Introduction

Efficiency, selectivity, and generality are major goals in contemporary synthetic organic chemistry, particularly important for total synthesis of natural products and the development of new medicines, agrochemicals, and materials1,2,3. To achieve these goals, concepts such as step-economy3, pot-economy4, catalytic enantioselective total synthesis, unified strategy5, and collective synthesis1 have been advanced. Central to these strategies are key reactions or key steps6, such as cycloadditions/annulations1, one-pot sequential/tandem/cascade reactions1, and multicomponent reactions7. These generally feature simultaneous formation of at least two bonds in target molecules. Although computer-aided key step identification has been introduced recently8, it relies on known key reactions, highlighting the demand for novel key reactions/steps for highly efficient total synthesis of natural products, new medicines, and agrochemicals.

Pyrrolidine and piperidine alkaloids, as well as fused bicyclic frameworks like indolizidine alkaloids, are prevalent in both plant9 and animal10,11 kingdoms. Ant venom alkaloids10 and over 800 alkaloids isolated from the skin of poison frogs as of 200511 mostly possess a stereogenic unbranched α-alkyl or a cis- or trans-α,α′-dialkyl pyrrolidine/piperidine motif (Fig. 1f). Such structural motifs are also found in medicinal agents12. Although preliminary biological studies on some alkaloids reveal diverse biological activities such as cytotoxic, insecticidal, hemolytic, antibacterial, antifungal, and necrotic properties10,11,13, comprehensive investigation has been hampered by a lack of samples from natural sources. Consequently, developing methods for enantioselective synthesis of these alkaloids has been a significant objective in organic synthesis14,15,16,17,18,19,20. However, most reported methods rely on stoichiometric chiral pools and chiral auxiliaries16, necessitating long synthetic sequences for transforming commercially available materials into target molecules5. The development of catalytic asymmetric reactions has led to elegant total syntheses of several pyrrolidine, piperidine, and indolizidine alkaloids17,18,19.

Fig. 1: Background and summary of this work.
figure 1

a Reported racemic direct reductive alkylation of amides. b The enantioselective reductive alkynylation of tertiary benzamides: method and limitation. c The enantioselective reductive alkynylation/alkylation of secondary amides: method and limitation. d Challenging substrates and targets for the highly efficient catalytic enantioselective total synthesis of alkaloids based on the catalytic enantioselective reductive alkylation/alkynylation of amides. e This work: The key reaction and its application to the general, one-pot, catalytic enantioselective syntheses of pyrrolidine piperidine, and indolizidine alkaloids. f Representative alkaloids/achieved targets: eight alkaloids and one anticancer antipode.

Full size image

Despite the importance of pyrrolidine, piperidine, and indolizidine alkaloids, only a few highly efficient total syntheses of these compounds have been reported. Between 2012 and 2013, Kroutil et al. documented a multi-enzymatic cascade process for the three-step total synthesis of either enantiomer of isosolenopsin (A-3) from commercially available compounds20,21. In 2013, Glorius’ group developed a four-step total synthesis of ent-monomorine I (ent-A-8) based on ruthenium-NHC-catalyzed asymmetric hydrogenation of indolizines22. However, due to the inherent substrate specificity of enzymatic reactions, the former method was limited to the synthesis of 2-methylpiperidine/pyrrolidine alkaloids, and the latter allowed access only to the unnatural enantiomer of monomorine I. Consequently, achieving highly efficient and general total synthesis of unbranched α-alkyl and α,α′-dialkyl pyrrolidine, piperidine, and indolizidine alkaloids with high enantioselectivity (≥90% ee) remains a significant challenge, despite recent advancements in the catalytic asymmetric synthesis of chiral α-alkylamines and allyl/propargylamines23,24,25.

