Fig. 1 Key nOe enhancements for diol 14.
isolation of a diastereomerically pure aldol product (14) in 65%
yield (based on 4) whose spectroscopic data were consistent
with the opposite stereochemistry for the hydroxyl group as
required for vannusal A (anti as opposed to the desired syn).
Since this stereochemical outcome could, in principle, be
rectified later on in the sequence through standard chemistry,
we proceeded to the next stage which required protection of the
newly generated hydroxyl group as a TES ether (TESCl,
imidazole, 72% yield) furnishing protected diol 15. Concerned
with potential epimerization and/or elimination problems
during the pending olefination of the cyclopentanone moiety in
Scheme 3 Synthesis of vannusal AAs polycyclic skeleton (2): (a) 3, LiHMDS
2.2 equiv.),THF, 278 °C, 1 h; then 4, 1 h, 278 °C, 65%; (b) TESCl (1.2
equiv.), imidazole (2.5 equiv.), DMAP (0.5 equiv.), DMF, 25 °C, 36 h, 72%;
c) Zn (10.0 equiv.), PbCl (0.01 equiv.), TMSCl (0.1 equiv.), CH (20.0
equiv.), THF, 25 °C, 0.5 h; then TiCl (3.0 equiv.), 5 h, 88%; (d) catalyst 17,
0.3 equiv.) CH Cl , 50 °C, 12 h, 84%; (e) PPTS (0.3 equiv.), MeOH,
CH Cl , 25 °C, 6 h, 91%; LiHMDS = lithium salt of 1,1,1,3,3,3-hexame-
1
5 under the standard Wittig conditions, we turned to a
(
8
modified Takai protocol in which the action of the obligatory
reagents (Zn, CH , and TiCl ) was enhanced by the addition of
catalytic amounts of PbCl and TMSCl. This reaction per-
formed admirably in this instance to afford the desired olefin
16) in 88% yield.9 Pleasantly, exposure of 16 to Grubbs’
2
I
2
4
(
2
2 2
I
2
4
(
2
2
(
2
2
thyldisilazane, TES = triethylsilyl, PPTS = pyridinium p-toluenesulfo-
nate.
second generation catalyst 17 formed the missing cyclohexene
ring and subsequent PPTS promoted desilylation led to the
targeted vannusal skeleton 2 in 76% yield.10 With the complete
fused polycyclic framework in hand, our attention was finally
turned to deducing the global relative stereochemistry of the
synthesized molecule. As shown in Fig. 2, all of the stereogenic
centers of 2‡ correspond, as indicated by the nOe enhance-
ments, to that of 1, except for the free hydroxyl group which
resides in the pseudo-equatorial position.
ing Grignard reagent under modified Kharasch3 conditions
(
CuI–LiCl) to afford the corresponding enolate, which was
4
trapped with TMSCl as the silyl enol ether. Treatment of the
latter product with aq. HF removed both silyl groups and
furnished intermediate 8, with the trans arrangement of the two
appendages on the six-membered ring, as a mixture of
inconsequential geometrical isomers in 71% overall yield.
Oxidation of the primary alcohol in 8 with PCC led to the
corresponding keto-aldehyde (89% yield), whose aldehyde
function was then selectively converted to a dimethoxy acetal 9
4
by treatment with catalytic amounts of NH Cl in refluxing
MeOH (95% yield) thereby setting the stage for the crucial
intramolecular spirocyclization via a Mukaiyama-type aldol
reaction. Gratifyingly, upon exposure of acetal 9 to excess
TMSI in the presence of HMDS, cyclization proceeded to afford
the coveted spiro compound 10 in 50% yield as a single
stereoisomer (except for the olefin geometry which was
Fig. 2 Key nOe enhancements for polycycle 2.
The described synthesis of vannusal A’s fused polycyclic
skeleton represents a novel approach to this natural product and
demonstrates the power of modern synthetic technologies to
construct complex molecular architectures. Efforts toward
implementing this strategy to effect the total synthesis of
vannusal A and B are now under way.
We thank the NIH, the Skaggs Institute for Chemical
Biology, and the George E. Hewitt Foundation (Fellowship to
M. P. J.) for financial support.
5
maintained as a mixture). With the first quaternary center fixed
in intermediate 10, we then turned our attention to the
introduction of the second such motif. Thus, refluxing spiro-
cycle 10 in THF with excess KH and (MeO)
quantitatively, keto-ester 5 as a mixture of two diastereomers
95% total yield). Noteworthy is the fact that, just as in the case
2
CO afforded,
(
of the geometrical isomerism, the stereochemistry of the ester
group in 5 was irrelevant to the stereochemical outcome of the
pending ring closure. Indeed, treatment of the latter compound
3 2
(5) with Mn(OAc) and Cu(OAc) in hot acetic acid (80 °C)
according to the Snider protocol furnished the expected keto-
Notes and references
ester 11 in 76% yield as a single diastereomer via the anticipated
1
G. Guella, F. Dini and F. Pietra, Angew. Chem., Int. Ed., 1999, 38,
134.
6
radical pathway. The next step toward 4 was the reduction of
1
1
1 to the corresponding diol, an objective which was accom-
2
3
B. H. Ridgway and K. A. Woerpel, J. Org. Chem., 1998, 63, 458.
M. S. Kharasch and P. O. Tawney, J. Am. Chem. Soc., 1941, 63,
plished stereoselectively with DIBAL. Much to our delight, the
resulting relative stereochemistry within 12 correlated with the
required configuration of 1 as evidenced by the observed nOe
enhancements (see Fig. 1). The primary hydroxyl of 12 was
2
308.
4 M. T. Reetz and A. Kindler, J. Organomet. Chem., 1995, 502, C5.
5 Y. Tokunaga, M. Yagihashi, M. Ihara and K. Fukumoto, J. Chem. Soc.,
Perkin Trans. 1, 1997, 189.
7
selectively oxidized by the BAIB-TEMPO protocol to furnish
6
7
B. B. Snider, Chem. Rev., 1996, 96, 339.
A. De Mico, R. Margarita, L. Parlanti, A. Vescovi and G. Piancatelli, J.
Org. Chem., 1997, 62, 6974.
aldehyde 13 in 75% overall yield. Subsequent acetylation
2
(Ac O, py) of the free secondary alcohol furnished the protected
aldehyde 4 in 72% yield.
8
9
K. Takai, Y. Hotta, K. Oshima and H. Nozaki, Tetrahedron Lett., 1978,
With the required fragment 4 readily available, attention was
then focused on coupling the two fragments and then executing
the final elaborations to the target structure, as shown in Scheme
1
9, 2417.
K. Takai, T. Kakiuchi, Y. Kataoka and K. Utimoto, J. Org. Chem., 1994,
9, 2688.
5
3
. Thus, treatment of cyclopentanone (3) with excess LiHMDS
10 M. Scholl, S. Ding, C. W. Lee and R. H. Grubbs, Org. Lett., 1999, 1,
in THF at 278 °C, followed by addition of aldehyde 4, led to the
953.
CHEM. COMMUN., 2002, 2480–2481
2481