The details of a total synthesis of (+)-minovine (3), patterned upon biogenetic considerations, are discussed.n 1962 Wenkert4 postulated that a protonated form I of dihydropyridine acrylic ester 2a might play an integral role in the biosynthesis of Aspidosperma and Zboga alkaloids. In recent years, in citro experiments have been reported with alkaloids of these classes in an attempt to verify the intermediacy of the acrylic esters 2b.j During this period secodine ( l a ) and 16,17dihydrosecodin-17-01 (lb) have been isolated6 and syn-the~ized.",~ Moreover, secodine has been shown to incorporate in Give into the Aspidosperma n u c l e~s .~~~ It was our intention to investigate the viability of the addition of an endocyclic enamine to a 2,2'-indole acrylic ester followed by a Mannich c10sure.~ Although the ultimate test would have been the construction of acrylic ester 2b or 2c, prudence dictated that the manipulation of such a reactive functionality would be best accommodated by isolating the reactive moieties in separate substrates and only at the designated time to allow their chemical union.Whereas an appropriate endocyclic enamine 412 was already at hand from previous work, it was necessary to consider a suitable means of constructing a (1) For a preliminary account of this work, see A considerable amount of controversy has arisen over these results. We leave the reader to formulate his own opinions.(6) (a) G.2,2'-indole acrylic ester. Since it was known13 that N-methylindole is capable of undergoing lithiation at the 2 position, it was an easily accessible method of appropriate functionalization of the indole nucleus which mandated that minovine (3) be a convenient test of the biogenetic-like sequence. la. R = CH. b. R = CH-OH H CO-CH , A s p id o s p e r m a (tabersonine) b.14,15 dehydro; Z = H, 2a.14,15 dehydro: Z = 0 c.14,15 dihydro; Z = H2
3I b o g a ( c a t h a r a n t h i n e ) When ethyl oxalate was added dropwise to an ethereal solution of the lithioindole, only trace amounts of the desired ethyl 1-methyl-2-indole glyoxylic ester 5a could be detected, the two major products being the diketone 6 and the hydroxy ketone 7. Since it was apparent that the glyoxylic ester was undergoing further reactions, the situation was ameliorated by the slow addition of the lithium reagent to an excess of ethyl oxalate in ether. The major product from the inverse addition was the glyoxylate 5a which was directly converted to its crystalline free acid 5b by saponification of the reaction mixture, which served to allow facile chemical separation from both neutral species 6 and 7. Treatment of the glyoxylic acid with ethereal diazomethane provided the methyl ester 5c as a yellow oil whose nuclear magnetic resonance spectrum revealed two three-proton singlets at 6 3.95 and 3.99 (NCHB and (13)