Amides are a class of bench-stable, readily available compounds widely used for synthesizing substituted amine motifs in the context of total synthesis of alkaloids and medicinal agents14,15. Nonetheless, due to the high stability of amides, their transformation into substituted amines requires multiple synthetic steps5. Over the past decade, the direct reductive functionalization of amides to yield amines (Fig. 1a) has emerged as a powerful strategy for total synthesis of alkaloids26,27,28,29,30,31,32,33,34,35,36,37,38. Despite the significant progress, the direct catalytic reductive alkylation of amides to yield chiral amines remains a formidable challenge. Recently, Huang and Wang established an Ir/Cu tandem catalysis protocol39 and an Ir/Cu/organocatalyst multicatalysis protocol40 for the enantioselective reductive alkynylation of amides (Fig. 1b, c). However, the former (Fig. 1b) is restricted to the transformation of tertiary benzamide derivatives39, and the latter (Fig. 1c), while broader in scope, does not permit the highly enantioselective reductive alkynylation of unbranched aliphatic secondary amides40. Therefore, the catalytic asymmetric reductive alkynylation of aliphatic tertiary amides, particularly the α-unbranched ones with high enantioselectivity, remains an unresolved challenge (Fig. 1d).

Results

Synthetic design

To develop a highly efficient and universal enantioselective approach to unbranched α-alkyl and α,α′-dialkyl pyrrolidine, piperidine, and indolizidine alkaloids, it was necessary to identify and establish a key reaction. To accomplish this, several challenges needed to be addressed: (1) The use of protecting groups should be minimized, and multi-step functional group manipulations should be avoided. Ideally, several reactions, including deprotection, could be performed sequentially in a single pot; (2) A versatile approach is required to access diverse targets with significant structural variation, including mono- and bicyclic, five- and six-membered saturated nitrogen heterocycles with varying substituents and the presence or absence of an N-methyl group; (3) The current state of organic synthesis does not allow for direct, catalytic enantioselective reductive alkylation of carbonyl compounds with Grignard or organolithium reagents; (4) As previously mentioned, the catalytic asymmetric reductive alkynylation of aliphatic tertiary amides, particularly the α-unbranched ones, has yet to achieve high enantioselectivity (Fig. 1d).

We hypothesized that by leveraging the rich chemistry of alkynes41,42,43, we could develop a versatile and highly enantioselective method for the direct reductive alkynylation of unbranched aliphatic amides. This would create a general and highly efficient entry point to the diverse alkaloids depicted in Fig. 1f. In this context, the formation of the C–C bond through catalytic asymmetric reductive alkynylation would not only establish the first asymmetric center but also induce the formation of other stereogenic centers during C–N bond formation (intramolecular reductive amination), constituting a key reaction. Moreover, this methodology highlights the use of alkynes as surrogates for alkyl carbanions44, ensuring the necessary chemoselectivity and generality.

We now report a general amide/alkyne-based reductive alkynylation/annulation strategy (referred to as AARA methodology) for the one-pot asymmetric synthesis of diverse α-alkyl or α,α′-dialkyl pyrrolidine, piperidine, and indolizidine alkaloids from aliphatic tertiary amides and terminal alkynes (Fig. 1e).

Catalytic enantioselective reductive alkynylation of tertiary aliphatic amides: reaction development

Upon identifying the crucial reaction, our strategic focus revolved around developing a general and highly enantioselective catalytic reductive alkynylation process for α-unbranched aliphatic amides. We chose N,N-dibenzylisobutyramide (1a) as our test aliphatic amide, and phenylacetylene (2a) as the alkyne partner, to investigate the catalytic asymmetric reductive alkynylation. Initial screening involved chiral ligands. Various chiral ligand classes have been employed in catalytic asymmetric alkynylation reactions to yield propargylamines25,45,46,47,48, including Li’s tridentate ligand pyridine bisoxazoline [PyBox, (S,S)-L2]45, Knochel’s chiral bisphosphine ligand [(R,R)-Quinoxp, L5]46, Carreira’s N-PINAP (L4)47, and Ma’s recently developed Pyrinap ligands48. We selected some cost-effective, commercially available ligands and tested a few new ones.

