme which catalyzes the intracellular conver-
sion of cortisone to bioactive glucocorticoid cortisol; such
modulation could provide a novel therapeutic pathway
against metabolic diseases such as obesity, osteoporosis, and
diabetes.[2] Structurally, 1 and its cousins represent the first
examples of natural products that contain not only a highly
substituted trans-hydrindane framework (colored in blue),
but also a unique oxa-bridged spirocyclohexanedione moiety
(colored in red). Indeed, related molecules (such as 9–11)[3]
are known to possess the first domain, but none have the
second except for allied family members such as 8.[4] Yet,
despite their uniqueness and overall chemical complexity,
only a modest number of synthetic studies toward these
systems and any structurally related natural products have
been reported to date.[5] Herein, we present a synthetic route,
designed around several challenging and chemo- and regio-
selective operations, which can afford access to these frame-
works. These efforts have culminated in a 19-step total
synthesis of manginoid A (1) and provide a blueprint suitable
for addressing related natural product architectures.
[*] Dr. Y.-A. Zhang, A. Milkovits, V. Agarawal, C. A. Taylor,
Prof. Dr. S. A. Snyder
Department of Chemistry, University of Chicago
5735 S. Ellis Avenue, Chicago, IL 60637 (USA)
E-mail: sasnyder@uchicago.edu
Scheme 1. Structures of selected members of the manginoid family
and natural products containing the multi-substituted trans-hydrindane
core (1–11) and a proposed retrosynthetic analysis of manginoid A (1)
based on a critical pinacol coupling and key regioselective transforma-
tions.
Supporting information and the ORCID identification number for
one of the authors of this article can be found under:
https://doi.org/10.1002/anie.202016178.
Angew. Chem. Int. Ed. 2021, 60, 11127 –11132
ꢀ 2021 Wiley-VCH GmbH
11127

Angewandte
Communications
Chemie
and regioselectivity onto a precursor (12) possessing three
ketones. We assumed that the hindrance provided by both the
adjacent spirocyclic quaternary carbon and the oxa-bridge
framework in general would appropriately shield the other
two ketones from attack.[6] In turn, following some changes in
oxidation state and cleavage of the ether bond within the
bicyclic unit, we arrived at compound 13, a material whose
syn-disposed 1,2-diol suggested its potential construction via
a pinacol coupling from keto-aldehyde 14.[7,8] In practice,
however, we expected this reaction to be difficult to execute
for several reasons. Namely, 1) the substrate 14 might be
unstable either on its own accord or under typical pinacol
reaction conditions due to its combination of three potentially
sensitive functional groups: an aldehyde, a b,g-unsaturated
ketone, and an a-oxygenated ketone; 2) given the potential
for the a-oxygenated ketone to undergo b-scission if initiation
began at the neighboring ketone,[9] the SET process would
ideally need to begin from the more reactive, but also more
hindered aldehyde; 3) if the coupling was successful, it was
unclear if the needed stereochemical disposition of the newly
formed diol within 13 would result.
Despite such concerns, we believed such a test was
worthwhile since its success would afford efficient access to
the near-complete architecture of the manginoids. Moreover,
we believed that the overall constraint of the system might
enable its achievement when it would be likely to fail for far
more flexible precursors, thus advancing knowledge regard-
ing challenging pinacol cyclization events. Critically, the idea
would be fairly easy to test since we anticipated that 14 could
be readily prepared from 15 via effective differentiation of its
two primary alcohols, with that intermediate deriving from
the merger of 16 and 17 through a Michael reaction and
subsequent intramolecular allylation using 16 as both a nucle-
ophile and electrophile based on the formal charges shown
(Scheme 1). Ultimately, while this plan was successful, it also
drew on the strategic use of the alkene within the core trans-
hydrindane framework on multiple occasions, both to shield
reactive functional groups, as well as to facilitate the
formation of the final ether linkage leading to 12 through
conformational locking.
