Asymmetric Synthesis of the Pentacyclic Framework of Kopsia Alkaloids

Abstract

Pyrroloazocine alkaloids from the Kopsia genus exhibit structural complexity and pharmacological potential. We report the concise and asymmetric synthesis of a pentacyclic core from l-proline. The key features include a C–N–C chirality transfer process to prepare the α-tertiary amine intermediate enantioselectively, an eight-membered C ring-forming Friedel–Crafts reaction at the indole C3 position, and a dearomative indole–Claisen rearrangement to form a bridged pentacyclic scaffold. This study provides a versatile platform for the divergent synthesis of diverse Kopsia alkaloids.

Pyrroloazocine alkaloids constitute a distinct subfamily of Kopsia alkaloids, predominantly isolated from species within the Kopsia genus, such as Kopsia grandifolia and Kopsia tenuis. The defining structural motif of these alkaloids is an indole-fused pyrroloazocine core (Figure A). This core adopts a rigid tetracyclic framework that incorporates a carbon bridge between C2 and C20. Individual members further display structural variations such as lactone, diester, or cyclopropane motifs, as exemplified by lapidilectines, , grandilodines, lundurines, and tenuisines (1–8). The Echavarren group proposed that these alkaloids are biosynthetically interconnected via the carbocation intermediate A, formed through oxidative decarboxylation of the C21 ester group of the diester subtype. However, the biosynthetic origin of this pathway has not been fully established experimentally. Preliminary biological studies have shown that pyrroloazocine family alkaloids, such as grandilodine A (1) and lapidilectine B (5), can reverse multidrug resistance in vincristine-resistant cancer cells, underscoring their considerable potential for pharmaceutical development. Beyond these promising biological activities, the highly intricate and architecturally unique structures of these alkaloids have drawn sustained attention from the synthetic community.

1.

Pyrroloazocine Kopsia alkaloids and synthetic approaches.

The synthesis of these alkaloids remains challenging due to the complexity of their core pentacyclic structures and the need for precise stereochemical control. These challenges have spurred the development of many synthetic tactics over the past decades. − As a result, several elegant total syntheses and innovative synthetic strategies have been reported, many of which have been comprehensively reviewed in the literature. Most synthetic strategies differ in their approaches to constructing the indole-fused azabicyclo[4.2.2] skeleton B. These strategies can be broadly categorized into two types: those proceeding through a spiro­[cyclohexane-2-indoline] intermediate C and those involving an indole-fused azocine intermediate D (Figure B). Even though many novel total syntheses of pyrroloazocine alkaloids have been developed, only a few reports describe divergent strategies that furnish multiple subtype alkaloids from a common intermediate. , As part of our ongoing program on the asymmetric synthesis of α-tertiary amine-containing natural products from chiral amino acids, we developed a concise synthesis of the pyrroloazocine indole core with a carbon bridge between C2 and C20, which may provide access to several members of this alkaloid family. Our synthetic strategy differs from previous approaches in that the eight-membered C ring of the tetracyclic core was constructed through an intramolecular Friedel–Crafts reaction of the indole moiety and the C2–C20 carbon bridge was installed through a dearomative indole–Claisen rearrangement (Figure C). Notably, this Claisen-based transformation has not, to our knowledge, been previously applied to the synthesis of pyrroloazocine alkaloids. In this report, we describe our efforts toward the asymmetric synthesis of pentacyclic core of pyrroloazocine alkaloids using l-proline as a chiral pool starting material.

We devised a synthetic route originating from l-proline, which enables access to all three structural variants of pyrroloazocine alkaloids. Retrosynthetic analysis of the target alkaloids, based on simplification of their functional groups, identified compound 9 as a common intermediate (Figure ) corresponding to the cationic intermediate A in Figure A. We envisioned that the Δ , double bond in 9 can be readily transformed into various functional groups leading to a divergent synthesis of several members. The carboxyl group at C16 of 9 was planned to be introduced from the exomethylene group in 10 with the proper stereochemistry required for the synthesis of the targeted natural products. We envisioned that the pentacyclic intermediate 10 could be accessed from 11 through a dearomative indolic-sigmatropic rearrangement, , which establishes the 2,2-disubstitution pattern and thereby constructs the carbon bridge between C2 and C20. The planned sequence involved a Tebbe olefination of the ester functionality, followed by a C–C bond-forming [3,3]-Claisen rearrangement.

2.

Retrosynthetic analysis.

Advanced intermediate 11 was planned to be synthesized by adopting our previously developed protocol, which involves sequential steps such as diastereoselective N-alkylation, stereospecific Stevens rearrangement, and an intramolecular Friedel–Crafts reaction. More specifically, the construction of 11 was designed to proceed through an intramolecular Friedel–Crafts reaction of cyclic O-acyl acetal 12. The asymmetric access to 12 from the bicyclic l-proline derivative 14 was based on the concept of our C–N–C chirality transfer, encompassing a diastereoselective N-quaternization of 14 with indole allylic halide (14/15 → 13) followed by a stereospecific [2,3]-Stevens rearrangement (13 → 12).

Our synthetic efforts commenced with the preparation of indole-derived allyl bromide 15 and bicyclic proline derivative 14 (Scheme ). Azabicycle 14 was prepared from l-proline in two-steps according to our established protocol. For the synthesis of allyl bromide 15, the known indole boronic ester 16 was first prepared from indole by following the protocol reported by Hartwig and co-worker. The boronic ester 16 was subjected to Suzuki cross-coupling conditions with the known vinyl iodide 17 to afford C2-substituted indole 18. At this stage, the methyl carbamate (Moc) group, which is found in the natural products, was introduced in high yield. Removal of the tetrahydropyranyl (THP) protecting group and the subsequent Appel reaction afforded allyl bromide 15 in good overall yield.

