Influence of Reaction Conditions on Enzymatic Enantiopreference: the Curious Case of HEwT in the Synthesis of THF-Amine

Abstract

Enzymatic enantiopreference is one of the key advantages of biocatalysis. While exploring the synthesis of small cyclic (chiral amines) such as 3-aminotetrahydrofuran (THF-amine), using the (S)-selective transaminase from Halomonas elongata (HEwT), inversion of the enantiopreference was observed at increasing substrate loadings. In addition, the enantiopreference could be altered by variation of the ionic strength, or of the co-solvent content in the reaction mixture. For example, using otherwise identical reaction conditions, the presence of 2 M sodium chloride gave (R)-THF-amine (14 % ee), while the addition of 2.2 M isopropyl alcohol gave the (S)-enantiomer in 30 % ee. While the underlying cause is not currently understood, it appears likely that subtle changes in the structure of the enzyme cause the shift in enantiopreference and are worth exploring further.

Keywords: THF ketones; amines; biocatalysis; enantioselectivity; enzyme; transaminases.

Scheme 1

Previous work: synthesis of THF‐amine, as well as 3‐aminotetrahydrothiophene, N‐methyl‐3‐aminopyrrolidine, and N‐methyl‐3‐aminopiperidine, showing variable ee values in batch vs. flow. Current work: Synthesis of THF‐amine from THF‐ketone, using the (S)‐selective TA HEwT, with variable ee values in batch depending on the reaction conditions.

Scheme 2

Simplified equilibrium for a generic reductive amination. Since both enantiomers have identical thermodynamic properties (standard Gibbs free energy), the equilibrium constants between the prochiral ketone (K) and either enantiomer of the amine (A) are identical. Since both equations have the same concentration of ketone in the denominator, the concentration of each enantiomer has to be identical for both equilibria to be satisfied.

Figure 1

A: The production of THF‐amine from THF‐ketone over time, in a biotransformation containing THF‐ketone (10 mM), (S)‐α‐methylbenzylamine (SMBA) (1 eq.) and purified HEwT (1.0 mg/mL). B: The production of THF‐amine from THF‐ketone with increasing concentrations of purified HEwT, in biotransformations containing THF‐ketone (10 mM) and SMBA (1 eq.). Samples taken after 3 h and C: samples taken after 24 h. D: The production of THF‐amine from THF‐ketone (THF=O) at increasing substrate concentrations, in biotransformations containing IPA (5 eq.), and purified HEwT (0.72 mg/mL). Samples taken after 3 h. All reactions contained PLP (1 mM), KP _i ‐buffer (100 mM); pH 8, 30 °C. Conversions determined by chiral RP‐HPLC, after FMOC‐derivatization. Signed ee values; +ve (S)‐enantiomer, ‐ve (R)‐enantiomer. Connecting lines added for clarity.

Figure 2

A: The production of THF‐amine from THF‐ketone (THF=O) at increasing concentration of THF‐ketone, in biotransformations containing a fixed concentration of IPA (50 mM), and purified HEwT (1 mg/mL). Samples taken after 3 h. B: The production of THF‐amine from THF‐ketone (THF=O) at increasing concentration of IPA, in biotransformations containing a fixed concentration of THF‐ketone (100 mM), and purified HEwT (1 mg/mL). Samples taken after 3 h. All reactions contained PLP (1 mM), KP _i ‐buffer (50 mM); pH 8, 30 °C. Conversions determined by RP‐HPLC, ee values determined by chiral RP‐HPLC, after FMOC‐derivatization. Error bars represent standard deviations (n=2). Signed ee values; +ve (S)‐enantiomer, ‐ve (R)‐enantiomer. Connecting lines added for clarity.

Figure 3

Docking of THF‐ketone into the entrance of the active site of wild‐type HEwT, showing the hydrogen bond to W56. Figure reproduced from ref. [10] (CC BY 4.0 license).

Figure 4

A−C: The production of THF‐amine from THF‐ketone (THF=O) in biotransformations containing THF‐ketone (10 mM), SMBA (1 eq.), purified HEwT (0.25 mg/mL), PLP (0.1 mM), and KP _i ‐buffer (50 mM); pH 8, with varying concentrations of additives. Samples taken after 3 h. A: ⁱ PrOH (0.05 to 5 M), B: NaCl (0.01 to 3 M)), C: NH₄Cl (0.05 to 1.5 M)). In the case of C, the decrease of conversion at increasing concentrations of ammonium chloride might be (partially) due to interference of ammonium in the FMOC derivatization, i. e. an artefact of the analysis rather than a real reduction in conversion. This is supported by the complete consumption of FMOC−Cl, the appearance of an unidentified peak (presumably FMOC‐NH₃), and a reduction in size of the FMOC‐derivatized SMBA peak (Figure S1). D: The production of THF‐amine from THF‐ketone (THF=O) in biotransformations containing THF‐ketone (10 mM), SMBA (1 eq.), purified HEwT (0.25 mg/mL), PLP (1 mM), and KP _i ‐buffer (100 mM), with varying pH. Samples taken after 3 h. All reactions carried out at 30 °C. Conversions (A−C) determined by RP‐HPLC, conversions (D) and ee values determined by chiral RP‐HPLC, after FMOC‐derivatization. Error bars represent standard deviations (n=2). Signed ee values; +ve (S)‐enantiomer, ‐ve (R)‐enantiomer. Connecting lines added for clarity.

Figure 5

A: The production of THF‐amine from THF‐ketone (THF=O) at increasing concentrations of cadaverine, in biotransformations containing a fixed concentration of THF‐ketone (100 mM), and purified HEwT (1 mg/mL). Samples taken after 3 h. B: The production of 2‐aminobutane from butanone at increasing substrate concentrations, in biotransformations containing IPA (5 eq.), and purified HEwT (1 mg/mL). Samples taken after 3 h. All reactions contained PLP (0.1 mM), KP _i ‐buffer (50 mM); pH 8, 30 °C. Conversions determined by RP‐HPLC, ee values determined by chiral RP‐HPLC, after FMOC‐derivatization. Error bars represent standard deviations (n=2). Signed ee values; +ve (S)‐enantiomer, ‐ve (R)‐enantiomer. Connecting lines added for clarity.

Figure 6

Effect of different co‐solvents on the enantioselectivity of HEwT for the production of THF amine from THF‐ketone (THF=O) in biotransformations containing THF‐ketone (10 mM), SMBA (1 eq.), purified HEwT (0.25 mg/mL), PLP (0.1 mM) in KP _i ‐buffer (50 mM); pH 8, 30 °C with varying co‐solvents (1.5 M). Samples taken after 3 h. The ee values determined by chiral RP‐HPLC, after FMOC‐derivatization. Error bars represent standard deviations (n=2). Signed ee values; +ve (S)‐enantiomer, ‐ve (R)‐enantiomer. Also shown are the log P and Conolly solvent excluded volume of each solvent, as reported by ChemDraw v. 20.0.