As shown in Scheme 2, our efforts commenced with the
preparation of the trans-hydrindane skeleton as expressed by
diol 15. Thus, following isopropenyl addition to commercially
available cyclopentenone to afford 16, subsequent treatment
with KHMDS and TIPSOTf at À788C in THF afforded
a 2.5:1 mixture of regioisomeric silyl enol ethers, favoring
19.[6] Subsequent exposure to SnCl4 in the presence of
methylene dimethyl malonate (17) then afforded compound
20 in 38% yield and undetermined d.r. over two steps.[10]
Significant efforts to improve the yield and regioselectivity of
this sequence were undertaken, but none proved successful;
nevertheless, material processing on scale did not prove
problematic. Pressing forward, the ketone within 20 was then
protected as a ketal using ethylene glycol, trimethyl ortho-
formate, and p-TsOH·H2O in hot toluene (908C); pleasingly,
these conditions also induced epimerization of the a-carbon
to afford 21 as a single diastereomer in 86% yield.[11] With the
trans-disposition of the two sidechains established, our focus
shifted to closing the final six-membered ring needed to
Scheme 2. Synthesis of key diol 15: a) KHMDS (1.4 equiv), TIPSOTf
(1.2 equiv), THF, À78 to 238C, 3 h; b) 17 (1.5 equiv), SnCl4
(0.20 equiv), CH2Cl2, À78 to 238C 1.5 h, 38% from 16, 10% recovered
16; c) ethylene glycol (5.0 equiv), CH(OMe)3 (5.0 equiv), p-TsOH·H2O
(0.10 equiv), toluene, 908C, 12 h, 86%; d) TCCA (1.1 equiv), EtOAc,
À788C, 45 min, 64%; e) NaH (3.0 equiv), DMF, 0 to 238C, 3 h, 81%;
f) LiAlH4, (2.0 equiv), THF, 08C, 15 min, 73%. TCCA=trichloroisocya-
nuric acid.
complete the hydrindane framework. Initial investigations
seeking a direct cyclization using both the White allylic
oxidation protocol[12] as well as Sniderꢀs radical-based oxida-
tive coupling[13] did not afford any desired products; only slow
decomposition of the starting material was observed.[14]
Pleasingly, a two-step alternative rose to the occasion by
using trichloroisocyanuric acid (TCCA) to effect an allylic
chlorination[8d,15] followed by NaH-induced intramolecular
[16]
cyclization. Global reduction of both esters with LiAlH4
then delivered 15 in 7 steps overall.[17] While we note that our
synthesis of 15 was performed racemically for material
processing, it is formally asymmetric given that several
enantioselective syntheses of 16 are known, with one repeated
in comparable yield and enantioselectivity in our hands.[18]
With this material synthesized, we sought next to advance
it to intermediate 29 (cf. Scheme 3), the anticipated precursor
for the key pinacol cyclization. The first critical operation in
that regard was selective functionalization of the equatorially
disposed primary alcohol, noting that such differentiation
simply by reagent control was unlikely to be successful given
the similar steric environments around both of the alcohols.[19]
Indeed, initial probes for mono-oxidation using Dess–Martin
periodinane and CrO3 showed poor selectivity. Thus, we
elected instead to take advantage of the proximity of the
axial-disposed alcohol to the neighboring alkene to forge
a
conformationally locked iodoether protecting group
through the action of NIS.[20] Following the in situ addition
of Dess–Martin periodinane at the end of the sequence,[21]
aldehyde 22 was obtained in 65% overall yield. Then, after
use of the Bestmann modification of the Seyferth–Gilbert
homologation, followed by deprotonation of the newly
formed terminal alkyne and the addition of electrophile 25,
compound 26 resulted as an inconsequential mixture of
diastereomers. Having served its role in facilitating the
construction of the side-chain, the iodoether was reductively
cleaved next through the action of Zn powder in the presence
of AcOH.[20b] The resultant product (27) was then hydro-
genated under the action of Pd/C poisoned by quinoline to
afford the cis-alkene of 28 in 88% overall yield from 26. Of
11128 www.angewandte.org
ꢀ 2021 Wiley-VCH GmbH
Angew. Chem. Int. Ed. 2021, 60, 11127 –11132

Angewandte
Communications
Chemie
Scheme 3. Synthesis of dicarbonyl 29: a) NIS (1.1 equiv), CH2Cl2,
238C, 1 h; then DMP (1.2 equiv), CH2Cl2, 238C, 1 h, 65%; b) Ohira–
Bestmann reagent (23, 1.5 equiv), K2CO3 (2.5 equiv), MeOH, 238C,
2 h, 89%; c) LiHMDS (1.3 equiv), BF3·Et2O (1.3 equiv), 25 (3.0 equiv),
THF, À78 to 238C, 3 h, 75%, 18% recoverd 24; d) Zn (30 equiv),
AcOH (10 equiv), MeOH/Et2O, 408C, 1 h, 96%; e) H2 (balloon), Pd/C
(0.05 equiv), quinoline (0.35 equiv), EtOAc, 238C, 2 h, 96%; f) DMP
(3.5 equiv), CH2Cl2, 238C, 2 h; then workup with aqueous Na2S2O3,
aqueous NaHCO3. NIS=N-iodosuccinimide, HMDS=hexamethyldisi-
lizane, DMP=Dess–Martin periodinane.
