Total synthesis and biological evaluation of jerantinine E.

Indole alkaloids have attracted the attention of synthetic chemists because of their intriguing structural features and remarkable bioactivities. This interest is perhaps best represented by the Vinca alkaloid vinblastine (1; Scheme 1), isolated from Catharanthus roseus, which is currently among the foremost drugs used to combat cancer. However, de novo synthesis of novel vinblastine analogues remains a daunting task owing to its highly complex structure. Consequently, the discovery of simpler bioactive Aspidosperma alkaloids is of high interest to facilitate the discovery of new anticancer agents. Not surprisingly, intensive synthetic efforts have been directed towards the synthesis of vindoline (2), or simpler Aspidosperma alkaloids such as aspidospermidine (3) and vincadifformine (4). In 2008, Kam and co-workers reported the isolation of seven new Aspidosperma indole alkaloids, jerantinine A–G, from a leaf extract of the Malayan plant Tabernaemontana corymbosa. Among them was jerantinine E (5), which is structurally related to vincadifformine (4) but has a more oxidized indole core. The jerantinines displayed significant cytotoxic activity (half-maximal inhibitory concentration; IC50= 0.27–0.96 mgL 1 (0.70–2.50 mm)) against human KB cells, which is rare among simple Aspidosperma alkaloids. Nevertheless, the mode of action for the cytotoxicity of the jerantinines is currently unknown. Furthermore, the highly oxygenated core of the jerantinines renders them challenging synthetic targets, and a total synthesis of jerantinines A-G has not yet been reported. Herein, we describe the first total synthesis of ( )jerantinine E (5) in 17 steps and 16% overall yield from dvalerolactam (10). The availability of a significant amount of the natural product permitted the separation of the enantiomers, in-depth biological evaluation of the cytotoxic activity in various human cancer cell lines, as well as a first investigation on the origin of the observed cytotoxicity. As shown in our retrosynthesis (Scheme 2, A), we envisaged that the free hydroxy group of jerantinine E (5) could be generated by selective demethylation during the last step of the total synthesis. This strategy would simplify the synthesis by avoiding a multistep hydroxy group protecting/ deprotecting sequence. We expected this selective methyl group removal to proceed under oxidative conditions owing to the presence of the electron-rich nitrogen in the para position. As such, the indole core could be oxidized to the corresponding iminoquinone, followed by in situ reduction to give 5. We next planned to build-up the e-ring system by a bis(alkylation) and to install the methyl ester group on the cring by acylation usingMander s reagent. This ring formation/ acylation sequence was developed during the synthesis of Aspidosperma alkaloids by Rawal and co-workers, albeit with a less oxygenated indole core. The requisite starting material for this sequence could be generated through Nprotecting group removal and deoxygenation of intermediate 6. Tetracyclic product 6 was envisaged to be accessed by a selective cyclization (formal homo-Nazarov reaction) of aminocyclopropane 7, previously developed in our group, forming the central ring system c. Cyclization precursor 7 would finally be obtained by adding an organometallic reagent derived from commercially available indole 9 onto Weinreb amide 8. Our synthesis commenced with the coupling of Weinreb amide 8, obtained in seven steps and 51% overall yield from Scheme 1. Vinca and Aspidosperma alkaloids.