A stereoselective synthesis of the tricyclic core in racemic form of the marine alkaloid Lepadiformine is described from 4-methoxy-3-pyrrolin-2-one (methyl tetramate). Key steps involve 5,5-dialkylation of the tetramate, metathesis closure to an A/C 1-azaspirocycle and stereoselective hydrogenation for the trans A/B 1-azadecalin system. Considerable interest has been shown recently in the synthesis of the marine alkaloid Lepadiformine 1, a tricyclic perhydropyrrolo[2,1-j]quinoline isolated 1 from the tunicate Clavelina lepadiformis and reported to have moderate in vitro cytotoxic activity against certain tumour cell lines, 1 as well as various cardiovascular effects in vivo and in vitro. 2 The early syntheses 3 utilised cycloaddition methodology for construction of the pivotal tertiaryaza stereogenic centre of the trans-1-azadecalin A/B ring junction. More recently, Weinreb has demonstrated the application of an intramolecular spirocyclisation of an Nacyliminium ion with an allylsilane to accomplish the first enantioselective synthesis 4 of the alkaloid, thus establishing its absolute configuration. A key intermediate in Weinreb's synthesis was a 1-azaspiro[4.5]decyl unit used as an A/C ring-system template. Recent publications on the construction of the natural products FR901483 5 and (-)-TAN1251A 6 have highlighted the use of substituted 1-azaspiro[4.5]decanes, and in this communication we disclose a novel strategy for accessing this ring system as a 6-ketone 7 with subsequent conversion of it in moderate stereoselectivity to the tricyclic core of Lepadiformine in racemic form (Figure 1).Our strategy centred around a sequence involving 5,5-dialkylation of a tetramate derivative, followed by chain extension and metathesis to the substituted azaspirocycle. 5-Alkylation of the lithium dienolate of N-protected tetramates with primary alkylating agents is well established in the literature, 8 however in our hands allylation of 1-benzyl-4-methoxy-3-pyrrolin-2-one produced significant quantities of the a-alkylation product. Thus we resorted to using a procedure reported by Jones 8a involving allylation of the dianion of 4-methoxy-3-pyrrolin-2-one 2 with one equivalent of allyl bromide. 5-Allyltetramate 3a was obtained in 79% isolated yield after column chromatography. Phase transfer 9 protection of the lactam nitrogen of 3a produced the N-protected derivative 3b in excellent yield setting the stage for construction of the pivotal tertiaryaza centre (Scheme 1).
Scheme 1 Reagents and conditions: (a) n-BuLi (2 equiv), THF, -78°C, allyl bromide (1 equiv), 79%; (b) KOH, Bu 4 NHSO 4 , THF, 88%.Given the poor regioselectivity of alkylation of the tetramate lithium dienolate, we opted to investigate the use of the much softer silyl dienol ether (2-silyloxypyrrole). Although N-Boc-2-tert-butyldimethylsilyloxypyrrole from N-Boc-3-pyrrolin-2-one has been demonstrated by Casiraghi 10 to be a versatile building block in natural product synthesis for 5-regioselective alkylation reactions, no equivalent studies have been carried out...