Scheme 4. Completion of the total synthesis of manginoid A (1) from
note, the traditional Lindlar catalyst did not effect this
reduction in our hands. In addition, the order of these final
two operations was critical; without the free primary alcohol
within 27, we were never able to complete any alkyne
reduction without significant loss of material through degra-
dation. Hence, this alcohol might be serving as an essential
directing group.[22]
Having reached this stage, all that was needed to access
the requisite 1,6-dicarbonyl of the pinacol coupling precursor
29 was a double oxidation. As originally feared, this event
proved challenging to execute, as initial conditions screened
(such as PCC, PDC, Swern oxidation, and Parikh–Doering
oxidation) led to minimal product formation along with
significant decomposition. Moreover, 29 proved to be highly
unstable on silica gel (even when base-deactivated), making
purification efforts nearly impossible.[9b,23] Fortunately, we
found that if we used Dess–Martin periodinane followed by
dicarbonyl 29: a) SmI2 (6.6 equiv), THF, À788C, 1 h, 35% from 28;
b) NBS (2.1 equiv), THF, 238C, 15 min; then add H2O, THF, 238C,
4 h; then add MeOH, K2CO3 (5.0 equiv), 508C, 4 h, 68%; c) n-BuLi
(4.0 equiv), THF, À788C, 10 min; then add MeOH (8.0 equiv), TBAF
(5.0 equiv), THF, 238C, 2 h; then add MeOH, K2CO3 (10 equiv), 508C,
4 h, 96%; d) DMSO (20 equiv), (COCl)2 (10 equiv), Et3N (30 equiv),
CH2Cl2, À78 to 238C, 4 h, 58%; e) FeCl3 (0.20 equiv), acetone, 238C,
12 h, 83%; f) MeMgBr (2.1 equiv), LaCl3·2LiCl (2.1 equiv), THF, À78
to 238C, 2 h, 42%, 16% recovered 12; g) m-CPBA (1.2 equiv), CH2Cl2,
238C, 2 h, 76%. NBS=N-bromosuccinimide, TBAF=tetra-n-butylam-
monium fluoride; m-CPBA=meta-chloroperoxybenzoic acid.
unidentified by-products. Although the yield was only 10%,
this initial result was encouraging. Efforts at optimization,[25]
some of which are shown in the inset table in the middle
portion of Scheme 4, included additives such as lithium
salts,[25c] alcohols,[7a] and species known to promote the
reducing ability of SmI2 (such as HMPA).[25a,b] In no case
were superior results obtained. However, simply by lowering
the temperature to À788C, we found that the yield could be
improved to a reproducible and readily scalable 35%. Given
the range of failures observed for this reaction in other
venues, this outcome was notable.
a
conventional work-up with both saturated aqueous
NaHCO3 and Na2S2O3 solutions, crude dicarbonyl 29 could
be obtained directly in good yield and of sufficient purity to
press forward. Notably, efforts to prepare simple model
congeners of this 1,6-dicarbonyl in the form of 30 and 31
failed, including under the successful conditions developed to
reach 29.[24]
With 29 finally in hand, we were pleased to find that
treatment with SmI2 in THF at 238C led to the desired diol 32
(Scheme 4) as a single diastereomer alongside a collection of
We presume that this yield enhancement may reflect
greater prevention of decomposition by b-scission,[9] either by
increased reaction initiation at the aldehyde or by modulated
fragmentation if a non-cyclized radical formed first at the
Angew. Chem. Int. Ed. 2021, 60, 11127 –11132
ꢀ 2021 Wiley-VCH GmbH
www.angewandte.org 11129

Angewandte
Communications
Chemie
ketone carbonyl. In addition, we rationalize the observed
formation of the desired 1,2-diol stereoisomer based on the
likely favorability of transition state 38 given the likely
repulsions found in its alternative (39). Of importance, as
intimated earlier the rigidity afforded by the full hydrindane
framework and the alkene linking the two carbonyls appears
necessary for the success of this coupling, as model com-
pounds 40, 41, and 42 all failed under allied conditions.