We initially established an efficient protocol consisting of sequential treatment of amide 1a with Vaska’s complex [IrCl(CO)(PPh3)2] (1.0 mol%), tetramethyldisiloxane49 (TMDS, 2.0 equiv) at room temperature for 10 min in toluene, followed by phenylacetylene (2a, 1.2 equiv), CuBr (5.0 mol%), and a chiral ligand (5.0 mol%). After evaluating several chiral ligand classes (Table 1), including those based on (S)-(+)-BNPA, L1, PyBox [(S,S)-L2], bisoxazoline [BOX, (S,S)-L3], (R,P)-N-PINAP (L4), (R,R)-QuinoxP* (L5), and Zhang’s chiral Ming-Phos L650, (R,P)-N-PINAP [(R,P)-L4)] emerged as the most promising, delivering propargylic amine 3a with an excellent enantiomeric excess of 93% but a moderate 58% yield (Table 1, entry 1).

Table 1 Optimization of reaction conditions
Full size table

Encouraged by this outcome, we assessed various conditions for reaction optimization, with key results summarized in Table 1. Solvent choice influenced the asymmetric reaction (see Table 1, entries 1–5), and the mixed solvent system of toluene/THF = 3:1 (v/v) proved optimal in terms of both yield (79%) and enantioselectivity (90% ee) (Table 1, entry 5). Multiple Cu(I) and Cu(II) copper salts demonstrated effectiveness (see Table 1, entries 5–10), with CuBr and CuI yielding comparable results (79%/80% yield; 90%/89% ee) (Table 1, entries 5 and 6).

We then investigated the impact of additives (see Table 1, entries 11–16). Employing 5.0 mol% triethylamine as an additive, in conjunction with CuBr or CuI, proved beneficial, yielding 3a with 93% ee (91% yield, entry 11) and 92% ee (94% yield, entry 15), respectively. Interestingly, we discovered that for CuBr- and CuI-catalyzed reactions, ee could be further enhanced to 96% and 98%, respectively, by inversely adding the partial reduction mixture to a combination of NEt3 and chiral ligand (R,P)-L4 at 0 °C (Table 1, entries 17 and 18). Consequently, the optimized conditions for reductive alkynylation of amide 1a were defined in entry 18 of Table 1. The absolute configuration of propargylamine 3a was determined to be S by comparing optical rotation data with those reported46,47.

Scope of amides and alkynes

Due to the unexplored potential of catalytic, enantioselective reductive alkynylation in α-unbranched aliphatic amides and the necessity of resulting propargylamines as intermediates for alkaloid synthesis (Fig. 1), the reaction applicability was examined. As demonstrated in Fig. 2A, various α-unbranched N,N-dibenzylamides (1b–1g), including propionamide 1g, reacted smoothly, yielding the corresponding propargylamines 3b–3g in 79–94% yields and 90–95% enantiomeric excess (ee). It is important to note that a one-pot alkaloid synthesis strategy requires the use of amides with diverse N,N-substituents. Consequently, N-allyl-N-benzylamide 1h and N-Me-N-PMB amide 1i were tested for asymmetric reductive alkynylation. To our delight, the resulting propargylamines 3h and 3i were produced with excellent ee (90% and 93%). The direct reductive alkynylation of cyclic amides was then investigated. Although the reductive alkynylation of six-membered lactam 1j resulted in moderate ee (53%), cyclic propargylamine 3k was formed with an outstanding ee (97%) when reacting with the seven-membered lactam 1k.

Fig. 2: Scope of Ir/Cu relay catalyzed enantioselective reductive alkynylation of aliphatic amides.
figure 2

aReaction conditions: amide 1 (0.4 mmol), IrCl(CO)(PPh3)2 (1 mol%), TMDS (0.8 mmol), toluene:THF = 3:1 (1 mL), rt, 10 min, 25 °C, then (R,P)-L4 (5 mol%), CuI (5 mol%), Et3N (5 mol%), alkyne 2 (0.48 mmol), toluene:THF = 3:1 (1 mL), rt, 5 d, 25 °C. b. Isolated yield. c. Determined by HPLC on a chiral stationary phase. dFor the structures of amides and alkynes, see Tables 1 and 2 in the supplementary information. A Aliphatic amide scope. B Alkyne scope.

Full size image

The versatility of our strategy heavily relies on the flexibility of the alkyne partners. Therefore, we examined the scope of different alkynes (Fig. 2B). The reaction displayed insensitivity to acetylene substituents, as arylacetylene 2b, trimethylsilylacetylene (2d), cyclopropylacetylene (2e), and alkylacetylene 2f yielded analogous propargylamines 3l–3o with 93–98% enantiomeric excess (ee).

