Asymmetric Total Synthesis of (+)‐Ineleganolide

ABSTRACT

Polycyclic cembrane diterpenoids and norditerpenoids have garnered sustained interest from the synthetic community due to their unique polycyclic frameworks and promising biological activities. Herein, we report a 13‐step total synthesis of (+)‐ineleganolide, a highly oxidized cembrane norditerpenoid possessing a synthetically challenging [6,7,5,5,5] cage‐type pentacyclic scaffold. This synthesis is highlighted by a novel strategy relying on an intramolecular Diels–Alder reaction that stereoselectively constructs the pivotal [6,6,5,5,5] pentacyclic framework as a single diastereoisomer, addressing stereochemical control challenges. In the late‐stage elaboration, a tandem epoxidation/Meinwald rearrangement efficiently forges the sterically congested central seven‐membered ring. This synthetic strategy provides a new blueprint for accessing other members of the cembrane norditerpenoid family.

The structurally complex cembrane norditerpenoid (+)‐ineleganolide has been synthesized by a route featuring an intramolecular Diels–Alder reaction for the stereoselective construction of the intricate [6,6,5,5,5] pentacyclic skeleton and a late‐stage tandem epoxidation/Meinwald rearrangement to forge the congested seven‐membered ring. This strategy provides a new blueprint for the synthesis of cembrane norditerpenoids.

Polycyclic cembrane diterpenoids and norditerpenoids are widely distributed across diverse marine organisms, with soft corals as the primary source [1, 2]. Endowed with unique, structurally intricate molecular architectures and prominent pharmacological activities (Figure 1) [3], this class of marine natural products has garnered sustained and intense interest from synthetic chemists over the past three decades, emerging as a research hotspot in synthetic chemistry. To date, numerous research groups have devoted considerable efforts to the total synthesis of polycyclic cembrane diterpenoid and norditerpenoid natural products [4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25]. This structurally diverse family encompasses several distinct carbon skeletons, predominantly categorized into tricyclic scaffolds of the 6/7/5, 7/6/5, and other types. Representative members include ineleganolide (1), scabrolide B (2), horiolide (3), yonarolide (4), scabrolide A (5), sinulochmodin C (6), pavidolide B (7), havellockate (8), and rameswaralide (9). Among these, ineleganolide is an early‐isolated representative that has attracted persistent research attention due to its exceptionally complex and unique molecular framework.

FIGURE 1.

Representative cembranoid natural products.

Ineleganolide was first isolated from the soft coral Sinularia inelegans by Duh and co‐workers in 1999 [26]. Pharmacologically, it exhibits potent cytotoxicity against P‐388 murine leukemia cell lines [27, 28, 29]. Structurally, ineleganolide possesses a highly rigid oxidized [6,7,5,5,5] pentacyclic framework, incorporating nine chiral centers, eight of which are contiguous, rendering it a formidable synthetic target. Prior to our work, several synthetic studies toward ineleganolide have been reported, including a biomimetic semisynthesis by the Pattenden group [9], key skeleton construction efforts by the Moeller group [14], convergent total synthesis attempts by the Vanderwal group [15], and decade‐long pursuits by the Stoltz group [16, 17, 18, 19, 20]. Notably, the elegant total synthesis of ineleganolide has only recently been accomplished independently by the Wood [21], Stoltz [22], Fürstner [23], Sarlah [24], and Maimone [25] groups. Inspired by our sustained interest in structurally intricate marine natural products, we herein report the asymmetric total synthesis of (+)‐ineleganolide in 13 steps [30].

In our synthetic design, an innovative synthetic strategy centered on intramolecular Diels–Alder reaction is proposed that enables the simultaneous construction of multiple chiral centers and the precise stereoselective assembly of the core [6,6,5,5,5] pentacyclic scaffold. Taking advantage of the high ring strain inherent to the central cyclohexane unit, sequential epoxidation/Meinwald rearrangement is employed to release strain and achieve ring expansion, thus efficiently affording the target [6,7,5,5,5] pentacyclic skeleton. This protocol addresses the challenge in constructing cycloheptane frameworks by utilizing strain‐driven rearrangement of a cyclohexane core, which simplifies the stereochemical control of multiple chiral centers. As outlined in Scheme 1, retrosynthetic analysis reveals that (+)‐ineleganolide (1) could be accessed from enol ether 10 through a tandem epoxidation/Meinwald rearrangement. Enol ether 10 could be obtained from enol 11 through an olefin migration reaction, which in turn could be derived from lactone 12 by an intramolecular Diels–Alder reaction followed by sequential oxidation steps. Lactone 12 could be prepared from allylic alcohol 13 through ester condensation combined with ring‐closing metathesis (RCM). 13 could be traced back to ketone 14 via Suzuki coupling and subsequent deprotection. Finally, ketone 14 could be synthesized from chiral allylic alcohol 15 through a cascade nucleophilic addition and gold‐catalyzed cyclization.