Having completed this key operation, we next sought to
construct the strained oxa-bridge of the manginoids. Success
required several design iterations and careful analysis of the
overall conformational properties of our intermediates. Our
global approach sought to place an appropriate epoxide onto
the 1,2-disubstituted alkene in hopes that it could be opened
by a primary alcohol to forge the required ether linkage. To
achieve that end, we first needed to engage the more reactive
exocyclic alkene, again as a haloether in the form of 43 (see
the bottom portion of Scheme 4). Unfortunately, upon
attempted oxidation with m-CPBA or DMDO, compound
44 resulted in which the newly formed epoxide was generated
on the undesired side; extensive efforts at a directed reaction
based on the free tertiary alcohol, shown here using VO-
(acac)2 and t-BuOOH, led only to recovered starting mate-
rial.[26] Success was ultimately obtained via more indirect
methods. Specifically, following the formation of 33 from 32
using excess NBS, the presence of water enabled a subsequent
and fully stereo- and regioselective formation of bromohydrin
34; the addition of base in the same pot then afforded the
desired b-disposed epoxide of 35 in 65% yield.[27] From here,
n-BuLi-promoted haloether ring opening[28] followed by silyl
deprotection generated 36, while subsequent base-promoted
etherification completed a second single pot cascade to afford
37 in 96% yield.
With the full tetracycle in place, only a few functional
group manipulations remained to generate 1. Those oper-
ations commenced with a double Swern oxidation followed by
ketal cleavage to afford 12 in 48% overall yield. The first of
these steps was the lower yielding event (58%), as an
unknown major side product was formed. Nevertheless, no
other condition set tested (including Dess–Martin period-
inane and PCC) proved capable of effecting both oxidations
efficiently. For the second, the use of FeCl3 was essential to
avoid racemization of the a-stereocenter, an event that was
observed when the deprotection was conducted with heating
in the presence of p-TsOH·H2O. Finally, a regioselective
addition of a methyl group was successful using a combination
of MeMgBr (2.1 equiv) and LaCl3·2LiCl (2.1 equiv) in THFat
À788C.[29] This process afforded manginoid A (1) as a single
diastereomer (in terms of the methyl addition) in 42% yield,
with all spectral data matching that of Zhu, Zhang, and co-
workers,[1] along with 16% recovered 12 and a complex
mixture of over-addition products. This result confirmed our
original hypothesis that the cyclopentanone was more reac-
tive than the other two ketones, though not perfectly so in
advance of full starting material consumption.
Scheme 5. Exploration of the factors determining success in the final
cyclization leading to 37: a) K2CO3 (5.0 equiv), MeOH, 508C, 12 h.
from 45 (see SI for full details on synthesis) and tested for its
capability to undergo the ring closure to afford 47. While we
anticipated that such a proposed 5-endo-tet cyclization would
be challenging to effect, as it is formally Baldwin forbid-
den,[30,31] computational analysis of the structure conforma-
tions demonstrated a potential to achieve the desired
cyclization. Calculations at the B3LYP/cc-pVTZ level of
theory indicated that of the two possible conformers of 46, B,
where both reactive sites were axially positioned, was favored
over A; such a preference suggested that suitable proximity
might exist to promote the anticipated 5-endo-tet cyclization.
Nevertheless, in the end we were not able to affect the desired
transformation with this substrate. By contrast, in 36 (the
material formed just prior to successful cyclization), although
the necessary conformers for cyclization (D and E) are higher
in energy than conformer C, the ring closure still occurred,
likely because a 5-exo-tet cyclization is a generally more
favorable process and the two reactive domains within E are
fairly close (3.1 ꢁ).
