cyclohexene acid as 3.3:1 mixture of stereoisomers (endo:exo).†
Much experimentation was required to discover the most suit-
able substituents, dieneophile, and reaction conditions to min-
imise polymerisation. Acetylation followed by separation of the
diastereomers gave the desired endo acid 7. The acid was con-
verted into the corresponding alcohol in a one-pot procedure
whereby the acid was first converted to the mixed anhydride
with diisopropylethylamine and isobutyl chloroformate and
graphic studies on the nitrone. In addition the heterocyclic ring
closure to the amide 11 would have not been possible if the
relative stereochemistry of the amine and the ester function-
alities had been incorrect.
Thus starting from the commercially available penta-1,4-
dien-3-ol, we have synthesised the desired A/E/F tricycle 2 in 14
steps and 5% overall yield. This ester models the A/E/F tricyclic
system of methyllycaconitine and contains 6 stereogenic
centres. Biological testing of this compound is currently
underway.
1
4
then reduced with sodium borohydride. Oxidation of the
1
5
alcohol with TPAP gave the aldehyde 8, which was heated to
reflux with N-ethylhydroxylamine in benzene to provide the
isoxazolidine 9. All attempts to cleave the N᎐O bond reduc-
tively were, to our surprise, unsuccessful. Thus we deployed a
two-step procedure in which the isoxazolidine was oxidised to
the N-oxide with MCPBA; a spontaneous elimination/ring-
opening resulted in formation of the nitrone 10. This was
reduced to the amine by catalytic hydrogenation over platinum.
Subsequent reactions showed that the stereochemical integrity
of the C᎐N bond was retained throughout this sequence. Ring-
closure to the amide 11 was effected by heating to reflux in
xylene for 24 h. O-Methylation of the alcohol functionality was
achieved by refluxing with methyl iodide in the presence of
silver() oxide: performing the O-methylation step before the
reduction of the amide circumvented problems which arose
in the model system in which the quaternary amine salt was
Acknowledgements
We thank Dr W.-S. Li for X-ray crystallographic analysis, the
EPSRC and Shell Research Ltd. for a CASE award to LCB and
the University of Nottingham for a Fellowship to JAS.
References
1 M. S. Yusunov, Nat. Prod. Rep., 1991, 8, 499.
2
3
4
S. W. Pelletier, N. V. Mody, K. Varughese, J. A. Maddry and H. K.
Desai, J. Am. Chem. Soc., 1981, 103, 6536.
A. Drasdo, M. Caulfield, D. Bertrand, S. Bertrand and S.
Wonnacott, Mol. Cell. Neurosci., 1992, 3, 237.
S. Wonnacott, E. X. Albuquerque and D. Bertrand, Methods
in Neuroscience, 1993, 12, 263.
1
0
formed as a significant side product. Reduction of the amide
5 S. P. Arneric, J. P. Sullivan, M. W. Decker, J. D. Brioni, A. W.
Bannon, C. A. Briggs, D. Donnelly-Roberts, R. J. Radek, K. C.
Marsh, J. Kyncl, M. Williams and J. J. Buccafusco, Alzheimer Dis.
Assoc. Disord., 1995, 9, Suppl. 2, 50.
to the amine 12, with concomitant removal of the acetyl pro-
tecting group was effected with LiAlH . Reaction of the alcohol
4
functionality with isatoic anhydride gave the anthranilate. The
succinimido ring was incorporated by refluxing this intermedi-
ate with (S)-methyl succinic anhydride in xylene to give the
6
E. Palma, S. Bertrand, T. Binzoni and D. Bertrand, J. Physiol., 1996,
91, 151.
4
7 (a) P. B. S. Clarke, Trends Pharmacol. Sci., 1992, 13, 407; (b) S. P. H.
Alexander and J. A. Peters, Trends Pharmacol. Sci., Receptor and
Ion Channel Nomenclature Supplement, 1997, 5.
desired A/E/F tricyclic fragment
diastereomers.‡
2
as
a
mixture of
8
J. M. Jacyno, J. S. Harwood, N. H. Lin, J. E. Campbell, J. P. Sullivan
and M. W. Holland, J. Nat. Prod., 1996, 59, 709.
The stereochemistry generated by the Diels–Alder reaction
and the 1,3-dipolar addition was confirmed by X-ray crystallo-
9
G. D. Manners, K. E. Panter and S. W. Pelletier, J. Nat. Prod., 1995,
5
8, 863.
1
1
0 L. C. Baillie, J. R. Bearder and D. A. Whiting, J. Chem. Soc., Chem.
Commun., 1994, 2487.
1 W. M. Roush, H. R. Gillis and A. I. Ko, J. Am. Chem. Soc., 1982,
†
To a solution of (E)-hepta-4,6-dienoic acid 5 (9.86 g, 78 mmol) in
water (30 ml) at 25 ЊC was added sodium hydrogen carbonate (6.57 g, 78
mmol) portionwise over 30 min. After gas evolution had subsided, ethyl
2
1
04, 2269.
-(hydroxymethyl)acrylate 6 (20.34 g, 156 mmol) and lithium chloride
1
1
1
2 J. Villieras and M. Rambaud, Synthesis, 1982, 924.
3 R. Breslow and U. Maitra, Tetrahedron Lett., 1984, 25, 1239.
4 M. Rodriguez, M. Llinares, S. Doulit, A. Heitz and J. Martinez,
Tetrahedron Lett., 1991, 32, 923.
(6.14 g, 145 mmol) were added. The mixture was warmed to 60 ЊC and
stirred for 120 h. The mixture was extracted with diethyl ether to remove
unreacted methacrylate, then acidified to pH3 with 2 aqueous hydro-
chloric acid and extracted into diethyl ether. The extracts were washed
with brine, dried and concentrated. The product was subjected to col-
umn chromatography (light petrol–ethyl acetate, 7:3) to yield the
desired compound as a colourless oil (18.16 g, 91%) (endo:exo 3.3:1, as
determined by NMR spectroscopy).
1
5 S. V. Ley, J. Norman, W. P. Griffith and S. P. Marsden, Synthesis,
1
994, 639.
Paper 7/04447J
Received 24th June 1997
Accepted 17th July 1997
‡
All new compounds gave satisfactory spectroscopic and analytical
data.
2
688
J. Chem. Soc., Perkin Trans. 1, 1997