.
Angewandte
Communications
thesis of bicyclo[3.3.1]nonadione 10, which could serve as
a pivotal intermediate for the assembly of various PPAPs (12)
through late-stage functionalization at C1 and C3 on inter-
mediate 11.
The synthesis began with a one-pot allylation/alkylation of
ketone 5 by using LDA and allylbromide, followed by a 1,2-
other hand, enone 13b was subjected to the above aldol
reaction conditions and propargyl alcohol 14b was isolated in
63% yield along with a mixture of three other minor
diastereomers (ca. 25% combined yield). At the outset, we
had concerns about the diastereoselectivity of the organo-
cuprate addition. Indeed, the addition of organocopper
reagents to d-substituted cyclohexenones[17] such as 14b
gives mainly the trans isomer through an attack anti to the
adjacent substituent.[18] In our case, the addition must proceed
through a syn pathway in order to obtain the required
stereochemistry at C8. To our delight, treating 14b with
MeMgBr in the presence of CuI followed by an oxidation
gave 15b as the sole diastereomer in 73% yield over two steps
(d.r. > 20:1). The unusual diastereoselectivity can be ration-
alized through the formation of a Mg-chelated intermediate
such as 16. The latter can adopt a conformation in which the
axial substituent shields the top face, thus favoring syn
addition of the organocopper reagent to provide 17b as the
major diastereomer.[19] Next, the replacement of the TMS
group by a bromide was achieved without incident by using
AgNO3 and NBS (15!8). The resulting bromoethynyl
compounds were transformed into the corresponding silyl-
enol ethers 8a–c in readiness for a gold(I)-catalyzed 6-endo-
dig carbocyclization.
=
addition of the appropriate alkylating agent (H2C
CHCH2CH2MgBr or MeLi) and subsequent treatment with
HCl 1n to give enones 6a and 6b in 66% and 83% yields,
respectively (Scheme 2). A second alkylation with LDA and
R2CH2–X provided enones 13a–c in 87%, 95%, and 72%
yields, respectively. Next, the installation of the propargyl unit
was realized through the addition of 3-(trimethylsilyl)-2-
propynal 7 to enones 13a–c in the presence of LDA. The
alcohols 14a and 14c were isolated as the major diastereo-
mers at C5–C7 in 67% and 66% yields, respectively (d.r. 3:1
at C4).[16]
Alcohols 14a and 14c underwent a sequence of conju-
gated addition/oxidation reactions to afford ketones 15a and
15c in 72% and 78% yields, respectively, over 2 steps. On the
With [(JohnPhos)Au(NCMe)][SbF6] as the catalyst
(5 mol%), the silylenol ethers 8a–c were converted into the
desired bicyclic ketones 10a–c in high yields (Scheme 3).
Remarkably, the gold(I)-catalyzed cyclization proceeded in
a sterically crowded environment and was amenable to large-
scale synthesis (> 10 g). Treating 10a–c with MeONa in
MeOH followed by silylation at C3 produced the bridgehead
ketones 18a–c. The latter compounds represent key inter-
mediates that can be transformed into a wide range of
naturally occurring PPAPs. For instance, compound 18c was
converted into ketone 19 through a high-yielding cross-
metathesis reaction by using isobutene and Grubbs II as the
catalyst.[10c] Simpkins and co-workers reported the synthesis
of nemorosone (3) in five steps from intermediate 19.[10b] The
synthesis of intermediate 19 thus completes a formal total
synthesis of the natural product.
The preparation of hyperforin (1) and papuaforin A (2)
from intermediates 18b and 18a, respectively, was then
examined (Scheme 4). Initial attempts at direct acylation at
the sterically congested C1 position by using a lithium amide
base and an acyl chloride, acyl cyanide, or aldehyde have thus
far been unproductive. In light of the previous work of
Danishefsky[10a,11b] and Simpkins,[10b,11c] we envisaged the
installation of the acyl group via a bridgehead iodo inter-
mediate. After considerable experimentation, we found that
deprotonation with LDA (3.5 equiv) in the presence of
TMSCl (5 equiv) followed by the addition of iodine gave
the corresponding iodoketones 20a and 20b in 29% and 18%
yields, respectively, along with significant amounts of reduced
by-products 21a,b (41–42%) and recovered starting materi-
als 18a,b (24–35%). Although a competitive LDA-mediated
reduction was operative,[20] the formation of the bridgehead
iodo intermediates 20 proved to be highly reproducible and
amenable to gram-scale synthesis.[21] Iodo-lithium exchange/
alkylation followed by oxidation and treatment with TBAF
Scheme 2. Syntheses of enol ethers 8a–c. Reagents and conditions.
=
a) 1. LDA, allyl bromide, THF, ꢀ788C, 2 h; 2. MeLi or H2C
CHCH2CH2MgBr, ꢀ788C then 3. HCl (1n); b) LiHMDS, R2CH2X, THF;
c) LDA, 7, THF, ꢀ788C; d) MeMgBr, CuI, DMS, THF, ꢀ788C; e) DMP,
dichloromethane, RT; f) AgNO3, NBS, RT; g) TBSOTf, DTMP, dichloro-
methane. LDA=lithium diisopropylamide, THF=tetrahydrofuran,
DMS=dimethyl sulfide, DMP=Dess–Martin periodinane, NBS=N-
bromosuccinimide, TBSOTf=tert-butyldimethylsilyl triflate,
DTMP=2,6-di-tert-butyl-4-methylpyridine.
2
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
These are not the final page numbers!