In conclusion, we have achieved the first total synthesis of
manginoid A (1) in 19 steps via a series of unique and highly
regio- and chemoselective transformations. Most critical
among these were the use of 1) reversible haloetherifications
to differentiate both alcohols and alkenes at critical junctures,
2) a facially specific pinacol coupling with a highly sensitive
keto-aldehyde to generate the key spirocyclic system, 3) two
highly orchestrated, one-pot multistep cascade reactions to
forge the final ether linkage of the oxa-bridged core, and 4) a
stereo- and regioselective addition of a methyl group to just
one of three ketones to complete the target. We believe the
developed solutions highlight the power of constrained
frameworks to facilitate difficult transformations and define
a roadmap for the efficient synthesis of other members of the
manginoid collection.
Finally, to put the success of the critical etherification
event in further context, alternate strategies to forge the oxa-
bridged ring were also explored but failed (Scheme 5). In one
of those initial efforts, epoxide 46 was successfully prepared
11130 www.angewandte.org
ꢀ 2021 Wiley-VCH GmbH
Angew. Chem. Int. Ed. 2021, 60, 11127 –11132

Angewandte
Communications
Chemie
[9] a) G. A. Molander, G. Hahn, J. Org. Chem. 1986, 51, 1135;
Acknowledgements
b) C. J. Brumsted, E. L. Carpenter, A. K. Indra, T. Mahmun,
Org. Lett. 2018, 20, 397.
[10] a) P. C. Bulman Page, D. C. Jennens, H. McFarland, Tetrahedron
Lett. 1997, 38, 6913; b) P. C. B. Page, G. Gambera, C. M.
Hayman, M. Edgar, Synlett 2006, 3411.
[11] a) V. V. Veselovsky, B. T. Zhuzbaev, K. M. Turdybekov, S. M.
Adekenov, Y. T. Struchkov, A. M. Moiseenkov, Russ. Chem.
Bull. 1993, 42, 106; b) B. T. Zhuzbaev, V. V. Veselovsky, A. V.
Buevich, A. M. Moiseenkov, Russ. Chem. Bull. 1993, 42, 110.
[12] a) A. J. Young, M. C. White, J. Am. Chem. Soc. 2008, 130, 14090;
b) A. J. Young, M. C. White, Angew. Chem. Int. Ed. 2011, 50,
6824; Angew. Chem. 2011, 123, 6956; c) J. M. Howell, W. Liu,
A. J. Young, M. C. White, J. Am. Chem. Soc. 2014, 136, 5750;
d) W. Liu, S. Z. Ali, S. E. Ammann, M. C. White, J. Am. Chem.
Soc. 2018, 140, 10658.
We wish to acknowledge Dr. Samantha Maki for some
explorations into alternative syntheses of intermediate 19. We
thank Dr. Josh Kurutz, Dr. Antoni Jurkiewicz, and Dr. C. Jin
Qin for assistance with NMR and mass spectrometry,
respectively. Support for the calculations performed was
completed in part with resources provided by the University
of Chicagoꢀs Research Computing Center. Financial support
came from the University of Chicago.
Conflict of interest
[13] a) B. B. Snider, R. Mohan, S. A. Kates, J. Org. Chem. 1985, 50,
3659; b) S. A. Kates, M. A. Dombroski, B. B. Snider, J. Org.
Chem. 1990, 55, 2427.
[14] We ascribe these failures to the weaker nucleophilicity of
malonates versus the b-keto esters more typically deployed in
such chemistries.
The authors declare no conflict of interest.
Keywords: cycloetherification · manginoid · pinacol coupling ·
regioselective · total synthesis
[15] V. V. Veselovsky, B. T. Shuzbaev, A. M. Molseenkov, Mendeleev
Commun. 1992, 2, 167.
[16] B. M. Trost, F. J. Fleitz, W. J. Watkins, J. Am. Chem. Soc. 1996,
118, 5146.
[1] K. Chen, X. Zhang, W. Sun, J. Liu, J. Yang, C. Chen, X. Liu, L.
Gao, J. Wang, H. Li, Z. Luo, Y. Xue, H. Zhu, Y. Zhang, Org. Lett.
