Total Synthesis of Terpenoids Employing a “Benzannulation of Carvone” Strategy: Synthesis of (−)-Crotogoudin

Abstract

Carvone is a sustainable and readily available starting material for organic synthesis. Herein, we present the syntheses of various natural product scaffolds that rely on a novel benzannulation involving the α-methyl group (C-10) of carvone to afford a versatile tetralin. The utility of our synthetic approach is highlighted by its application to a short synthesis of the ent-3,4-seco-atisane diterpenoid (−)-crotogoudin. The 13-step enantiospecific synthesis features a regioselective double oxidative dearomatization, a Diels—Alder cycloaddition with ethylene gas (to construct the bicyclo[2.2.2]octane framework), and a final acid-mediated lactonization. The versatility of this benzannulation strategy is demonstrated by its utility in the preparation of the carbon skeleton of ent-3,4-seco-abietane diterpenoids using an intramolecular oxidative dearomatization.

A key aspiration in pursuing total syntheses of complex molecules in the modern era is to maximize sustainable practices.¹ Designing highly efficient synthesis strategies, as well as powerful methods to implement them, is paramount to realizing this objective.² In addition to considerations of strategies and methods, the choice of readily available and sustainable starting materials contributes substantially to achieving the goals of a modern synthesis. In this context, the pool of chiral compounds including amino acids,³ sugars,⁴ and terpenes⁵ (the “chiral pool”)⁶ has served admirably as starting materials for many practical and inspirational total syntheses over the past century. With regard to the total synthesis of terpenoid natural products,⁷ carvone has been a frequently employed starting material due to its ready availability in both enantiomeric forms, as well as the potential for the orthogonal derivatization of its functional groups.⁸

Despite the wealth of reactivity that has been established for the α-methyl (C-10), isopropenyl, and enone (i.e., double bond and carbonyl) groups of carvone (Figure 1), we recognized that direct C—C bond formation involving the α-methyl group has been underexplored. Direct C—C bond formation to this methyl substituent holds significant potential in the context of natural product synthesis. Specifically, we envisioned that if benzannulation of the carvone six-membered ring could be achieved by engaging the C-10-methyl and enone carbonyl groups, the stage could be set to access myriad natural product classes. In particular, numerous natural product scaffolds could arise from benzannulation following sequential diastereoselective functionalization α to the enone carbonyl group (i.e., at C-6) of carvone.⁹ For example, 3,4-seco-atisane natural products¹⁰ such as agallochaol C^10a (Figure 1A) could be accessed from (S)-carvone whereas 3,4-seco-abietanes including seco-hinokiol¹¹ or callicarpic acid A¹² (Figure 1B) could arise from (R)-carvone.

Figure 1.

Benzannulation of carvone: a unified approach toward terpenoids.

In this Communication, we report our initial investigations into developing this potentially unifying strategy, which has afforded the frameworks of several terpenoid secondary metabolites via short diastereoselective sequences. The virtues of this approach are borne out in a short, enantiospecific total synthesis of the ent-3,4-seco-atisane diterpenoid (−)-crotogoudin (1)¹³ in 13 steps from (S)-carvone.

We commenced our studies with the preparation of benzo-fused bicycle 6 (Scheme 1A), bearing allyl and methyl groups at C-6 (carvone numbering). The methyl group is resident in many of the natural products that could arise from this benzannulated intermediate, whereas the choice of the allyl substituent was dictated by its facile introduction as well as its versatility for subsequent derivatizations. Following a well-established sequence, known carvone derivative 2 was easily prepared through a sequential methylation/allylation protocol.^9c Conjugate reduction using l-Selectride followed by oxidative workup affords the corresponding ketone,¹⁴ which is converted to vinyl triflate 3 upon deprotonation and treatment of the resulting enolate with Comins’ reagent.^15–16 Heck reaction of 3 with ethyl acrylate as the cross-coupling partner yields an ethyl enoate (4), which upon saponification provides acid 5, the substrate for benzannulation.

Scheme 1.

Synthesis of Hexadienoic Acid 5 and Initial Exploration of the Proposed Benzannulation

We anticipated that benzannulation would be achieved by conversion of carboxylic acid 5 to the corresponding ketene¹⁷ (9, Scheme 1B) by ε-deprotonation in mixed anhydride intermediate 8.¹⁸ In turn, 6π electrocyclization of 9, aromatization, and acylation of the resulting phenol would yield 6, consistent with the precedent of Murali and Krishna Rao.¹⁹ Several conditions (A–C), as outlined in Scheme 1A, were explored to effect the benzannulation. Using the conditions reported by Murali and Krishna Rao (Condition A), only a 13% yield of 6a was isolated from a complex reaction mixture.²⁰ A switch to propionic anhydride as the solvent, which could be heated to 160 °C, led to a substantial increase in yield to 42% and the isolation of desired bicycle 6b and, surprisingly, constitutional isomer 7b in a 1:1 ratio. A Cope rearrangement²¹ of 8 prior to ketene formation and electrocyclization likely explains the genesis of 7b through conformer 8′. Full conversion of starting material 5 was achieved by heating the reaction mixture to 180 °C for 5 days, resulting in a combined yield of 59% of 6b and 7b (1:1.4 ratio).

