Total Synthesis of (−)‐Kopsinine by an Asymmetric One‐Pot [N+2+3] Cyclization

Abstract

One-pot processes, the sequential conjugation of multiple bond-forming reactions, are powerful tools in organic synthesis. [1] Conjugate addition [2] of lithium amide 2 to enoate 3 forms an N À C bond and generates lithium enolate 5, which is further applicable as a carbonucleophile reacting in an S N 2 manner with alkyl halide. When the halide has two reaction sites like dihalide 6, the initially introduced nitrogen atom in 7 should undergo an intramolecular S N 2 reaction to form piperidine 8 (Scheme 1). This one-pot [N+2+3] cyclization, however, has not been realized, despite several reports on sequential C-alkylation following conjugate addition of lithium amide. [3,4] We describe herein the total synthesis of (À)-kopsinine (1) [5] using the asymmetric one-pot [N+2+3] cyclization based on our asymmetric conjugate aminolithiation [6] of a lithium amide with an enoate.An asymmetric conjugate addition of lithium N-benzyltrimethylsilylamide (2) was performed with tert-butyl N-Boc-indole-3-propenoate (3) using chiral diether 4 in toluene at À78 8C for 0.25 h to afford, after quenching with saturated aqueous ammonium chloride, 3-aminoester 9 with 97 % ee in 89 % yield (Table 1, entry 1). Extension of the reaction time to 3 h did not affect the results, giving 9 with comparable enantiomeric excess and yield (Table 1, entry 2). The same reaction in the presence of internal chlorotrimethylsilane [6a] (TMSCl) gave 9 with 97 % ee in 93 % yield after 1 h (Table 1, entry 3). These data indicated the completion of the conjugate aminolithiation at À78 8C within 0.25 h.Then, alkylation of enolate 5 as the second step was examined. After the aminolithiation at À78 8C for 1.5 h and then at À40 8C for 2 h, alkylation was performed at À40 8C for 4 h by the addition of 3-tert-butyldimethylsilyloxypropyl iodide (6 a; X= I, Y = OTBS) in dimethylformamide [7] (DMF; 233 equiv) to give, after treatment with TBAF to remove the TBS group, 10 with 95 % ee as a sole diastereomer in 60 % yield and enoate 11 in 18 % yield ( Table 2, entry 1). N,N'-Dimethylpropyleneurea (DMPU; 133 equiv) was a better co-solvent, giving 10 with 95 % ee in 73 % yield and 11 in 13 % yield (Table 2, entry 2). Surprisingly, 11 was [a] Dr. Scheme 1. One-pot [N+2+3] cyclization strategy. Bn = benzyl, TMS = trimethylsilyl, Boc = tert-butoxycarbonyl, Ind = indolyl, TBAF = tetrabutylammonium fluoride. X,Y = halide. Table 1. Asymmetric conjugate aminolithiation followed by protonation. Entry T [8C] t [h] Quench Yield 9 [%] ee 9 [%] Yield 3 [%] Yield 12 [%] 1 À78 0.25 aq NH

Critical profiles of chiral diether-mediated asymmetric conjugate aminolithiation of enoate with lithium amide as a key to the total synthesis of (−)-kopsinine

“…DMF was also a good additive, giving anti-7a in 60% yield with high 94% and 95% ee and 99:1 dr in the absence and presence of HMPA, respectively (entries 5 and 6). 2 When the quantity of DMF was decreased to 30 equiv, alkylation became sluggish and after 18 h gave anti-7a with lower 88% ee and 86:14 dr in 57% yield, 2b in 3% yield, and 2c in 10% yield (entry 7). DMPU was the best additive among those examined, giving anti-7a with 95% ee and 99:1 dr in 73% yield, but along with 2b in 2% and 2c in 13% yield (entry 8).…”

“…DMPU was the best additive among those examined, giving anti-7a with 95% ee and 99:1 dr in 73% yield, but along with 2b in 2% and 2c in 13% yield (entry 8). 2 The lower yield and poorer % ee production of 7a compared with 5a, and the production of 2b and 2c in significant amounts implied a couple of possibilities: (1) incomplete conjugate addition reaction of lithium amide 1 with 2a at the time of addition of the additive, resulting in 1 and 2a remaining in the reaction mixture, led to the progression of further conjugate addition reactions without chirality control, and/or (2) addition of the additives to a completed reaction mixture, resulting in retro-Michael-type reac-tion of 4 to 1 and 2a, and again conjugate addition but without chirality control. Since the 89% high yield production of 5a with 97% ee was certainly confirmed by the isolation (Table 1, entry 1), it would be nonsense to doubt the incomplete conjugate addition of 1 to 2a.…”

“…Under these specified conjugate addition conditions 7a of 98% ee was satisfactorily obtained in 89% yield by adding DMPU as the best activator of lithium enolate 4, as described previously (entry 14). 2 Production of deBoc products 2b and 2c from 4 rather than directly from 2a would be possible at a higher temperature-40 °C by nucleophilic attack of lithium amide 1 to an activated vinylogous type urethane carbonyl group of 4, as shown in Scheme 1. This could explain the absence of the formation of a deBoc product of 5a.…”

“…Cyclization of anti-7a with mesyl chloride and triethylamine in methylene chloride at room temperature for 18 h gave piperidine cis-10 in 98% yield (Scheme 2). 2 In the same way, syn-7a was cyclized into piperidine trans-10 in 83% yield. The coupling constants 5.2 and 10.1 Hz of the adjacent methine protons, respectively, indicated the relative configuration of cis-and trans-10.…”

“…We describe herein that the speed of the chiral diether-mediated asymmetric conjugate addition is dependent on the structure of an enoate and is sometimes very slow at low temperature. Another approach toward the total synthesis of (-)kopsinine 1,2 is also the subject of the present study.…”