Chemoselectivity and functional group tolerance

Achieving chemoselectivity and functional group tolerance are two critical aspects for the success of our one-pot strategy. Consequently, we investigated the reactions of various functionalized amides containing either ester or keto groups, with the results summarized in Fig. 3, part A. The reaction accommodated ester (1l), cyclic ketone (1m), aryl ketone (1n and 1o), and aliphatic (1p–1r) keto groups, yielding the corresponding functional propargylamines 3t–3y. Furthermore, the alkyne partner was able to support different functional groups, such as a chlorine substituent in 4-chlorobut-1-yne (2i, Fig. 3, part B).

Fig. 3: Chemoselectivity and functional group tolerance of the catalytic enantioselective reductive alkynylation of aliphatic amides.
figure 3

Part A Chemoselectivity and functional group tolerance of the amides. Part B Chemoselectivity and functional group tolerance of the alkyne. aFor reaction conditions, see footnote a in Fig. 2. bIsolated yield. cDetermined by HPLC on a chiral stationary phase.

Full size image

One-pot, catalytic asymmetric total synthesis of eight pyrrolidine, piperidine, and indolizidine alkaloids and an anticancer unnatural enantiomer from amides and alkynes

After developing a versatile and highly chemoselective catalytic asymmetric reductive alkylation of amides, we were well-positioned to explore its application in the concise enantioselective synthesis of biologically active alkaloids. To begin, we conducted a gram-scale synthesis of propargylamine 3a. A 5 mmol (1.337 g) scale reaction of 1a yielded 1.589 g of 3a (90% yield) with outstanding enantioselectivity (96% ee) (Table 1).

Next, our attention turned to the catalytic asymmetric total synthesis of alkaloids found in ant venom and poison-frog. As previously mentioned, ant venoms and poison-frog skins are rich sources of 2,6-dialkylpiperidines, 2,5-dialkylpyrrolidines, and indolizidine alkaloids. For instance, cis-2-methyl-6-nonyl-piperidine51 (A-3), cis-2-methyl-6-undecyl-piperidine52 (A-4), and the N-methyl derivative: cis-1,2-dimethyl-6-nonyl-piperidine53 (A-6), as well as 2,5-dialkylpyrrolidines like 2-n-butyl-5-n-heptylpyrrolidine (A-2) are ant venom alkaloids isolated from various fire ant species910.

Intriguingly, cis-2-methyl-6-nonyl-piperidine11 (A-3), cis-2-methyl-6-undecyl-piperidine11 (A-4), and 2-n-butyl-5-n-heptylpyrrolidine (A-2) have also been extracted from amphibian skin, where they are designated as 225I11, 253J11, and 225C11, respectively. Among these alkaloids, the absolute configurations of those sourced from three ant samples—S. geminutu workers, S. invictu workers, and S. invictu alates—have been determined. However, the configurations of alkaloids isolated from other origins have not yet been ascertained due to sample scarcity, and they may differ based on possible variations in biosynthetic pathways51.

We chose the alkaloids cis-2-methyl-6-nonyl-piperidine (cis-225I)52 (A-3) and cis-2-methyl-6-undecyl-piperidine (cis-253J)53 (A-4) as our initial targets, as they are potent inhibitors of neuronal nitric oxide synthase and of [3H]-perhydrohistrionicotoxin binding to sites associated with the nicotinic receptor-gated ion channel in Torpedo californica electric organ, respectively. To determine the enantioselectivity of the one-pot total synthesis, we first explored a two-step synthesis (Fig. 4). Consequently, δ-keto amide 1p, easily prepared in a single step from commercially available 5-oxo-hexanoic acid, was subjected to chemoselective and enantioselective alkynylation with non-1-yne (2j), yielding propargylamine ketone 3z in 90% yield and 98% ee. Scaling up the reaction to 1.238 g (4.0 mmol) provided 3z in 86% yield and 92% ee. In the presence of concentrated HCl, Pd(OH)2/C-catalyzed tandem hydrogenation-hydrogenolysis and reductive cyclization of keto-propargylamine 3z yielded the desired (−)-cis-2-methyl-6-nonyl-piperidine (cis-225I) (A-3), isolated as hydrochloride salt, in 90% yield and excellent cis-diastereoselectivity (dr > 30:1).