In order to obviate the competing Cope rearrangement and with an eye toward application of the benzannulated bicycle to the synthesis of the diterpenoids illustrated in Figure 1, the allyl group of ester 4 was converted to an n-propyl hydroxy group (Scheme 2). This was achieved by chemoselective hydroboration of the allyl group in the presence of the isopropenyl group using Wilkinson’s catalyst (1 mol % loading) and catecholborane followed by oxidation of the resulting alkylborane.^22,23 Saponification of the intermediate hydroxyester gave acid 10 in 83% yield over two steps. Benzannulated bispropionate bicycle 11 was formed in 82% yield upon heating 10 in propionic anhydride to 180 °C for 5 days.

Scheme 2.

Completion of the Synthesis of Crotogoudin via a Double Oxidative Dearomatization Strategy

Following procedures adapted from Kunesch and Kondo,²⁴ the phenyl propionate in 11 was selectively cleaved using tetramethylguanidine. This set the stage for a position selective oxidative dearomatization to afford dienone 12 (along with the corresponding para-quinol ether and the isomeric masked ortho-benzoquinone as side products in 11–13% yield, respectively).^25,23 The observed selectivity in this iodine(III)-mediated oxidative dearomatization is rather unusual and has, to the best of our knowledge, only been reported by Mal and co-workers on simpler substrates.²⁶ Inspired by Fukuyama’s recent synthesis of (−)-lepenine,²⁷ a diastereoselective [4 + 2] cycloaddition of cyclohexadienone 12 with ethylene was envisioned. However, in accordance with investigations by Liu and co-workers, compound 12 did not readily undergo the desired Diels—Alder reaction.²⁸ Cycloaddition only proceeded under pressure and at elevated temperature (70 bar, 140 °C, 5 d) to afford tricycle 13 in 90% yield (6:1 d.r.).²⁹ At this stage, Wittig olefination of the ketone group followed by acid treatment removed both the propionyl group and cleaved the dimethyl ketal. The resulting primary hydroxyl was oxidized to the carboxyl group to provide seco-crotogoudin (14) in 75% yield over two steps.^30,31 Lactonization of 14 to afford (−)-crotogoudin was fraught with complicating side reactions.²³ Ultimately, conditions were identified that provided crotogoudin (1) in 16% yield (2.9% total yield over 13 steps), along with rearranged lactone 15 in 14% yield.³² Current efforts are directed at identifying conditions that provide 1 more selectively and in higher yield.³³ Crotogoudin prepared using the strategy outlined here provided spectral and analytical data consistent with those obtained during its previous syntheses by Carreira^13c and Liu^13b as well as from its isolation by Dumontet and Rasoanaivo from croton goudotii.^13a

Notably, our synthesis plan affords opportunities to access other diterpenoid secondary metabolites including the atisane and abietane frameworks outlined in Figure 1. For example, ester cleavage of bispropionate bicycle 11 (Scheme 3) and subsequent intramolecular oxidative dearomatization³⁴ of the intermediate phenol (not shown) provided dienone 16. Selective reduction of the less substituted double bond of the cyclohexadienone moiety of 16 to yield α,β-unsaturated ketone 17 was achieved using a combination of MAD³⁵ and l-Selectride.³⁶ A 1,2-addition of an isopropyl group using Knochel’s method³⁷ readily delivered allylic alcohol 18. The direct treatment of this tertiary alcohol with a proton source results in elimination to key intermediates (19 and 20) for the synthesis of seco-abietane congeners such as 9-hydroxycallicarpic acid A and seco-hinokiol, respectively.

Scheme 3.

Synthesis of Secondary Metabolite Congeners via an Intramolecular Oxidative Dearomatization Pathway

In conclusion, a novel strategy for the synthesis of diterpenoids using carvone as a starting material has been developed. Several key transformations led to the success of this approach. These include (1) a benzannulation sequence that employs propionic anhydride and (2) a site-selective double oxidative dearomatization reaction that sets the stage for (3) a highly diastereoselective cycloaddition of ethylene to forge the key [2.2.2] bicycle. Our approach has led to an enantiospecific 13-step synthesis of the diterpenoid (−)-crotogoudin and provided a platform for the synthesis of other terpenoids. The application of this plan to the syntheses of other natural products is the subject of ongoing studies in our laboratory.