|Home | About | Journals | Submit | Contact Us | Français|
We report a concise, enantioselective, and highly efficient synthesis of the marine actinomycete-derived natural product saliniketal B. Our approach was motivated with an eye on future structure-function studies of this inhibitor of phorbol ester-mediated ornithine decarboxylase induction via an unknown mechanism. Our strategy highlights the utility of Pt(II)-mediated cycloisomerization of alkynediols developed in our laboratory to construct the dioxabicyclo[3.2.1]octane ring system, a highly selective aldol fragment coupling of which the stereochemical outcome is influenced by a γ-stereogenic methyl group, and an interesting one-pot desilylation/dihydropyranone fragmentation/amidation sequence. As such, saliniketal B was obtained in 11 steps and 23% overall yield from commercially available starting material via a convergent coupling of two equally complex fragments assembled in 7 and 8 steps (39% and 45%) respectively.
Due to the groundbreaking work of Fenical and coworkers, obligate marine actinomycetes were shown to be a rich source of novel bioactive natural products.1 In 2007, they reported the isolation of the polyketides saliniketals A (1) and B (2) from the marine actinomycete Salinispora arenicola,2 the structure of which was confirmed by a total synthesis of Paterson and coworkers.3 Besides unusual structural features, including a dioxabicyclo[3.2.1]octane ring system, an E,Z-dienamide unit reminiscent of the ansa chain of rifamycin, and nine stereocenters (eight of which are contiguous), saliniketals are of biological interest due to their ability to inhibit ornithine decarboxylase (ODC) induction. The first enzyme in the polyamine biosynthesis pathway and direct transcriptional target of the oncogene MYC, ODC was shown to be a potential target for chemotherapeutic or chemopreventive intervention.4 Unlike α-DFMO, saliniketals do not inhibit ODC enzyme activity, but attenuate tumor promoter-mediated induction of ODC.2 Herein, we report a concise and flexible synthesis of saliniketal B (2) featuring a strategy aimed at enabling future structure-function and mode-of-action studies.
Our synthetic strategy is based on a convergent aldol coupling of fragments 3 and 4 following an anti-selective reduction of β-hydroxyketone 2 (Scheme 1). We envisioned a late stage installation of the E,Z-dienamide via an interesting but rarely utilized fragmentation of a dihydropyranone.5 The 2,8-dioxabicylo[3.2.1]octane moiety will be assembled via cycloisomerization of alkynediol 5 exploiting methodology developed in our laboratory.6
An efficient synthesis of coupling partners 3 and 4 is depicted in Scheme 2. Reagent-controlled aldol reaction of the stannyl enolate derived from known oxazolidinone 87 with aldehyde 7, obtained from oxidation of commercially available alkynol 6 (92% yield), provided aldol product 9 in 82% yield and >16:1 dr. Anti-selective reduction with Na(OAc)3BH (>20:1 dr)7a followed by desilylation (74% yield, 2 steps) set the stage for a cycloisomerization of alkynediol 5. Using 5 mol% of Zeise’s dimer,6 2,8-dioxabicyclo[3.2.1]octane 10 was obtained in quantitative yield and processed to ketone 3 via Weinreb amide formation and Grignard reaction with EtMgBr (87%, 2 steps).8
Dihydropyranone 4 was synthesized from ester 11 ($). According to a sequence by Nicolaou and coworkers, p-methoxybenzyl ether formation (87%) was followed by semi-reduction to aldehyde 12 (93%) and allylation with Brown’s reagent (90%).9 The resulting syn-homoallylic alcohol 139,10 was esterified with acid 14, a material prepared from methyl acrylate via Baylis-Hillman reaction,11 silylation, and saponification (73%, 3 steps). Dihydropyranone formation (→ 15) was accomplished via ring-closing metathesis with Grubbs’ second generation catalyst under high dilution conditions in 67% yield (15% starting material was recovered).12 Final oxidative deprotection (DDQ, 91%) and oxidation with Dess-Martin periodinane13 (quantitative) delivered aldehyde 4 in seven steps and 36–39% overall yield.
The final aldol coupling between ethyl ketone 3 and aldehyde 4 yielded the anti-Felkin adduct 2 with high selectivity (>11:1 dr) in 81% yield (Scheme 3). The stereochemical outcome of this reaction deserves some comment. The Z(O)-lithium enolates of syn α-Me, β-alkoxy-substituted ethyl ketones typically yield the 1,3-syn-1,4-syn aldol adducts,14 a situation that is mismatched with the inherent anti-Felkin bias of aldehyde 4.10 We surmise that the observed high selectivity for our reaction can be attributed to the presence of the additional γ-Me stereocenter. As shown in eq 1, the Si-enolate face is normally exposed via conformation A minimizing A1,3-strain in the TTS, whereas the additional γ-Me group disfavors this conformation (A′, eq 2) due to unfavorable syn-pentane interactions. This exposes the enolate Re-face via B′ for a matched reaction with aldehyde 4.15 Next, reduction of β-hydroxy ketone 2 delivered anti-diol 16 (89%, >20:1 dr).16,17 Finally, combining a fluoride-mediated desilylation and concomitant fragmentation5d of dihydropyranone 16 with an in situ amidation of the liberated carboxylic acid provided saliniketal B (2) in 72% for this one-pot operation.18
In summary, we have achieved a short, highly efficient synthesis of saliniketal B (2) in 11 steps (longest linear) and 23% overall yield. Our approach features the utility of our Pt(II)-catalyzed cycloisomerization methodology for the construction of the dioxabicyclo[3.2.1]octane core, a stereoselective aldol coupling of which the selectivity was positively influenced by the ketone γ-stereocenter, and an unusual one-pot desilylation/dihydropyranone fragmentation/amidation sequence.
Financial support for was provided by the NIH (CA90349) and the Robert A. Welch Foundation. We thank Christopher F. Bender for helpful suggestions.