Fig. 4: One-pot, catalytic enantioselective total syntheses of pyrrolidine and piperidine alkaloids from δ/γ-keto amides and alkynes.
figure 4

(1) One-pot synthesis of cis-225I (A-3). (2) One-pot synthesis of cis-253J (A-4). (3) One-pot synthesis of cis-197F (A-5). (4) One-pot synthesis of cis-225C (A-2). (5) One-pot synthesis of cis-1,2-dimethyl-6-nonyl-piperidine (A-6).

Full size image

Encouraged by this outcome, we undertook a one-pot tandem reductive alkynylation-hydrogenation-N-debenzylation-reductive cyclization of δ-keto amide 1p, resulting in alkaloid (−)-cis-2-methyl-6-nonyl-piperidine (cis-225I) (A-3) in one-pot with 71% yield and 30:1 dr in favor of the cis-diastereomer. The enantiomeric excess of (−)-A-3 was deduced from the intermediate 3z to be 92%.

In subsequent investigations, we first conducted the catalytic enantioselective reductive alkynylation to determine the ee of the propargylamine intermediate and subsequently undertook the one-pot tandem reaction to synthesize the natural product. Thus, the one-pot tandem reaction of keto amide 1p with undec-1-yne (2g) produced alkaloid (−)-cis-2-methyl-6-undecyl-piperidine (253J) (A-4) in 72% yield and >20:1 dr (cis/trans). The ee of (−)-cis-A-4 was 90%, deduced from that of the keto amine 3v.

While numerous enantioselective syntheses of 2-methyl-6-alkyl-piperidine ant venom alkaloids like A-3 and A-4 have been reported20,21,54, synthesis of poison-frog alkaloid cis-197F11 (A-5), a 2-ethyl-6-alkyl-piperidine alkaloid, has not been reported. Extending our methodology to keto amide 1q and hex-1-yne (2f) using (R,P)-L4) as the chiral ligand directly afforded alkaloid (−)-cis-197F (A-5) in 62% yield with >20:1 dr (cis/trans) and 91% ee, determined on 3w.

The one-pot tandem reaction could also be applied to the synthesis of 2,5-cis-dialkylpyrrolidine alkaloids. Unlike their piperidine counterparts, 2,5-cis-dialkylpyrrolidines are rare among ant venom alkaloids10,55. Interestingly, while 2-butyl-5-heptylpyrrolidine (A-2) from the poison gland of the South African ant Solenopsis punctaticeps was identified as the trans-diastereomer, the stereochemistry of the same alkaloid found in one Colombian population of Dendrobates histrionicus, named with the code 225C, has not yet been determined11,56. Since little enantioselective synthesis of cis-2-butyl-5-heptylpyrrolidine cis-225C (A-2) has been reported57 and a racemic synthetic product of cis- and trans-diastereomers was found to be a high-affinity ligand58, we decided to undertake a one-step synthesis of this alkaloid. Employing (R,P)-L4 as the chiral ligand, the reductive coupling of γ-keto amide 1r with hept-1-yne (2h) in tandem with reductive cyclization produced pyrrolidine alkaloid (−)-cis-225C (A-2) in 65% yield with >20:1 dr (cis/trans) and 90% ee, assessed at the propargylamine 3x stage.

It is important to note that N-methyl piperidine and N-methyl pyrrolidine structures are found in certain alkaloids. Our methodology’s adaptability to N-substituents provides a direct approach to these structural motifs, where the N-methyl group is incorporated as part of the starting material. Using (R,P)-L4 as the chiral ligand, the one-pot reaction of N-benzyl-N-methyl-5-oxohexanamide 1s with non-1-yne (2j) directly produced the N-methylpiperidine alkaloid (−)-cis-1,2-dimethyl-6-nonyl-piperidine (A-6) in 76% yield, with a >20:1 dr (cis/trans) and 94% ee, as determined on 3aa.

