Biomimetic 1,2-amino migration via photoredox catalysis

Abstract

Synthetic organic chemists continually draw inspiration from biocatalytic processes to innovate synthetic methodologies beyond existing catalytic platforms. Within this context, although 1,2-amino migration represents a viable biochemical process, it remains underutilized within the synthetic organic chemistry community. Here we present a biomimetic 1,2-amino migration accomplished through the synergistic combination of biocatalytic mechanism and photoredox catalysis. This platform enables the modular synthesis of γ-substituted β-amino acids by utilizing abundant α-amino-acid derivatives and readily available organic molecules as coupling partners. This mild method features excellent substrate and functionality compatibility, affording a diverse range of γ-substituted β-amino acids (more than 80 examples) without the need for laborious multistep synthesis. Mechanistic studies, supported by both experimental observations and theoretical analysis, indicate that the 1,2-amino migration mechanism involves radical addition to α-vinyl-aldimine ester, 3-exo-trig cyclization and a subsequent rearrangement process. We anticipate that this transformation will serve as a versatile platform for the highly efficient construction of unnatural γ-substituted β-amino acids.

Fig. 1. Biomimetic 1,2-amino migration via photoredox catalysis.

a, The development of new organic methods is often inspired by enzymatic transformation. b, FG migratory transformation is a valuable but underdeveloped strategy for molecular modification, especially for late-stage functionalization. c, Regioisomers often pose quite different properties in drug discovery. d, One particular type of FG migratory transformation that intrigued us is 1,2-amino migration, when an amino group is translocated to its neighbouring location on an sp³ scaffold. Given the abundance of alkyl amines as feedstock reagents and their importance as synthetic building blocks, we believe that the chemical community could greatly benefit from such a method. However, to our knowledge, 1,2-amino migration has not yet been established as a practical technique in organic chemistry. Here we have developed a biomimetic photoredox-catalysed 1,2-amino migration method. This method was inspired by a LAM-catalysed α-lysine to β-lysine transformation. AA, amino acid; TPP, thiamine pyrophosphate; NHC, N-heterocyclic carbene; NMDA, N-methyl-d-aspartate; EC₅₀, half maximal effective concentration.

Fig. 2. Design plan, optimization and control experiments.

a, Design plan for 1,2-amino migration. b, Optimization and control experiments. ^aReaction conditions: 2 (0.1 mmol), CF₃SO₂Na (0.15 mmol), PC (1 mol%), K₃PO₄ (2.0 equiv.), MeCN (3 ml), λ_max = 440 nm Kessil (40 W), N₂, room temperature, 15 h, then work-up with 1 M HCl in THF, 6 h; ¹H NMR yield with 1,3,5-trimethylbenzene as internal standard. ^b1H NMR yield before work-up with 1 M HCl. ^cReaction performed on a 0.3 mmol scale with PC (3 mol%), isolated yield. PC, photocatalyst; SET, single-electron transfer; ND, not detected; r.t., room temperature.

Fig. 3. Synthetic applications.

a, Gram-scale preparation of γ-CF₃-β-Ala-OMe·HCl 5 and further transformations for the synthesis of compounds 48–52. b, Primary attempts to construct chiral γ-phosphonyl-β-amino acids 53 and 54 using (salen)Mn(III) complex. c, Late-stage modification of a drug block and synthesis of tripeptides 55–58. d, Divergent synthesis of a number of γ-FG-β-amino acids 59–62 from one common vinyl amino acid. e, Synthesis of sitagliptin analogue 64. Supplementary Section 6 presents details of the experimental conditions. All yields are isolated yields in their hydrochloride salt forms except for product 64, which is isolated in its free amine form.

Fig. 4. Mechanism investigations.

a, TEMPO-radical inhibition experiment. The reaction was completely inhibited in the presence of TEMPO as a radical scavenger, which is consistent with the radical nature of the 1,2-amino migration process. b, Radical-trapping experiment. The experiments indicate the involvement of phosphonyl radical intermediates. c, Capture of the γ-FG-β-radical of the α-amino-acid derivative and the γ-FG-α-radical of the β-amino-acid derivative. The experiments indicate that the reaction probably proceeds through a cyclization–ring-opening sequence involving the intermediate azacyclopropyl carbinyl radical. d, DFT free-energy profile of 1,2-amino migration. ΔG_sol, Gibbs free energy of solvation. e, Spin density isosurface of the radical intermediates. Details are provided in Supplementary Section 13.

Extended Data Fig. 1. Additional examples of γ-fluoroalkyl-β-AAs and γ-phosphonyl-β-AAs.

All yields are isolated yield over three steps from vinyl-α-AA in their hydrochloride salt forms except for product S14, which is isolated in its free amine form. See Supplementary Information Section 4.1 to 4.2 for full experimental details.

Extended Data Fig. 2. Additional examples of γ-sulfonyl-β-AAs and γ-thioalkyl-β-AAs.

All yields are isolated yield over three steps from vinyl-α-AA in their hydrochloride salt forms except for product S28, which is isolated in its free amine form. See Supplementary Information Section 4.3 to 4.4 for full experimental details. ^b The yield of the one-pot procedure, see Supplementary Information Section 5 for details.