Part 2: Enantioselective Total Synthesis of (−)-Citrinadin A and Revision of Its Stereochemical Structure
Zhiguo Bian, Christopher C. Marvin, and Stephen F. Martin
So folks, here it is. Took me a few days longer as promised but finally I made it. Though citrinadin A is closey related to citrinadin B the synthetic approach of the Martin group is much different from that of Wood et al.. Their retrosynthetic considerations are summarized in scheme 1. Interestingly the introduction of the epoxyketone utilizes almost the same chemistry under similar reaction conditions. Contrary to the Wood approach the spiro-oxindole is build up by an epoxidation / semipinacol rearrangement in a diastereoselective manner. This disconnection leads back to the bromoindole shown which in turn is introduced by a Fischer indole synthesis. The tertiary alcohol and amine are derived from selective epoxidation / epoxide opening to give a lactam which tracks back to a vinylogous Mannich reaction between 2 and A.
Dimethylcyclohexadione 1 is monoprotected and methoxycarbonylated with dimethylcarbonate. Subsequent triflation and copper-mediated introduction of another methylgroup then leads to ester 2. In the presence of LDA and in situ transmetallation with zinc chloride a vinylogous enolate is formed which reacts with in situ formed pyridinium ion A. After acidic hydrolysis ester 4 is formed which undergoes base mediated lactam formation to give 5.
Next TIPS cleavage sets the stage for the stereoselective introduction of a methyl group which had to be accomplished in a two-step sequence. After cuprate addition derived from PhMe2SiCH2MgCl and reduction of the ketone the silylgroup was removed under harsh conditions to provide alcohol 8. Epoxidation of the unsaturated lactam with peracid and ensuing opening with dimethylamine then leads to 10.
The dioxolane is then directly used in a Fischer indole synthesis with bromophenylhydrazine in aqueous sulfuric acid to give indole 11 in pretty good yield. Successive reduction of the lactam carbonyl was accomplished by combined alane / borohydride reduction which proved to give the best yields of 12. In situ protection of the sensitive amine moieties with PPTS and epoxidation with Davis oxaziridine yields an intermediate indoline which undergoes semipinacol rearrangement in the presence of acetic acid to give the core structure of the citrinadins with complete control of the quaternary carbon centre. Sonogashira coupling then provides alkyne 15.
All that remains was to transform the triple bond into the epoxyketone which was accomplished after amide formation with dimethylvaline utilizing again the Gold mediated oxygenation and subsequent Enders epoxidation protocol (cf. Wood et al.).
Pretty cool synthesis. It was very intriguing to me to see two almost completely different approaches of the Wood and Martin group which were also published back to back and ultimately corrected the proposed structure of the citrinadins.