Next, we focused on the enantioselective synthesis of (−)-bgugaine (A-1) (Fig. 5), a plant alkaloid derived from Arisarum vulgare tubers, known for its antibacterial activity against Gram-positive bacteria and antimycotic activity against some Candida and Cryptococcus strains. Employing (R,M)-L4 as the chiral ligand in the reductive alkynylation process, the one-pot reaction of simple amide 1t with commercially available 3,3-diethoxyprop-1-yne (2k) yielded natural (−)-bgugaine (A-1) with an 81% yield [94% ee for (−)-3ab] (Fig. 5). Since the unnatural enantiomer, (+)-bgugaine (ent-A-1), has reported anticancer activity59,60, its synthesis was also pursued61. Using chiral ligand (R,P)-N-PINAP [(R,P)-L4], the one-pot reductive transformation yielded (+)-bgugaine (ent-A-1) with an 82% yield and 92% ee, as determined on propargylamine intermediate (ent-)-3ab.

Fig. 5: One-pot, catalytic enantioselective total syntheses of pyrrolidine and indolizidine alkaloids from amides/δ-keto amides and functionalized alkynes.
figure 5

(6) One-pot synthesis of (−)-bgugaine (A-1). (7) One-pot synthesis of (+)-bgugaine (ent-A-1). (8) One-pot synthesis of (−)-5,9Z-indolizidine 209 D (A-7). (8) One-pot synthesis of (+)-monomorine I (A-8).

Full size image

Lastly, we directed our attention to the asymmetric total synthesis of indolizidine alkaloids. To create the enantiomer represented by (−)-5,9Z-indolizidine alkaloid 209 D18,62 (A-7), (R,M)-L4 was employed as the chiral ligand for the reductive alkynylation. A combination of δ-keto amide 1u with functionalized alkyne 2k directly yielded indolizidine alkaloid (−)-5,9Z-indolizidine 209 D (A-7) with a 60% yield (90% ee, determined at the intermediate 3ac stage). The spectral and physical data of our synthetic product matched the reported values.

In our final demonstration of the power of our methodology, the enantioselective total synthesis of (+)-monomorine I (A-8) was envisioned. This 5Z,9Z-3-methyl-5-butyl indolizidine alkaloid is a pharaoh ant trail pheromone isolated from Monomorium pharaonis L63. and later from the ant Myrmicaria melanogaster from Brunei55. In 1985, Husson and Royer achieved the first enantioselective total synthesis of (−)-monomorine I64, determining the absolute configuration of the natural (+)-monomorine I as 3R,5S,9S. This alkaloid has since become the target of extensive synthetic efforts19,65. For the synthesis of (+)-monomorine I (A-8), (R,P)-L4 was used as the chiral ligand. The one-pot asymmetric reaction of δ-keto amide 1p with functionalized alkyne 2l directly yielded alkaloid (+)-monomorine I (A-8) with a 51% yield, excellent ee (93%), and diastereoselectivity, predominantly forming the naturally occurring (5Z,9Z)-diastereomer. The spectral and physical data of our synthetic product matched the reported values65. Remarkably, this catalytic enantioselective tandem reaction formed one C–C bond and two C–N bonds, establishing the correct relative and absolute stereochemistries of three stereogenic centers, and producing one isomer from eight possible stereoisomers.

Discussion

A plausible mechanism for the one-pot, catalytic enantioselective total synthesis of indolizidine alkaloid monomorine I is outlined in Fig. 6. The first stereogenic center was established by the catalytic enantioselective reductive alkynylation of δ-keto amide 1p. Under acidic and Pd-catalyzed hydrogenation/hydrogenolytic conditions, the cleavage of ketal protecting group, hydrogenation of alkynyl moiety, cleavages of two N-benzyl groups, and intramolecular reductive amination occurred sequentially to give intermediate 4a, which could be isolated, and its structure was determined by comparison with the data reported for a racemic 4a (see the Supplementary Information)65. Further one-pot intramolecular reductive amination occurred via presumed intermediate 4b to deliver the final product A-8. The stereochemical course of the two intramolecular reductive aminations could be understood in terms of stereoelectronic effect as showcased by Stevens and Lee in their seminal racemic total synthesis of monomorine I66.

