Asymmetric Total Synthesis of (−)-Plicatic Acid via a Highly Enantioselective and Diastereoselective Nucleophilic Epoxidation of Acyclic Trisubstitued Olefins

Abstract

The first total synthesis of (−)-plicatic acid has been achieved by a concise and enantioselective route. In this synthesis, a conceptually new strategy featuring an asymmetric epoxidation-intramolecular epoxy-ring-opening Friedel-Crafts reaction sequence was developed for the stereoselective construction of the 2,7’-cyclolignane skeleton bearing contiguous quaternary-quaternary-tertiary stereocenters. The implementation of this strategy was enabled by the development of a modified protocol for the Seebach epoxidation with TADOOH, which affords an unprecedented, highly enantioselective and diastereoselective epoxidation with a range of α-carbonyl-β-substituted acrylates 3.

Plicatic acid has been identified as the causative agent of occupational asthma.^1a–d Furthermore, plicatic acid has been shown to cause inflammatory and allergic reactions, including increased concentrations of immunoglobulins, histamine, leukotrienes, eosinophil and T-cell levels in the blood.1e–g Plicatic acid was isolated in 1959 by MacLean and co-workers from western red cedar (Thuja plicata).^2a The relative and absolute configurations were assigned by X-ray crystallographic analysis and optical rotatory dispersion (ORD) studies, respectively.^2b–c A concise total synthesis of this natural product could establish a means for access to analogues that could be valuable for biomedical studies aiming to elucidate the molecular mechanism underlying the biological activities of plicatic acid. Our interest in the total synthesis of plicatic acid is also motivated by the synthetic challenges imposed by its rather unusual lignan skeleton that is densely functionalized and bears a motif of contiguous quaternary-quaternary-tertiary stereocenters.³

As illustrated in Scheme 1, our retrosynthetic analysis presents a new strategy to create the 2,7’-cyclolignane skeleton in order to achieve a stereoselective construction of the B ring bearing the contiguous quaternary-quaternary-tertiary stereocenters. We envisaged that the key intermediate, α-hydroxy ketone 2a, could be synthesized from olefin E- 3a by an asymmetric epoxidation followed by an intramolecular Friedel-Crafts reaction to open the epoxide ring in 4a. The stereoselective construction of the quaternary center at C(8) is to be accomplished via a C(8’) hydroxy-directed nucleophilic addition to the ketone in 2a.

Scheme 1.

A Retrosynthesis for (−)-Plicatic Acid (1)

The implementation of this synthetic strategy, however, required us to fill a significant gap in the current repertoire of asymmetric epoxidations, namely the lack of a highly diastereoselective and enantioselective nucleophilic epoxidation of trisubstituted electron-deficient olefins such as the α-carbonyl-β-substituted acrylates 3a.⁴ Accordingly, we first focused on the establishment of a highly enantioselective and diastereoselective asymmetric epoxidation of α-carbonyl-β-substituted acrylate 3b. Through considerable experimentation, we discovered that Seebach’s nucleophilic epoxidation with TADOOH as the terminal oxidant⁵ could be promoted by a catalytic amount of LiOH, which, interestingly, proceeded smoothly in THF at 0 °C to transform 3b into epoxide 4b as a single diasteoreomer^4f in 92% ee and 98% yield (4b, Table 1). Significantly, the scope of this modified Seebach epoxidation could be extended to a variety of acrylates (3b–i) bearing either an α-alkoxycarbonyl or an α-ketone group (4b-i, Table 1).

Table 1.

Asymmetric Nucleophilic Epoxidation of Various Acyclic Trisubstituted Olefins

^aSee Supporting Information for details of reaction conditions.

^bIsolated yields are reported.

^cEe’s were determined by HPLC analysis.

