Toward the Total Synthesis of Alpkinidine: Synthesis of Haloquinone CE Ring System Synthons and Attempted Nucleophilic Bisannulation

Abstract

Model chemistry involving the bisannulation of 2,3-dichloro-1,4-naphthoquinone with the ester enolate derived from ethyl o -nitrophenylacetic acid, which rapid assembled the ABCD ring system of a pentacyclic pyrroloacridine, has been applied to the attempted synthesis of the marine natural product alpkinidine. The reaction of ethyl o -nitrophenylacetic acid with 6,7-dichloro-2-methylisoquinoline-1,5,8(2 H )-trione, required to extend the model strategy to alpkinidine, was unfruitful, giving only complex mixtures. Efforts to direct the regiochemistry of the key Michael substitution step using 6-bromo-2-methylisoquinoline-1,5,8(2 H )-trione afforded an adduct sharing the complete carbon skeleton of alpkinidine, but this could not be elaborated to the natural product.

Toward the Total Synthesis of Alpkinidine: Michael Addition to Isoquinolinetrione CE Ring-System Synthons

“…At the outset, attempts to use Michael addition reactions to couple 1,4-naphthoquinone ( 11 ) and o -nitrophenylacetonitrile ( 12 ) had been unsuccessful, producing what appeared to be oligomeric products instead of the expected hydroquinone 13 or quinone 14 adducts ( Scheme 2 ), 1 hence our prior focus on Michael substitution of haloquinones 6 and 8 ( Scheme 1 ). 2 …”

“…In an effort to resurrect the general strategy of connecting the CE- and A-ring systems of potential precursors to alpkinidine, through the reaction of carbanions with quinone electrophiles, 2 the reaction of ethyl o -nitrophenylacetate ( 2 ) with isoquinolinetrione 15 2 was investigated and indeed provided a hydroquinone adduct in modest yield ( Scheme 3 ). Unfortunately, heteronuclear multiple bond correlation (HMBC) spectroscopy revealed the product to be the undesired regioisomer 16 .…”

“…Serendipity played a role in allowing us to explore the general strategy set out in Scheme 5 . During an attempt to oxidatively demethylate 35 , 2 Ag 2 O was mistakenly used instead of AgO ( Scheme 6 ). Unsurprisingly, no quinone 15 was isolated from this reaction, but the electron-rich ring system was sufficiently reactive to undergo nitration with the nitric acid.…”

“…Amidation of 2,5-diacetoxybenzoic acid ( 20 ) with commercial aminoacetal 21 gave 22 , along with the partially deacetylated product 23 (Scheme ). The mixture was treated with concentrated H 2 SO 4 but, while the analogous N -methylamide cyclized under these conditions, in this case the isoquinolone 26 could not be detected in the complex mixture of products formed. In an attempt to avoid phenol protecting groups, methyl 2,5-dihydroxybenzoate ( 24 ) was heated with neat amino acetal 21 , providing the amide 25 in acceptable yield.…”

“…We previously described an efficient route to the model pentacyclic pyrroloacridine 5 by Michael substitution of 2,3-dichloronaphthoquinone ( 1 ) with the anion derived from ethyl o -nitrophenylacetate ( 2 ) to give 3 , D-ring formation through nucleophilic lactamization with methylamine, and reductive cyclization of 4 to close the B ring . Despite this precedent, attempts to effect an analogous Michael substitution of dichloroisoquinolinetrione 6 were unsuccessful . Formation of the key carbon–carbon bond ultimately comprising the BD-ring fusion was achieved in the analogous reaction of monobromide 8 ; however, we were not able to progress adduct 9 toward alpkinidine ( 10 ).…”