the unwanted Z isomer. Although adjusting the reaction
conditions had little impact on this ratio, it was possible to
equilibrate the mixture. Exposure of 8 to thiophenol and
AIBN in refluxing benzene8 led to a 9:91 mixture of Z and
E alkenes from which pure 8E could be isolated by flash
chromatography. In practice, pure 8E could be isolated from
7 in 93% yield from a process involving two olefination
reactions and three isomerization cycles. The stereochemical
of the cyclohexenone A ring in 13 by aldol cyclization.
Several sets of reaction conditions were tried, the best
proving to be treatment with pyrrolidine and acetic acid in
benzene at ambient temperature. The C-3 R-methoxyl group
was installed by Luche reduction11 of the enone followed
by mesylation and methanolysis. It is significant that
methanolysis at room temperature produced a 4:1 mixture
of the R- and â-methoxyl isomers. In contrast, extended
methanolysis at lower temperatures (-25 °C) afforded a 10:1
mixture of the same compounds. It should be noted that 1H
NMR analyses of all intermediate compounds between 11
and 14 were complicated by the presence of formamide
rotational isomers (∼3:2 ratio). Compound 14 (racemic) has
been previously prepared and carried on to racemic haeman-
thidine, pretazettine, and tazettine, and spectral data for the
optically active 14 described here were in complete accord.3h,i
1
assignment of alkenes 8E and 8Z was based on H NMR
data, specifically the chemical shift of the vinyl protons in
the two isomers. This proton was observed at δ 6.29 for the
desired isomer 8E and at δ 5.97 for isomer 8Z, a trend that
has been consistent for several pairs of related compounds
in this series. Further justification for this assignment was
the significant NOE enhancement observed between the vinyl
proton of 8Z and the isolated ortho proton in the aryl ring;
a similar enhancement was not seen for isomer 8E.
Transformation of 14 to the title alkaloids pretazettine (1),
haemanthidine (2), and tazettine (3) was accomplished
efficiently following the original Wildman protocol,2 as
practiced by Martin.3h,i Thus, Bischler-Napieralski12 cy-
clization of 14 followed by pivalate hydrolysis afforded
optically pure haemanthidine (2) in 63% yield. Spectroscopic
data for this material, including specific rotation, were
consistent with those reported for the natural alkaloid.
Treatment of haemanthidine (2) with methyl iodide followed
by stirring with 0.01 M aqueous HCl afforded, after workup
and flash chromatography, optically pure pretazettine (1) in
95% yield. Finally, exposure of pretazettine to 0.1 M aqueous
NaOH led to rapid (<30 min) conversion to tazettine (3) in
91% purified yield. All spectroscopic data were in complete
accord with those reported for this material. Moreover, the
specific rotations of these materials verify that the absolute
stereochemistry of (+)-pretazettine, (+)-tazettine, and (-)-
haemanthidine are as indicated in structures 1, 2, and 3,
respectively.
The necessary stereochemical outcome of the cycloaddition
step at C-4a and C-6b (cis ring fusion, pretazettine number-
ing) was anticipated on the basis of earlier work. For instance,
one set of model studies suggested that the acetonide ring
would lead to a slight preference (∼3:1) for a cis/anti/cis
cycloadduct.5b Furthermore, other studies had indicated that
the (R)-R-methylbenzyl substituent on the nitrone nitrogen
would favor the creation of the C-4a stereogenic center with
the desired S configuration, creating the opportunity for an
exercise in double diastereoselection.5c,d Hydrolysis of the
dimethyl acetal followed by treatment with (R)-R-methyl-
benzyl hydroxylamine9 then afforded the nitrone, which
without purification was heated at 80 °C in benzene for 17
h to yield cycloadduct 9 in 65% yield after purification by
1
flash chromatography. The room temperature 400 MHz H
NMR spectrum of 9 was difficult to interpret because of
significant broadening of the signals associated with the
isoxazolidine ring. At 90 °C, however, the spectrum was well
resolved.
The considerable length of the synthetic scheme reported
here is largely the result of our interest in confirming the
absolute stereochemistry of the target alkaloids. Thus, of the
total synthetic manipulations required to produce alkaloids
1-3 from R-methyl-D-mannopyranoside, nearly half (12)
were needed to obtain aryl ketone 7. We are currently
developing alternative efficient routes to 7 and related
compounds that take advantage of asymmetric alkene oxida-
tion protocols.
The transformation of cycloadduct 9 to the target alkaloids
1-3 was accomplished as outlined. Thus, hydrogenolysis
of 9 afforded hydroxy lactam 10 in 95% yield after cleavage
of the N-O and benzylic nitrogen bonds followed by
intramolecular transamination. That the structure of 10 is as
indicated was verified by single-crystal X-ray analysis,
proving that the important trans relative stereochemical
relationship between the aryl and C6a-hydroxyl group had
been established. Moreover, by relating these stereocenters
to C-4 of the mannose starting material (* in 10), the absolute
configuration of 10, and of pretazettine (1), haemanthidine
(2), and tazettine (3), have been verified.10
Acknowledgment. We gratefully acknowledge NSF
(CHE 95 22580) for financial assistance in the purchase of
the 400 MHz NMR instrument used in support of this work.
We also thank Dr. Peter White of the Department of
Chemistry at the University of North Carolina - Chapel Hill
Conversion of 10 to formamide 14 was accomplished as
indicated, the critical transformation being the construction
(7) (a) Shimoji, K.; Taguchi, H.; Oshima, K.; Yamamoto, H.; Nozaki,
H. J. Am. Chem. Soc. 1974, 96, 1620. (b) For a review of Peterson-type
olefination reactions, see: Kelly, S. E. In ComprehensiVe Organic Synthesis;
Trost, B. M., Fleming, I., Eds.; Pergamon Press: Elmsford, NY, 1991; Vol.
1, pp 729-817.
(8) Schwarz, M.; Graminski, G. F.; Waters, R. M. J. Org. Chem. 1986,
51, 260. This equilibration method, originally applied to disubstituted
alkenes, has been applied to the trisubstituted analogues for the current
work. Details will be provided in due course.
(10) The original assignment of absolute configuration to these alkaloids
was based on tenuous circular dichroism studies, eventually confirmed by
the conversion of tazettine to (+)-3-methoxyadipic acid (known configu-
ration). Highet, R. J.; Highet, P. F Tetrahedron Lett. 1966, 4099.
(11) (a) Luche, J. L.; Rodrigues, H. L.; Cragge, P. J. Chem. Soc., Chem.
Commun. 1978, 601. (b) Luche, J. L. J. Am. Chem. Soc. 1978, 100,
2226.
(9) Wovkulich, P. M.; Uskokovic, M. R. Tetrahedron 1985, 41, 3455.
An improved preparation of this material will be reported in the full paper.
(12) Fodor, G.; Gal, J.; Phillips, B. Angew. Chem., Int. Ed. Engl. 1972,
11, 919.
Org. Lett., Vol. 2, No. 2, 2000
101