We next turned our attention to the application of this new protocol to the asymmetric epoxidation of the olefin intermediate E-3a (Scheme 2). Benzylation of eugenol (5) followed by oxidative cleavage of the olefin afforded aldehyde 7, which was transformed into β-ketoester 8 in 92% yield using Roskamp’s protocol.⁶ The Knoevenagel condensation of 8 and 9 furnished olefin 3a as a 5:3 E/Z mixture, which could be separated by chromatography. The Z-3a was found to isomerize to E-3a in the presence of pyridine in refluxing benzene. Thus, E-3a could be obtained in 80% overall yield from one cycle of Knoevenagel condensation-isomerization. Gratifyingly, the key LiOH-catalyzed asymmetric epoxidation of E3a - with (S,S)-TADOOH generated epoxide 4a in 98% ee and 83% yield. A screening of various Lewis acids revealed TfOH as the optimal catalyst, which effectively promoted the Friedel-Crafts reaction in a loading of 4.0 mol %. Thus, α-hydroxy ketone 2 was obtained as a 4:1 diastereomeric mixture in favor of the desired diastereomer 2a (Scheme 2), which was isolated in 70% yield by silica gel chromatography. Notably, the Friedel-Crafts reaction also proceeded in a highly regioselective manner as the other regioisomer was not detected.

Scheme 2.

Total Synthesis of (−)-Plicatic Acid (1)

^a Conditions: (a) NaH, BnBr, quant. (b) (1) NMO, cat. OsO₄; (2) NaIO₄, quant. (c) N₂CHCOOEt, cat. SnCl₂, 92%. (d) 9, piperidine, PhCOOH, 80% (after one cycle), E/Z = 5:3. (e) (S,S)-TADOOH, cat. LiOH, 83%, 98% ee. (f) TfOH (0.04 eq.), 0 oC to r.t., 70% 2a, 17% 2b. (g) ClSi(Me)₂CH₂Br, imidazole, 75% (94% brsm). (h) SmI₂, NiI₂ (0.1 eq.), 0 oC, 58%. (i) H₂O₂, NaHCO₃, 87% (90% brsm). (j) n-PrSNa, DMF, 97%. (k) H₂, Pd/C, MeOH, then Dowex-50, 72%.

With α-hydroxy ketone 2a in hand, a stereoselective addition of the hydroxymethyl group to the ketone in 2a stood as the final obstacle for the construction of the full carbon skeleton of plicatic acid. Our initial attempts to realize a C8’-OH-directed addition with a metal reagent such as vinylmagnesium bromide,^7a vinyl-lithium,^7b vinylcesium chloride^7c or benzyloxymethyl magnesium chloride^7d were unsuccessful. These basic metal reagents only deprotonated the benzylic proton at C7 in 2a, thereby leading to enolization, rather than nucleophilic addition to the C8-carbonyl group. We then explored an alternative strategy to execute a formal stereospecific addition of a hydroxymethyl group to the ketone with the C-C bond formation implemented under nearly neutral conditions (Scheme 2). Thus, the C8’-OH in 2a was first silylated with ClSi(Me)₂CH₂Br to form 10 in 75% isolated yield (94% yield brsm). To our delight, 10 underwent a SmI₂-mediated, intramolecular Barbier reaction⁸ in the presence of 10 mol % of NiI₂⁹ to afford hydroxysilane 11, which was subjected to a Fleming-Tamao-Kumada oxidation¹⁰ to furnish the triolester 12 in 50% overall yield from 10. However, triolester 12 decomposed rapidly when subjected to hydrolysis by LiOH. On the other hand, upon treatment with the slightly basic sodium propanethiolate, 12 was converted to carboxylate 13 in 97% yield.¹¹ Global debenzylation of 13 followed by cationic exchange delivered synthetic (−)-plicatic acid (1) in 72% yield. Extensive spectroscopic and chromatographic analysis of a 1:1 mixture of synthetic and natural (−)-plicatic acid showed the two to be indistinguishable.¹²

In summary, the first asymmetric total synthesis of (−)-plicatic acid was accomplished in 12 steps and 14% overall yield from eugenol. In this synthesis a conceptually new strategy featuring an asymmetric epoxidation-intramolecular epoxy-ring-opening Friedel-Crafts reaction sequence was developed for the stereoselective construction of a structurally complex 2,7’-cyclolignane skeleton. The implementation of this strategy was enabled by the development of a modified protocol for the Seebach epoxidation with TADOOH, which affords an unprecedented, highly enantioselective and diastereoselective epoxidation with a range of α-carbonyl-β-substituted acrylates 3.