SCHEME 1.

Retrosynthetic analysis of (+)‐ineleganolide.

As shown in Scheme 2a, our synthetic endeavors commenced with commercially available chiral allylic alcohol 15. Through hydroxy protection and subsequent addition of organozinc reagents, diol 17 was obtained in 53% yield, enabling the installation of the second pivotal stereocenter. After addressing this stereochemical issue, we investigated the gold‐catalyzed cyclization [31]. Notably, in addition to the previously reported influential factors (ligands, anion pairs, protic acids, and nitrogen oxides), the reaction concentration was found to critically impact the yield. Higher concentrations significantly enhanced the reaction efficiency. Ultimately, the target ketone 14 was obtained in 66% yield. Subsequently, we focused on the site‐specific introduction of a cyclohexenyl group at the C1 position via a cross‐coupling reaction. Initially, the carbonyl moiety of the substrate was converted into the corresponding trifluoromethanesulfonate derivative. To modulate the regioselectivity, a series of sterically hindered bases (Sodium bis(trimethylsilyl)amide (NaHMDS), Potassium bis(trimethylsilyl)amide (KHMDS), Lithium bis(trimethylsilyl)amide (LiHMDS), and Lithium diisopropylamide (LDA)) were subjected to systematic evaluation. Among these, NaHMDS proved to be the optimal choice, affording enol ether 19 in 77% yield with exclusive regioselectivity. Utilizing this intermediate as the coupling partner, Suzuki cross‐coupling and desilylation sequence furnished allylic alcohol 13 in high yield. Next, attention was directed toward the construction of the dienophile unit. An ester condensation protocol was employed to achieve efficient side chain incorporation and the formation of acrylate 22. Subsequently, the key lactone 12 was successfully accessed through RCM reaction.

SCHEME 2.

Preparation of 12 and studies of Diels–Alder reaction.

With the key precursor in hand, we explored the feasibility of constructing the [6,6,5,5,5] polycyclic framework via the intramolecular Diels–Alder reaction of lactone 12 and its analogues. To explore the reactivity of this intramolecular Diels–Alder reaction, a series of substrates were designed and subjected to systematic condition screening [32]. Preliminary results demonstrated that the target Diels–Alder adduct 23 could only be obtained from lactone 12; in contrast, lactone analogues 24, 25, and 26 underwent decomposition (Scheme 2b). When 12 was used as starting material, 23 was isolated in 28% yield, accompanied by a substantial amount of decomposition by‐products (Scheme 2b, entry 5). We hypothesized that this could be attributed to substrate oxidation induced by the high reaction temperature. To address this issue, two remedial strategies were examined sequentially: the addition of the antioxidant 2,6‐di‐tert‐butyl‐4‐methylphenol (BHT) and performing the reaction under a nitrogen atmosphere. Nitrogen protection was found to be highly effective, increasing the yield to 42%. Moreover, reaction concentration exhibited a considerable influence on the yield. Further optimization identified 0.01 mol/L as the optimal concentration, which raised the yield to 72%. In accordance with Hammond's postulate for late transition state, pronounced steric congestion and ring strain destabilize the exo transition state, thereby disfavoring the generation of isomer 23’. By contrast, the strain‐minimized endo transition state facilitates efficient assembly of the desired polycyclic skeleton (Scheme 2c.) Notably, residual Grubbs catalyst from the preceding RCM step showed no significant adverse effect on the Diels–Alder reaction. Capitalizing on this observation, the RCM and the key Diels–Alder reaction were integrated into a one‐pot sequence to streamline the synthetic route. To this end, toluene was replaced with p‐xylene as the solvent for the RCM reaction, thereby achieving full substrate conversion. These results validate the feasibility of the one‐pot RCM/Diels–Alder strategy.