Fig. 6: Plausible mechanism.
figure 6

Plausible mechanism for the one-pot, catalytic enantioselective total synthesis of indolizidine alkaloid monomorine I.

Full size image

In summary, we have established and refined a multifaceted approach for the enantioselective catalytic reductive alkynylation of linear aliphatic tertiary amides. Utilizing our highly effective Amide/Alkyne-based Reductive Annulation Strategy (AARA methodology), we have achieved a one-pot, catalytic enantioselective total synthesis of eight pyrrolidine, piperidine, and indolizidine alkaloids, as well as an anticancer antipode, with 90–98% enantiomeric excess (ee). Given that the amides/alkynes are either commercially available or can be synthesized in just one or two steps from readily accessible compounds, our approach represents the most concise and adaptable catalytic asymmetric total syntheses of the target alkaloids to date. This robust and highly stereoselective method sets the stage for the efficient catalytic enantioselective total synthesis of other bioactive pyrrolidine, piperidine, indolizidine, and pyrrolizidine alkaloids, as well as therapeutically significant agents.

Methods

General procedure for the one-pot catalytic asymmetric reductive alkynylation of tertiary amides

To a flame-dried Schlenk tube were added CuI (3.8 mg, 0.02 mmol), (R,P)-N-PINAP [(R,P)-L4] (11.2 mg, 0.02 mmol) and toluene: THF = 3:1 (1 mL) under a N2 atmosphere. After being stirred at room temperature for 5 min, triethylamine (6 μL, 0.02 mmol) and alkyne 2 (0.48 mmol) were added, and the resulting mixture was stirred at room temperature for 30 min.

To another flame-dried Schlenk tube were added sequentially IrCl(CO)(PPh 3)2 (3.12 mg, 1 mol%), an amide 1 (0.4 mmol, 1 equiv), TMDS (144 μL, 0.8 mmol, 2 equiv) and toluene:THF = 3:1 (1 mL) under N2 atmosphere at room temperature. After being stirred for 10 min, the resulting mixture was added to the abovementioned Schlenk tube containing CuI, (R,P)-L4, triethylamine and a terminal alkyne at 0 °C. The mixture was stirred at room temperature for 5 d. The reaction mixture was concentrated under reduced pressure, and the residue was purified by flash chromatography on silica gel eluting with petroleum ether/ethyl acetate to afford the corresponding chiral propargylic amine 3.

General procedure for the one-pot catalytic asymmetric synthesis of alkaloids

To a flame-dried Schlenk tube were added CuI (9.5 mg, 0.05 mmol), (R,P)-L4 (28.0 mg, 0.05 mmol), and toluene: THF = 3:1 (2 mL) under a N2 atmosphere. The resulting mixture was stirred at room temperature for 5 min. Triethylamine (13 μL, 0.05 mmol) and a terminal alkyne 2 (1.2 mmol) were added, and the mixtures was stirred at room temperature for 30 min.

To another flame-dried Schlenk tube were sequentially added IrCl(CO)(PPh 3)2 (7.8 mg, 1 mol%), an keto amide 1 (1.0 mmol, 1 equiv), TMDS (0.36 mL, 2 mmol) and toluene:THF = 3:1 (3 mL) under a N2 atmosphere at room temperature. After being stirred for 10 min, the resulting mixture was added to the abovementioned Schlenk tube containing CuI, (R,P)-L4, triethylamine and alkyne at 0 °C. The mixture was stirred at room temperature for 5 d, then filtered through a short pad of Celite. The filtrate was concentrated under reduced pressure, and the residue was dissolved in MeOH (5 mL). To the resulting mixture, Pd(OH)2/C was added, and the flask was purged three times with hydrogen. The suspension was stirred for 12 h at room temperature under a hydrogen atmosphere (1 atm). The resulting mixture was filtered through a short pad of Celite, and washed with methanol (50 mL). The filtrate was concentrated and the residue was purified by flash chromatography on silica gel to afford the corresponding alkaloid.