2017, 19, 5956.
[2] a) B. S. Hamilton, F. Himmelsbach, H. Nar, A. Schuler-Metz, P.
Krosky, J. Guo, R. Guo, S. Meng, Y. Zhao, D. S. Lala, L. Zhuang,
D. A. Claremon, G. M. McGeehan, Eur. J. Pharmacol. 2015, 746,
50; b) C. G. Fenton, C. L. Doig, S. Fareed, A. Naylor, A. P.
Morrell, O. Addison, C. Wehmeyer, C. D. Buckley, M. S. Cooper,
G. G. Lavery, K. Raza, R. S. Hardy, Arthritis Res. Ther. 2019, 21,
188.
[3] a) N. Tanaka, S. Abe, K. Hasegawa, M. Shiro, J. Kobayashi, Org.
Lett. 2011, 13, 5488; b) J. J. Zhang, X. W. Yang, J. Z. Ma, X. Liu,
L. X. Yang, S. C. Yang, G. Xu, Nat. Prod. Bioprospect. 2014, 4,
73; c) W. J. Xu, J. Luo, R. J. Li, M. H. Yang, L. Y. Kong, Org.
Chem. Front. 2017, 4, 313.
[4] K. Chen, X. Chen, J. Guo, W. Sun, J. Liu, J. Yang, X. Liu, J. Wang,
X. Luo, H. Zhu, Y. Zhang, Phytochemistry 2019, 164, 236.
[5] a) T. Kobayashi, I. Takizawa, A. Shinobe, Y. Kawamoto, H. Abe,
H. Ito, Org. Lett. 2019, 21, 3008; b) Z. Yan, C. Zhao, J. Gong, Z.
Yang, Org. Lett. 2020, 22, 1644; For work towards targets with
related architectures not exactly of the type indicated here, see:
c) Y. Tao, K. Reisenauer, J. H. Taube, D. Romo, Angew. Chem.
Int. Ed. 2019, 58, 2734; Angew. Chem. 2019, 131, 2760.
[6] For examples of total syntheses where high regioselectivity was
enabled by the steric effect of quaternary centers, see: a) P. Hu,
H. M. Chi, K. C. DeBacker, X. Gong, J. H. Keim, I. T. Hsu, S. A.
Snyder, Nature 2019, 569, 703; b) H. M. Chi, C. J. F. Cole, P. Hu,
C. A. Taylor, S. A. Snyder, Chem. Sci. 2020, 11, 10939.
[7] For recent reviews, see: a) D. J. Edmonds, D. Johnston, D. J.
Procter, Chem. Rev. 2004, 104, 3371; b) K. S. Nicolaou, S. Ellery,
J. S. Chen, Angew. Chem. Int. Ed. 2009, 48, 7140; Angew. Chem.
2009, 121, 7276; c) M. Szostak, N. J. Fazakerley, D. Parmar, D. J.
Procter, Chem. Rev. 2014, 114, 5959.
[8] For selected recent examples of SET-mediated couplings in
natural product synthesis, see: a) K. V. Chuang, C. Xu, S. E.
Reisman, Science 2016, 353, 912; b) L. Cai, K. Zhang, O. Kwon,
J. Am. Chem. Soc. 2016, 138, 3298; c) D. Chen, P. A. Evans, J.
Am. Chem. Soc. 2017, 139, 6046; d) E. P. Farney, S. S. Feng, F.
Schꢂfers, S. E. Reisman, J. Am. Chem. Soc. 2018, 140, 1267; e) Y.
Imamura, S. Yoshioka, M. Nagatomo, M. Inoue, Angew. Chem.
Int. Ed. 2019, 58, 12159; Angew. Chem. 2019, 131, 12287; f) A.
Han, Y. Tao, S. E. Reisman, Nature 2019, 573, 563; g) K. R.
Owens, S. V. McCowen, K. A. Blackford, S. Ueno, Y. Hirooka,
M. Weber, R. Sarpong, J. Am. Chem. Soc. 2019, 141, 13713.
[17] For an alternate synthesis of diol 15, see the Supporting
Information. This second route was initially utilized to make
several grams of material, but was ultimately abandoned as the
route described here was both more concise and could deliver 15
in comparable overall yield.
