Li et al.
Page 2
Retrosynthetically, we envisaged that cyclohexene derivative 1 with an appropriate R
substituent would serve as a common advanced intermediate for all three natural product
targets: (+)-asperpentyn, (-)-harveynone and (-)-tricholomenyn A (Scheme 2). A direct
precursor of 1 would be an acyclic 1,3-diyne-contaning compound 2 or its simpler variant
lacking the relay device. The pivotal metathesis substrate 2 would be prepared through
flouoride-catalyzed addition of silylated 1,3-diyne 4 to epoxy aldehyde 3, which in turn would
14
be derived from aldehyde 5 and allyl propargyl ether 6. To test the feasibility of the key step,
alkene-tethered 1,3-diyne 11 was prepared from commercially available cis-2-buten-1,4-diol
7 in 8 steps (Scheme 3). Following the known procedure involving the Sharpless asymmetric
15
epoxidation , the diol 7 was elaborated to epoxide 8. Oxidation of the primary alcohol with
PCC to generate aldehyde 5 was followed by addition of vinyl magnesium bromide, MOM-
protection of the resultant secondary alcohol (1:1 mixture) and removal of the TBS group
16
afforded primary alcohol 9. After Dess-Martin oxidation of the primary alcohol, a lithiated
diyne derived from 10 was added to provide RCM substrate 11. Disappointingly, however,
treatment of 11 with Grubb's second-generation catalyst did not effect the ring closure to
generate 12. We assumed that the initiation was hindered by the steric congestion around the
17
vinyl group.
To overcome this hindered initiation, a relay metathesis strategy was adopted, which entails
18
the preparation of more elaborated RCM substrate 2 containing the relay device. Along this
modified plan, the first goal is to synthesize a common intermediate that can branch off to
different target molecules, which is aldehyde 16 (Scheme 4). The synthesis of 16 was
commenced with the addition of acetylide derived form allyl propargyl ether 6 to an aldehyde
derived from alcohol 8, providing separable diastereomeric alcohols 13 and 14 in a 1:2 ratio.
The desired β-epimer 13 was easily elaborated to acetate 15 via controlled partial
19
hydrogenation (H , Pd/CaCO , Pb(OAc) , quinoline, Hexane/EtOAc 1:1) and acetylation
2
3
4
(Ac O, pyr, DMSO, DCM). For a more practical material throughput, without separation the
2
mixture of two epimers, 13 and 14 were subjected to a four-step sequence to convert to 15,
which involves Dess-Martin oxidation of the secondary alcohols, (R)-Me-CBS mediated
20
reduction, partial hydrogenation of the triple bond, and Mitsunobu reaction with acetic acid.
21
Removal of the TBS group from 15, followed by oxidation of the corresponding primary
alcohol gave the aldehyde 16.
For the synthesis of asperpetyne and harveynone, aldehyde 16 was reacted with
22
triethylsilyl-1,3-pentadiyne 17 and a catalytic amount of the anhydrous fluoride source
23
tetrabutylammonium difluorotriphenylsilicate (TBAT), providing enyne RCM substrate
18 after silyl concomitant protection of the secondary alcohol (Scheme 5). Treatmenent of
24
18 with Grubbs second-generation catalyst in a dilute solution of CH Cl , a mixture of
2
2
epimers 23 and 24 was isolated in 62% yield along with unidentified byproducts.
Based on the level of our understanding, we believe that the overall metathesis process started
from the teminal alkene of the allyl ether relay device to form 19 initially, which then delivers
the ruthenium moiety intramolecularly to the cis- alkene to generate a new propagating
alkylidene 20. Subsequent enyne RCM generating alkynyl Ru- alkylidene 21 would induce
facile metallotropic [1,3]-shift to generate fully conjugated alkylidene 22. The termination at
the sterically less hindered carbon through 22 would ultimately deliver the final products 23
13
and 24, thereby establishing 1,5-diene-3-yne substructure.
After separation, the C1-β-epimer 23 was elaborated to (+)-asperpentyn through the removal
of both the C1-TES group and C4-acetate in one step (KCN, EtOH 95%) (Scheme 6). Also,
the TES-group on C1-β-epimer of 23 was selectively deprotected to generate alcohol 25
(HF·pyr, pyr, CH Cl , 0 °C), which was then oxidized to acetylated harveynone 26.
2
2
Unfortunately, under a variety of conditions, the C4-acetate of 26 could not be removed
Org Lett. Author manuscript; available in PMC 2010 February 5.
Li, Jingwei
