Oxidative Entry into the Illicium Sesquiterpenes: Enantiospecific Synthesis of (+)-Pseudoanisatin

Abstract

Illicium sesquiterpenes have been the subject of numerous synthetic efforts due to their ornate and highly oxidized structures as well as significant biological activities. Herein we report the first chemical synthesis of (+)-pseudoanisatin from the abundant feedstock chemical cedrol (~$50 USD/kg) in 12-steps using extensive site-selective C(sp³)–H bond functionalization. Significantly, this work represents a novel oxidative strategic template for future approaches to these natural products and their analogs.

Graphical abstract

Since the report of pseudoanisatin (1) and anisatin (2) in 1968 by Yamada and coworkers from Japanese star anise (Illicium anisatum), more than fifty sesquiterpenes containing the seco-prezizaane skeleton have been isolated from plants of the Illicium genus.^1,2 For decades, these natural products have attracted significant synthetic attention owing to their diverse oxidation patterns and notable biological activities (Figure 1). Pseudoanisatin (1) and its congeners contain a bridging ε-lactone, while anisatin (2) and associated compounds present an unusual spiro β-lactone motif. Other members present elaborate variations on this theme; most notably, hydroxylation is found at every potentially oxidizable position on the 15-carbon framework. Each of these structures is also marked by unique and varied phenotypic outputs, although such effects appear to stem from modulation of the γ-aminobutyric acid (GABA) receptors.^2,3 For instance, anisatin (2) is one of the most powerful poisons of plant origin (murine LD₅₀ < 1 mg/kg),⁴ while pseudoanisatin (1) is non-toxic and shows marked selectivity for insect GABA receptors (IC₅₀^fly = 376 nm, IC₅₀^rat > 10⁴ nm) implying potential as an insecticide.⁵ Related jiadifenolide (4) and jiadifenin (5) promote neurite outgrowth at low concentrations,^3c while 1 and 2 are not reported to have this ability.²

Figure 1.

Illicium sesquiterpenes possessing the seco-prezizaane framework.

Many highly creative syntheses have been disclosed for various seco-prezizaane Illicium family members, with anisatin (2),⁶ jiadifenolide (4),⁷ and jiadifenin (5)⁸ attracting the most synthetic attention to date.⁹ The efficiency with which these compounds can be prepared synthetically, however, varies greatly according to structural type and oxidation pattern, with syntheses ranging from a remarkable 8 steps to 4,^7e 22 steps to 3,^6c and over 40 steps to 2.^6a,b Given that the hydroxylation pattern of the Illicium sesquiterpenes also appears to govern their biological effects, we sought a synthetic strategy that would be flexible in its introduction of such units and viewed 1 as an entry point to this family. To accomplish this objective, we drew upon the proposed biosynthesis of these compounds as inspiration.

The biosynthesis of seco-prezizaane sesquiterpenes has not been fully elucidated biochemically, but has been proposed to commence with cyclization of farnesyl pyrophosphate to the bisabolyl cation (6) (Figure 2).² Through a series of hydride shifts and cationic cyclizations, the polycyclic cedrane skeleton (7) is rapidly forged. A 1,2-alkyl shift converts 7 into the allo-cedrane framework (8) which undergoes C–C bond cleavage furnishing the core seco-prezizaane ring system (9). Individual members are then tailored by extensive C-H hydroxylation. In contemplating a semi-synthetic approach to this family, a strategy which has had a profound impact on the synthesis of steroids,¹⁰ we were dismayed to find that no readily abundant starting materials with the skeleton of 9 are available. The sesquiterpene (+)-cedrol (10), however, which bears the ring system of 7, is an exceedingly inexpensive terpene feedstock extracted from Texas cedarwood ($52 USD/kg). Our goal therefore became the conversion of this abundant and unfunctionalize terpene into the rare Illicium sesquiterpenes. Such an endeavor entails not only the exploration of extensive C(sp³)–H bond functionalization,^11,12 but also carbon-carbon bond reorganization. Herein we document the realization of this blueprint resulting in the first chemical synthesis of pseudoanisatin (1) in 12 steps from cedrol.