[18] a) T. Robert, J. Velder, H. G. Schmalz, Angew. Chem. Int. Ed.
2008, 47, 7718; Angew. Chem. 2008, 120, 7832; b) G. Lalic, E. J.
Corey, Tetrahedron Lett. 2008, 49, 4894; c) Q. Naeemi, T. Robert,
D. P. Kranz, J. Velder, H. G. Schmalz, Tetrahedron: Asymmetry
2011, 22, 887; d) A. Cernijenko, R. Risgaard, P. S. Baran, J. Am.
Chem. Soc. 2016, 138, 9425; We specifically repeated the
procedure of ref. [18b] on 0.5 mmol scale twice (with some
variation in ligand loading) and garnered material in 83% and
93% ee, respectively in comparable yields as that reported. See
SI for full details.
[19] N. Hartrampf, N. Winter, G. Pupo, B. M. Stoltz, D. Trauner, J.
Am. Chem. Soc. 2018, 140, 8675.
[20] a) K. C. Nicolaou, N. A. Petasis, R. E. Zipkin, J. Uenishi, J. Am.
Chem. Soc. 1982, 104, 5555; b) B. M. Trost, M. J. Krische, J. Am.
Chem. Soc. 1996, 118, 233.
[21] J. Zhang, M. Liu, C. Wu, G. Zhao, P. Chen, L. Zhou, X. Xie, R.
Fang, H. Li, X. She, Angew. Chem. Int. Ed. 2020, 59, 3966;
Angew. Chem. 2020, 132, 3994.
[22] G. Stork, D. E. Kahne, J. Am. Chem. Soc. 1983, 105, 1072.
[23] a) I. Storch de Gracia, H. Dietrich, S. Bobo, J. L. Chiara, J. Org.
Chem. 1998, 63, 5883; b) J. L. Chiara, ꢃ. Garcꢄa, E. Sesmilo, T.
Vacas, Org. Lett. 2006, 8, 3935.
[24] These results suggest that such an array of functional groups is
highly unstable, and that the presence of the hydrindane system
confers higher stability to 29 than its simplified congeners.
[25] a) J. Inanaga, M. Ishikawa, M. Yamaguchi, Chem. Lett. 1987, 16,
1485; b) J. B. Shotwell, J. M. Sealy, R. A. Flowers, J. Org. Chem.
1999, 64, 5251; c) R. S. Miller, J. M. Sealy, M. Shabangi, M. L.
Kuhlman, J. R. Fuchs, R. A. Flowers, J. Am. Chem. Soc. 2000,
122, 7718.
[26] Other efforts included the use of Ti(Oi-Pr)4 in varied solvents,
but no directed epoxidation method delivered the desired
product; recovered starting material and/or decomposition was
the more common outcome.
[27] C. P. Ting, G. Xu, X. Zeng, T. J. Maimone, J. Am. Chem. Soc.
2016, 138, 14868.
Angew. Chem. Int. Ed. 2021, 60, 11127 –11132
ꢀ 2021 Wiley-VCH GmbH
www.angewandte.org
11131

Angewandte
Communications
Chemie
[28] Both the aforementioned Zn-mediated ring-opening conditions
and SmI2 gave no desired product.
Cook, Org. Lett. 2012, 14, 3340; c) M. Iwakura, H. Tokura, K.
Tanino, Tetrahedron Lett. 2017, 58, 1223.
[29] A. Krasovskiy, F. Knopp, P. Knochel, Angew. Chem. Int. Ed.
2006, 45, 497; Angew. Chem. 2006, 118, 511.
[30] J. E. Baldwin, J. Chem. Soc. Chem. Commun. 1976, 743.
[31] For selected examples of successful 5-endo-tet cyclizations, see:
a) G. E. Veitch, A. Pinto, A. Boyer, E. Beckmann, J. C.
Anderson, S. V. Ley, Org. Lett. 2008, 10, 569; b) P. Gao, S. P.
Manuscript received: December 4, 2020
Revised manuscript received: February 17, 2021
Accepted manuscript online: February 28, 2021
Version of record online: April 6, 2021
11132 www.angewandte.org
ꢀ 2021 Wiley-VCH GmbH
Angew. Chem. Int. Ed. 2021, 60, 11127 –11132