and co-workers achieved the first asymmetric total synthe-
sis of solanoeclepin A in 2011.6
(Scheme 2) was found to generate the requisite trans-fused
rings in model experiments on 9a (X = H, Y = OTBS).
The terminal acetylenic moiety of 9a was first deproto-
nated with nBuLi and protected with TMSCl. Then a
smooth deprotonation of the propargylic proton took
place with the addition of a second equivalent of nBuLi
inTHF atꢀ78°C (ConditionsA), and a spontaneous[2,3]-
Wittig rearrangement proceeded at the same temperature
to provide the propargylic alcohol 10a as a single stereo-
isomer in one pot with an 85% overall yield. Furthermore,
after methylation and removal of the TMS group, the
NMR analysis of 12 revealed that the newly generated
stereogenic C11 was of an R-configuration.12 These experi-
ments also confirmed that the A/B rings of the hexahy-
droindene derivative 10 was trans-fused.
One of the challenges of synthesizing solanoeclepin
A is the construction of the highly strained tricyclo-
[5.2.1.01,6]decene skeleton 4, which includes four-, five-,
and six-membered rings and bears three contiguous qua-
ternary stereogenic centers. There have only been three
reports of the syntheses of the tricyclo[5.2.1.01,6]decane core
of solanoeclepin A to date: (i) intramolecular [2 þ 2] photo-
cycloaddition of alkene-dioxenone or allene butenolide,3 (ii)
base-induced intramolecular cyclization of an epoxynitrile,6
and (iii) 4-exo-trig radical cyclization of the cyclobutane.5
In this letter, we showcase a fourth example as an alter-
native synthesis of the tricyclic, cyclobutane-containing
framework 4 by exploiting a HosomiꢀSakurai type cycli-
zation of an acetyleneꢀdicobalthexacarbonyl complex.
The key step is generation of the Nicholas type carbenium
ion,7 which had been employed in our previous work as the
main strategy for the cyclization of seven-, eight-, and nine-
membered ether rings, in the total synthesis of ciguatoxin.8
In our retrosynthetic analysis of solanoeclepin A, the
functionalized seven-membered carbocyclic B-ring is
cleaved into two segments (Scheme 1): the 7-oxabicyclo-
[2.2.1]heptanone 2 and the highly strained tricyclic moiety
3 bearing three quaternary centers. We intended to develop
a stereocontrolled synthesis of the tricyclic subunit 4 and
envisaged that its cyclobutane moiety could be acquired by a
HosomiꢀSakurai type cyclization through a Nicolas cation
5 generated in situ from an acetyleneꢀdicobalthexacarbonyl
complex under Lewis acidic conditions. According to the
stereochemistry of the cyclobutyl moiety, a trans-stereo-
chemistry at the fused rings of 6 is required. Therefore,
the propargyl group at the C4 bridgehead position was
required to be trans with respect to the methyl group at the
C9 junction. This stereochemically defined intermediate 6
could be procured through a [2,3]-Wittig rearrangement9 of
β-propargyl ether 7, which could be prepared in turn from
HajosꢀParrish ketone 8.
Scheme 1. Retrosynthetic Analysis of Solanoeclepin A
In the [2,3]-Wittig rearrangement of 9b having a β-
substituent on C1 (X = ;CtC;SiMe3, Y = OTBS),
the rearrangement was impeded by the bulky β-TMS
acetylenic group. Under the same reaction conditions, only
a very small amount of the desired rearrangement product
10b was observed by TLC. When the reaction temperature
was raised from ꢀ78 to 28 °C for 3.5 h (Conditions B), a
9% yield of 10b was obtained, along with elimination
product 11b (75% yield) as the major product. On the
other hand, the reaction of 9c bearing an exocyclic olefin
(X = Y = ;CdCHCH2OTBS) at C1 provided the
desired [2,3]-Wittig rearrangement product 10c in 70%
yield over two steps (Conditions A).
Having accomplished the [2,3]-Wittig rearrangement to
give the desired trans-octahydroindene system, we pro-
ceeded to synthesize the right-hand segment 4 of solanoe-
clepin A from HajosꢀParrish ketone 8 (Scheme 3). First,
the R,β-unsaturated ketone was selectively protected
to give monoethylene-ketal 13 in 90% yield following
Wicha’s method using 1,2-bis(trimethylsiloxy)ethane and
The incorporation of substituents to generate the qua-
ternary C4 with the correct stereochemistry at the bridge-
head for producing the trans-fused octahydroindane ring
in solanoeclepin A was a synthetic challenge. Conjugate
addition of cuprates to 8 is known to produce the cis-fused
bicyclo[4.3.0] framework, except in the case of the Nagata-
hydrocyanation reaction,10 which also yielded a minor
amount of the trans-fused product. All attempts to use
[3,3]-sigmatropic rearrangements (e.g., the Irelandꢀ
Claisen rearrangement) failed to build the trans-fused ring
junction, including the use of 6-β-thiophenyl acetate.11
On the other hand, the use of a [2,3]-Wittig rearrangement
(6) Tanino, K.; Takahashi, M.; Tomata, Y.; Tokura, H.; Uehara, T.;
Miyashita, M. Nat. Chem. 2011, 3, 484–488.
(7) Connor, R. E.; Nicholas, K. M. J. Organomet. Chem. 1977, 124,
C45–C47.
(8) Hamajima, A.; Isobe, M. Angew. Chem., Int. Ed. 2009, 48, 2941–
2945.
(9) (a) Nakai, T.; Mikami, K. Chem. Rev. 1986, 86, 885–902. (b)
Nakai, T.; Tomooka, K. Pure Appl. Chem. 1997, 69, 595–600.
(10) Nagata, W.; Yoshioka, M. Org. React. 1977, 25, 255–476.
(11) See the Supporting Information.
(12) It is mechanistically assumed from Nakai et al. (ref 9) and
from similar [2.3]Wittig rearrangement product 22 and crystalline
compound 23.
B
Org. Lett., Vol. XX, No. XX, XXXX