Figure 2.

Proposed biosynthetic pathway to seco-prezizaane Illicium sesquiterpenes inspires a semisynthetic approach.

We began our synthetic studies by first oxidizing the gem-dimethyl group of cedrol, a task easily accomplished with Suarez’s radical-based method (I₂/PhI(OAc)₂, floodlamp) on 50–100 mmol scales.¹³ The strained tetrahydrofuran ring subsequently formed could be methylated and eliminated via the action of Meerwein’s salt (Me₃OBF₄) and mild base (proton sponge) affording methoxy cedrene 11 in near quantitative yield (10-gram scale). The trisubstituted olefin of 11 was oxidatively cleaved using in situ generated RuO₄ (via NaIO₄/cat. RuCl₃) and afforded keto-acid 12 in good yield on decagram scales. Significant difficulties were quickly encountered in attempting to elicit an α-oxidation of the methyl ketone at the hindered internal position. Canonical strategies such as thermodynamic enolate formation and subsequent oxidation all failed. Ultimately it was found that simply heating keto-acid 12 to 150 °C in the presence of anhydrous CuBr₂ (3.0 equiv) and t-BuOH (3.0 equiv) cleanly forged lactone 13 via direct, intramolecular acyloxylation.¹⁴

With lactone 13 in hand, we were poised to investigate methods to convert the 5,5-fused ring system of the cedranes into the 5,6-fused seco-prezizaane skeleton and the venerable α-ketol rearrangement appeared to be a logical choice.¹⁵ After extensive experimentation we found that treating 13 with KOH/KOt-Bu in DMSO initiated lactone hydrolysis, presumably forming potassium alkoxide 14, which then underwent an α-ketol rearrangement to generate 15 as a 4:1 mixture of diastereomers in 45% combined yield. This reaction could be easily scaled to the gram level, and the stereochemistry of the major product–which bears the correct stereochemistry for all seco-prezizaane Illicium sesquiterpenes–was confirmed by X-ray crystallographic analysis. Although many conditions–both acidic and basic–have been reported to promote α-ketol rearrangements, we found the potassium counterion to be crucial in our system. Furthermore, subjecting either pure diastereomer of 15 to the reaction conditions established the same 4:1 mixture implying the observed diastereoselectivity is a result of thermodynamic control.

Moving forward, the tertiary alcohol of 15 was silylated (NaH, TBSCl) affording compound 16 and setting the stage for a crucial C-H activation reaction of the unactivated tertiary methine center. Considering quaternary carbons on both sides flank its position, we reasoned that a successful oxidation of this position would likely need to be directed by the internal carboxylic acid moiety. Taking inspiration from the general observations of White and others, we reasoned that non-heme iron catalysis could potentially forge a lactone ring directly.^16–18 Nevertheless, with an alkyl ether, silyl ether, and several C-H bonds within reach of the pendant carboxylic acid present, substrate 16 would test the limits of this chemistry. After extensive experimentation, we found that iron catalysts bearing the pinene-based mepp ligand architecture popularized by Costas were able to promote this reaction, resulting in a combined 37% yield of lactone products (see 17–19) along with 18% recovered starting material.^19,20 The structure of the major product (17) was confirmed by X-ray crystallographic analysis of a desilylated derivative (Scheme 1). Loss of the silicon-based protecting group during the reaction was significantly problematic, as the iron catalyst was found to decompose this unprotected material competitively, without promoting the desired, directed C-H oxidation. Nevertheless, both 17 and 18 could be directly utilized in the synthesis without further manipulation, and thus the combined 30% yield of these products proved to be workable in completing exploratory studies.²¹ Lactone 17 was then ethylated and eliminated (Et₃OPF₆, proton sponge) affording alkene 20. Much to our delight, in-situ formed TMSI (TMSCl, NaI) cleanly dealkylated the methyl ether and following treatment with TBAF, compound 21, which contains the ε-lactone characteristic of pseudoanisatin was formed. Dihydroxylation of the trisubstituted alkene in 21 was then attempted and previous work, particularly Kende’s pioneering synthesis of 3, hinted that reagent approach would come from the desired α-face.⁶ However, treatment of 21 with OsO₄ in pyridine afforded exclusively the undesired diastereomer (22) in which the jiadifenolide-type lactone has been reconstructed and the primary hydroxyl group has formed a hemiketal. Fortunately, Donohoe’s procedure for dihydroxylation directed by a homoallylic alcohol (OsO₄•TMEDA), overrode this facial bias resulting in the formation of 23.²² Finally, mesylation of 23 (MsCl/pyridine) followed by treatment with aqueous sodium hydroxide promoted secondary alcohol inversion thus completing a 12-step synthesis of (+)-pseudoanisatin (1) from (+)-cedrol.

