Highly Stereoselective Synthesis of Polycyclic Indoles through Rearrangement/[4+2] Cycloaddition under Sequential Catalysis

The indole moiety is a privileged structural motif in many biologically active and medicinally valuable molecules. Polycyclic frameworks lead to relatively rigid structures that could be expected to show substantial selectivity in their interactions with enzymes or receptors. Construction of polycyclic indoles usually requires multistep approaches. The preparation of polyfunctional indoles is therefore an important research field. Sequential catalysis involving a binary catalytic system often reduces labor and waste and therefore has attracted much attention recently. Homogeneous catalysis by gold complex has also received considerable attention in recent years. The combination of mechanistically distinct organocatalysis and transition-metal catalysis, especially gold catalysis, has enabled novel transformations beyond those possible with single catalytic systems. During our ongoing investigation on the nitrogenor phosphine-containing Lewis base-catalyzed chemical transformation, we found that nitrogen-containing Lewis bases are efficient catalysts for highly regioselective and stereoselective cycloadditions of allenoates. Thus, we envisaged that it might be possible to explore a direct route to polycyclic indoles by means of a sequential catalysis of gold complex and a nitrogen-containing Lewis base. In 2010, Gagosz’s group reported a novel gold-catalyzed rearrangement of propargyl benzyl ethers that allows for rapid preparation of variously substituted allenes (Scheme 1A). As for isatin-derived propargyl benzyl ether 1a, the a,b-unsaturated ketone 2a could be formed in 20% yield along with the release of HOBn (determined by GC analysis) rather than the allene product in wet dichloromethane (Scheme 1B). Herein, we wish to report an interesting rearrangement/cycloaddition based on sequential catalysis of gold complex and a nitrogencontaining Lewis base to construct polycyclic indoles. In order to clarify the effect of water on the rearrangement of benzyl ether 1a, we first carried out the reaction in freshly distilled dichloromethane containing various concentrations of water. The results are summarized in Table 1, and as can be seen the concentration of water has an obvious effect on this reaction: 1.0 equiv of water is enough to give 2a in good yield. Next, we used propargyl benzyl ether 1a (0.1 mmol) as the substrate to optimize the reaction conditions. The results are summarized in Table 2. Examination of solvent effects revealed that chloroform was the solvent of choice giving 2a in 67% yield, whereas, in other organic solvents such as 1,2-dichloroethane, toluene, acetonitrile or 1,4-dioxane, 2a was formed in lower yield (Table 2, Entries 1–5). Carrying out the reaction in the presence of [(tBu3P)AuCl] or [(Me3P)AuCl] (5 mol%) afforded the desired product 2a in 40% and 52% yields, respectively (Table 2, Entries 6 and7). Using [AuCl] or [AuCl3] instead of [(Ph3P)AuCl] as the gold catalyst gave 2a in 46% and 42% yields, respectively, and [Ph3PAu]3OBF4 as well as [(tBuXPhos)Au(NCMe)]SbF6 were not effective gold catalysts in this reaction (Table 2, Entries 8–11). Changing silver salt to Scheme 1. Gold-catalyzed rearrangement of propargyl benzyl ethers according to A) Gagosz et al. and B) this work. Reagents and conditions: a) [XPhosAu(NCMe)SbF6] (4 mol%), CHCl3, 20 8C or 60 8C, 1–3h. b) [(Ph3P)AuCl]/AgOTf (5 mol%), CH2Cl2 (wet), RT, 2h.