Sequential Riley and Ley oxidations were then carried out, affording enone 11 in 70% yield. Initially, we attempted to construct enol ether 10 via olefin migration, followed by a tandem epoxidation/Meinwald rearrangement to access diketone 31 via intermediate 30. However, despite extensive screening of various reaction conditions (see the Supporting Information for details), the targeted enol ether 10 remained inaccessible. This outcome is mainly ascribed to the high acidity of the α‐hydrogens adjacent to the carbonyl group, which is readily deprotonated to generate an enolate intermediate. The enolate intermediate diverts the reaction pathway and precludes the desired olefin migration. Accordingly, the synthetic route was slightly modified, with enol ether 10 redesigned to be constructed through C4‐hydroxylation, C2–C3 double bond reduction, and subsequent dehydration (Scheme 3a). We first targeted the synthesis of hemiacetal 27 via allylic oxidation of enone 11 [33, 34, 35, 36]. Classical oxidation systems, such as SeO₂ and CrO₃, were evaluated but failed to deliver the desired product. Structural analysis revealed that the target reaction site corresponds to the allylic position of a tetrasubstituted olefin, where steric hindrance and intrinsically low reactivity impede the oxidation process. We therefore turned to radical oxidation, a transformation that proceeds via allylic hydroperoxidation followed by base‐mediated elimination to form the carbonyl group [37]. Based on this mechanistic insight, we proposed to trap the intermediate as an alcohol by omitting the base; instead, a reducing agent was added after the reaction to convert the hydroperoxide intermediate into the desired alcohol. This strategic modification delivered hemiacetal 27 in 62% yield. Subsequent tert‐butyldimethylsilyl (TBS) protection of hemiacetal 27 afforded acetal 28 [38], whose structure was unambiguously confirmed by single‐crystal X‐ray diffraction (Scheme 3b).

SCHEME 3.

Total synthesis of (+)‐ineleganolide (1).

The subsequent olefin reduction presented another significant challenge. As a tetrasubstituted alkene, 27 proved resistant to both conventional H₂/Pd‐C hydrogenation and hydrogen atom transfer (HAT) protocols [39, 40]. Notably, we discovered that tellurium (Te) synergized with sodium borohydride [41, 42, 43] efficiently accomplished the reduction of this unreactive olefin, affording hemiacetal 29 with a diastereomeric ratio (dr) of 10:1. This pivotal transformation provided a critical foundation for the subsequent hydroxyl elimination and tandem epoxidation/Meinwald rearrangement. Enol ether 10 was then constructed via mesylation followed by thermally promoted in situ elimination. It was found that 10 exhibited poor stability at room temperature, which limited its handling in subsequent manipulations. Our initial strategy aimed to construct the diketone 31 through oxidation and rearrangement in a single step, employing either epoxidation or dihydroxylation. However, extensive condition screening (see the Supporting Information for details) led to substrate decomposition at ambient temperature, which was primarily attributed to the intrinsic instability of 10. To address this issue, the reaction conditions were optimized by adding a solution of dimethyldioxirane (DMDO) in dichloromethane (DCM) at −78°C, followed by gradual warming of the reaction mixture to 5°C. This protocol successfully furnished the target rearrangement product 31 in 32% yield, with the recovery of substrate 27 in 33% yield [44]. Subsequent Saegusa oxidation of 31 delivered cyclohexenone 32 in 78% yield. Finally, investigations focused on the stereospecific installation of the isopropenyl moiety. Initial attempts at 1,4‐addition employing organocopper reagents [23] gave moderate to high yields but poor diastereoselectivity. In contrast, rhodium‐catalyzed 1,4‐addition [45, 46] enabled the efficient synthesis of the single diastereoisomer in 65% yield, thus completing the total synthesis of (+)‐ineleganolide.

In summary, we have accomplished a concise 13‐step asymmetric total synthesis of polycyclic cembrane norditerpenoid (+)‐ineleganolide. Key to the success of this work was the construction of the [6,6,5,5,5] ring system, which was accomplished in a single step through a highly selective Diels–Alder reaction with the simultaneous formation of four chiral centers. Taking advantage of the high ring strain inherent to the central cyclohexane ring, a sequential epoxidation/Meinwald rearrangement was performed to release ring strain and realized ring expansion, thereby efficiently constructing the targeted [6,7,5,5,5] pentacyclic skeleton. This strategy elegantly addresses the challenge associated with the construction of complex cycloheptane frameworks through strain‐driven rearrangement of a cyclohexane precursor, which shortens the synthetic route with excellent stereocontrol.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File: anie73612‐sup‐0001‐SuppMat.docx.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.