Scheme 1.

12-step synthesis of (+)-pseudoanisatin (1) from (+)-cedrol (10). Reagents and conditions: a) PhI(OAc)₂ (3.0 equiv), I₂ (1.0 equiv), cyclohexane, hν (visible), 1.5 h, 73%; b) Me₃OBF₄ (1.5 equiv), proton sponge (1.5 equiv), DCM, 55 °C, 12 h, 97%; c) NaIO₄ (5.0 equiv), RuCl₃•xH₂O (0.1 equiv), CCl₄:MeCN:H₂O (3:3:4), 1 h, 72%; d) CuBr₂ (3.0 equiv), t-BuOH (3.0 equiv), diglyme, 150 ºC, 12 h; e) KOH (1.0 equiv), KOt-Bu (3.0 equiv), DMSO, 14 h, 45% (two steps), d.r. = 4:1; f) NaH (5.0 equiv), TBSCl (4.0 equiv), THF, 65 °C, 8 h then add 3.0 M HCl (16 equiv), 65 °C, 16 h, 88%; g) TBHP (5.0 equiv), [Fe] (0.5 equiv), Tl(OTf) (0.5 equiv), MeCN, 1 h, 20% 17, 11% 18, 6% 19; h) Et₃OPF₆ (3.0 equiv), proton sponge (3.0 equiv), DCE, 85 °C, 12 h then add TFA/H₂O (1:1), rt, 45 min, 66%; i) TMSCl (10.0 equiv), NaI (5.0 equiv), MeCN, 80 °C, 12 h; j) TBAF (5.0 equiv), AcOH (1.0 equiv), THF, 1 h, 64% (two steps); k) OsO₄ (1.5 equiv), pyr., 12 h, 63%; l) OsO₄•TMEDA (1.5 equiv), DCM, −78 °C to rt, 2 h; m) MsCl (10.0 equiv), pyr. (10.0 equiv), DCM, 12 h, then add aq. NaOH (2.0 M), 2 h, 80% (two steps). Proton sponge = 1,8-bis(dimethylamino)naphthalene, TBS = tert-butyldimethylsilyl, [Fe] = [Fe((R)-mepp)(MeCN)₂][(SbF₆)₂], TBHP = tert-butyl hydroperoxide, DCE = 1,2-dichloroethane, TMS = trimethylsilyl, TMEDA = tetramethylethylenediamine.

In summary, biosynthesis-guided synthetic planning coupled with the exploration of selective C(sp³)–H oxidation reactions and a skeletal rearrangement strategy, has allowed for the first chemical synthesis of (+)-pseudoanisatin (1) in 12-steps. The route to 1 from cedrol represents the first fully oxidative strategy toward this coveted sesquiterpene family, a blueprint we envision allowing for both the synthesis of numerous Illicium sesquiterpenes and their heteroatom analogs. Moreover, this work further attests to both the power, as well as current practical limitations, of C-H oxidation pathways in complex synthetic planning.²³ We foresee future efforts in this area leading to an enhanced toolset of compounds for neuroscience as well as improved reagents for oxidative chemistry in complex molecular settings.