1. Synthesis of Pyrroloazocine Core of Kopsia Alkaloids .

^a a) 17, Pd­(dppf)­Cl₂, Cs₂CO₃, 1,4-dioxane/H₂O, 80 °C, 3 h, 80%; b) NaH, ClCO₂Me, DMF, 0 °C to rt, 3 h, 91%; c) 1 N HCl (aq), MeOH, 0 °C to rt, 3 h; d) PPh₃, CBr₄, CH₂Cl₂, 0 °C to rt, 2 h, 81% (2 steps); e’) (i) 14, MeCN, rt, 5 d, (ii) KOtBu, CH₂Cl₂, –78 °C, 2 h, 80%, 39:1 dr (2 steps); e) 14, K₂CO₃, MeCN, rt, 5 d, 63% (15 37% recovered); f) MsOH, CH₂Cl₂, rt, 3 h, 79% (11); f’) pTsOH·H₂O, CH₂Cl₂, rt, 19 h, 75% (20), >20:1 dr.

Next, we explored a stereoselective Cα-alkylation of 14 with 15 to access the α-tertiary amine-containing product 12 via our C–N–C chirality transfer strategy. N-Allylation of 14 with allyl bromide 15 proceeded sluggishly at room temperature; however, after 5 days, ammonium salt 13 was afforded as the sole detectable diastereomer. Evaporation of the MeCN solvent followed by treatment of the crude 13 with KOtBu in CH₂Cl₂ at −78 °C effected a [2,3]-Stevens rearrangement, furnishing α-tertiary amine 12 in 80% overall yield with 39:1 diastereomeric ratio and >98% ee. This two-step sequence could also be executed in a one-pot fashion by adding K₂CO₃ during the N-allylation, providing 12 in 63% yield (37% RSM) with >98% ee. The observed stereochemical outcome is consistent with our earlier studies, which suggest that the reaction proceeds via a Curtin–Hammett-controlled diastereoselective N-alkylation, followed by a stereospecific [2,3]-Stevens rearrangement. The high enantiopurity of the product indicates that the Cα-chirality of l-proline was efficiently conserved through the C–N–C chirality transfer process, in accordance with our previous findings. ,

With intermediate 12 in hand, we turned to an intramolecular Friedel–Crafts reaction to construct the eight-membered C ring. Treatment of acetal 12 with MsOH as an acid catalyst in CH₂Cl₂ afforded 11 in 79% yield, whose structure was confirmed by X-ray crystallography. To the best of our knowledge, this is the first example of an eight-membered ring-forming Friedel–Crafts reaction at the indole C3. A trace amount of the seven-membered isomer 20 was also detected, the formation of which might proceed via an intramolecular Prins-type reaction of the methylene group with an oxonium intermediate, followed by migration of the ester group (see Supporting Information). Interestingly, this isomer became the predominant product when p-toluenesulfonic acid monohydrate was employed as the acid catalyst. The structure of 20 was unambiguously established by single-crystal X-ray crystallographic analysis of its nitrobenzoyl derivative.

After establishing the pyrroloazocine core, we directed our efforts toward constructing the carbon bridge with C20 ester and indole functionalities. We envisioned that a C–C bond-forming dearomative indole–Claisen rearrangement would provide a direct means to achieve this connectivity. , The Tebbe olefination of 11 introduced the requisite enol ether functionality for the rearrangement, affording the desired enol ether 21 in high yield, albeit with limited stability under ambient conditions (Scheme ). Accordingly, 21 was immediately heated in a sealed tube, enabling the [3,3]-sigmatropic Claisen rearrangement to furnish rearranged product 10. The formation of 10 can be rationalized by the concerted suprafacial [3,3]-sigmatropic shift, which installs the carbon bridge while simultaneously effecting dearomatization of the indole nucleus. The rearranged product 10 was also somewhat unstable under ambient conditions. Thus, the crude mixture was directly reduced in situ with NaBH₄ in MeOH, thereby delivering the pentacyclic product 22 in 84% overall yield. The structure of 22 was unambiguously confirmed by single-crystal X-ray analysis of its ether derivative. Consequently, the pentacyclic framework of the target alkaloids has been constructed in only 6 steps from l-proline or 9 steps from indole.

2. Synthesis of Bridged Structure via Claisen Rearrangement .

^a a) Tebbe reagent, pyridine, toluene/THF, 0 °C, 1 h, 98%; b) xylene, 150 °C, 1.5 h, then, NaBH₄, MeOH, rt, 2 h, 84%.

In conclusion, we developed a stereoselective asymmetric synthesis of the pyrroloazocine core of Kopsia alkaloids, starting from l-proline and indole-derived intermediates. By leveraging our C–N–C chirality transfer strategy, the α-tertiary amine–containing intermediate was prepared enantioselectively. A novel intramolecular Friedel–Crafts reaction was followed to construct the eight-membered core ring. A subsequent Tebbe reaction and Claisen rearrangement enabled the installation of a critical carbon bridge and an indole substitution pattern. This concise route not only provides a foundation for the synthesis of several pyrroloazocine natural products but also expands the synthetic toolkit for constructing complex alkaloid scaffolds with precise control. We are currently pursuing the total synthesis of several pyrroloazocine alkaloids based on this strategy, and the results will be reported in due course.

The data underlying this study are available in the published article and its Supporting Information.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.orglett.5c03771.

• Detailed experimental procedures, X-ray crystallographic data, NMR data, and copies of ¹H and ¹³C NMR spectra (PDF)

The authors declare no competing financial interest.

Dedicated to the memory of the late Professor Amos B. Smith III, whose work continues to inspire us.