Total synthesis of the akuammiline alkaloid picrinine.

We report the first total synthesis of the complex akuammiline alkaloid picrinine, which was first isolated nearly five decades ago. Our synthetic approach features a concise assembly of the [3.3.1]-azabicyclic core, a key Fischer indolization reaction to forge the natural product's carbon framework, and a series of delicate late-stage transformations to complete the synthesis. Our synthesis of picrinine also constitutes a formal synthesis of the related polycyclic alkaloid strictamine.

Natural products belonging to the akuammiline family of alkaloids have a rich history in the chemical literature.[1] Representative akuammilines 1–6 are depicted in Figure 1, each of which possesses an indoline or indolenine core fused to a daunting polycyclic framework.[2−7] Since the initial isolation report of echitamine (1) in 1875,[2] more than 30 compounds in this family have been isolated from plants predominantly in Southeast Asia. The complex structures of these molecules, along with their interesting biological properties,[1] have recently prompted a number of synthetic investigations. Notable achievements in this arena include total syntheses of aspidophylline A (2) by Zhu,[8] Ma,[9] and our group,[10] vincorine (3) by Qin,[11] Ma,[12] and MacMillan,[13] and scholarisine A (4) by Smith[14] and Snyder.[15,16]

Figure 1

Representative akuammiline alkaloids 1–6.

The present study focuses on picrinine (5), which has not previously been prepared by total synthesis. First discovered in 1965 from the leaves of Alstonia scholaris,[6] picrinine (5) is a highly complex, cage-like molecule that contains a furoindoline core fused to a densely functionalized cyclohexyl ring. In turn, the central cyclohexyl ring is part of a bridged [3.3.1]-azabicyclic framework. Picrinine (5) possesses six stereogenic centers, five of which are contiguous, and contains two N,O-acetal linkages within its polycyclic skeleton. In vitro studies show that picrinine (5) exhibits anti-inflammatory activity through inhibition of the 5-lipoxygenase enzyme.[17] In this Communication, we report the first total synthesis of picrinine (5).

Our retrosynthetic analysis of picrinine (5) is highlighted in Scheme 1.[18] We envisioned that picrinine (5) could be accessed by late-stage introduction of the bis(N,O-acetal) linkage via a proximity-driven cyclization of aminolactol 7.[19] This intermediate would arise from pentacycle 8 through oxidative cleavage of the cyclopentene and subsequent functional group interconversions. Next, in a key step, it was thought that pentacycle 8 could be constructed from a Fischer indolization reaction[20] between phenylhydrazine 9 and tricyclic cyclopentene 10. If successful, this would allow for the introduction of the C7 quaternary stereocenter and provide the complex carbon framework of the natural product. Tricycle 10 would be assembled from enal 11, which could be generated from bridged [3.3.1]-azabicycle 12. Finally, it was hypothesized that bicycle 12 would be accessible from readily available vinyl iodide 13 using a Pd-catalyzed enolate cyclization.

Scheme 1

Our synthesis commenced with the preparation of enal 11 (Scheme 2). Sulfonamide 14, which is available from commercial sources or can be readily prepared,[21] was identified as a suitable starting material. Alkylation of 14 with tosylate 15(22) afforded vinyl iodide 13 in excellent yield. Next, treatment of 13 with PdCl2(dppf) and K2CO3 in MeOH at 70 °C furnished bicycle 12 via a Pd-catalyzed enolate cyclization.[23] Bicyclic ketone 12 was processed to enone 16 by IBX oxidation.[24] Subsequent epoxidation, using sodium perborate as a mild oxidant,[25] transformed 16 into epoxide 17 in 89% yield with complete diastereoselectivity. To arrive at the desired enal intermediate 11, epoxide 17 was subjected to Wittig olefination using (methoxymethyl)triphenylphosphonium chloride and potassium tert-butoxide.[26] Fragmentation of the presumed epoxy enol ether intermediate occurred spontaneously in situ to deliver 11 in 82% yield.

Scheme 2

As shown in Scheme 3, enal 11 could be elaborated to tricyclic cyclopentene 10 through a five-step sequence. Reduction of 11 proceeded smoothly upon exposure to tributyltin hydride, catalytic Pd(PPh3)4, and zinc chloride[27] to furnish 18.[28] Wittig olefination, followed by Dess–Martin oxidation delivered vinyl ketone 19 in 80% yield over two steps. Next, ketone 19 was treated with LHMDS and DMPU, followed by allyl iodide. This step afforded allyl vinyl ketone 20, which was subjected to the Hoveyda–Grubbs second-generation catalyst in refluxing dichloromethane to give tricyclic cyclopentene 10.[29]

Scheme 3

With access to tricycle 10, we explored the critical Fischer indolization and the ensuing furoindoline formation, as summarized in Scheme 4. We were delighted to find that treatment of 10 with phenylhydrazine (9) and trifluoroacetic acid in dichloroethane at 40 °C furnished the desired pentacycle 8 in 74% yield. However, exhaustive efforts to oxidatively cleave the cyclopentene of 8 en route to lactol aldehyde 21 were unsuccessful.[30]

Scheme 4

Hypothesizing that access to the disubstituted olefin of cyclopentene 8 was obstructed by the [3.3.1]-bicycle and the C9 hydrogen, we pursued an alternate strategy, which involved olefin oxidation prior to the Fischer indolization (Scheme 4). Thus, cyclopentene 10 was exposed to an oxidation[31] and protection[32] sequence, providing carbonate 22 in 78% yield over two steps. In the critical Fischer indolization, TFA-promoted reaction of 22 with phenylhydrazine 9 afforded the hexacyclic indolenine product 23 with complete diastereoselectivity. This transformation marks one of the most complex examples of the Fischer indolization reaction[20] and is testament to the reliability of this venerable synthetic method.[33] It should be noted that indolenine 23 exists in equilibrium with its hydrate,[34] so careful purification and 2D NMR analysis were necessary to facilitate structure elucidation. Nonetheless, cleavage of the cyclic carbonate of 23,[35] followed by oxidative cleavage[36] generated lactol 21 as an inconsequential mixture of diastereomers.

Having assembled the carbon scaffold of the natural product, three transformations remained in order to access picrinine (5). From late-stage intermediate 21, this would involve conversion of the aldehyde to the corresponding methyl ester, cleavage of the nosyl group, and cyclization to forge the second N,O-acetal linkage. As shown in Scheme 5, the first of these challenges was addressed by implementing a two-step sequence involving oxidation to the carboxylic acid,[37] followed by esterification. This generated ester 24 in 58% yield over the two steps without disturbing the lactol. Next, our attempts to remove the nosyl protecting group from 24 proved challenging. We ultimately found, however, that denosylation could be achieved using a solid-supported thiol resin.[38,39] Much to our gratification, proximity-driven cyclization of the presumed aminolactol intermediate 7 occurred under the reaction conditions to provide the bis(N,O-acetal) linkage and furnish picrinine (5). Synthetic picrinine (5) was found to be identical to a sample of the natural material in all respects. It should be noted that our total synthesis of 5 also constitutes a formal synthesis of strictamine (6).[40]

Scheme 5

In summary, we have completed the first total synthesis of the akuammiline alkaloid picrinine (5, 18 steps from known ketone 14), which was first isolated nearly 50 years ago. Our synthetic approach features a concise assembly of the [3.3.1]-azabicyclic core, a key Fischer indolization reaction to forge the natural product’s carbon framework, and a series of delicate late-stage transformations to complete the total synthesis. We expect our approach to 5 will enable the syntheses of other alkaloids in the akuammiline family of natural products.