Stereochemical Control of Enzymatic Carbon–Carbon Bond‐Forming Michael‐Type Additions by “Substrate Engineering”

Abstract

The enzyme 4‐oxalocrotonate tautomerase (4‐OT) promiscuously catalyzes the Michael‐type addition of acetaldehyde to β‐nitrostyrene derivatives to yield chiral γ‐nitroaldehydes, which are important precursors for pharmaceutically active γ‐aminobutyric acids. In this study, we investigated the effect of different substituents at the aromatic ring of the Michael acceptor on the catalytic efficiency and stereoselectivity of the 4‐OT‐catalyzed acetaldehyde addition reactions. Highly enantioenriched (R)‐ and (S)‐γ‐nitroaldehydes and 4‐substituted chroman‐2‐ol could be obtained in good to excellent yields by applying different substituents at appropriate positions of the aromatic substrate. Stereochemical control of these enzymatic Michael‐type additions by “substrate engineering” allowed the enantioselective synthesis of valuable γ‐aminobutyric acid precursors. In addition, the results suggest a novel enzymatic synthesis route towards precursors for chromans and derivatives, which are valuable scaffolds for preparing biologically active natural products.

Introduction

γ‐Aminobutyric acid (GABA) is the primary inhibitory neurotransmitter widely distributed in the mammalian central nervous system to regulate neuronal excitation.1 GABA deficiency has been associated with several neurodegenerative disorders such as Parkinson's disease,2 Huntington's disease,[2b], 3 and Alzheimer's disease4 in addition to psychiatric disorders including depression,5 alcoholism,6 and anxiety.[5b], [5c], 7 Over the past few decades, a number of GABA derivatives showing biological activities have been developed as potential drugs for the treatment of neurological disorders.8, 9 It has been shown that the biological activities of these compounds are largely dependent on the configuration of the chiral center. For example, only the (S) enantiomer of vigabatrin,10 the (R) enantiomer of baclofen,11 and the (S) enantiomer of pregabalin12 are the active pharmaceutical ingredients. The asymmetric synthesis of GABA derivatives has therefore attracted considerable attention, and several methodologies involving the use of metal catalysts, organocatalysts, and biocatalysts, or their combinations, to construct these chiral molecules have been established.[8a]

We recently reported a biocatalytic approach for the asymmetric Michael‐type addition of acetaldehyde to nitroalkenes to afford enantioenriched γ‐nitroaldehydes,13 which can be readily converted into chiral GABA derivatives by two simple chemical steps. This methodology utilizes the catalytic promiscuity of the enzyme 4‐oxalocrotonate tautomerase (4‐OT).14 In this study, we investigated the effects of different substituents at the aromatic ring of the Michael acceptor on the catalytic efficiency and stereoselectivity of 4‐OT‐catalyzed Michael‐type addition reactions. We report the asymmetric 4‐OT‐catalyzed Michael‐type addition of acetaldehyde (1; Scheme 1) to a series of monosubstituted β‐nitrostyrene derivatives (see compounds 2c–m) to yield enantioenriched (R)‐ or (S)‐γ‐nitroaldehydes (see compounds 3c–m), as well as the 4‐OT‐catalyzed Michael‐type addition of 1 to (E)‐2‐hydroxy‐β‐nitrostyrene (2b), which, after cyclization, affords 4‐(nitromethyl)chroman‐2‐ol (5; Table 1).

Scheme 1.

Michael‐type addition of acetaldehyde (1) to β‐nitrostyrene derivatives 2a–m.

Table 1.

Preparative‐scale asymmetric 4‐OT‐catalyzed Michael‐type additions of acetaldehyde 1 (50 equiv.) to β‐nitrostyrene derivatives 2a–n (1.5–2.0 mm) in NaH₂PO₄ buffer (pH 5.5) to yield γ‐nitroaldehydes 3a–n or chroman‐2‐ol 5

Yield of isolated product.

Determined by HPLC analysis with a chiral stationary phase.

Absolute configurations of the major enantiomers of 3c–m were determined by optical rotation measurements and comparison with literature data (see Table S3).

Compared to β‐nitrostyrene derivatives 2a–n.

Previously reported result; reaction performed with 1 (50 mm), 2a (1.3 mm), 4‐OT (73 µm) in phosphate buffer (pH 7.3).[13a]

Diastereomeric ratio (dr) was determined by ¹H NMR spectroscopy and comparison with literature data;15 enantiomeric excess (ee) of 5 was not determined.

n.d. = not determined.

Previously reported result; reaction performed with 1 (65 mm), 2g (1.3 mm), 4‐OT (36 µm) in phosphate buffer (pH 5.5).[13c]

Previously reported result; reaction performed with 1 (50 mm), 2n (2 mm), 4‐OT (28 µm) in phosphate buffer (pH 5.5).[13b]

Results and Discussion

The 4‐OT‐catalyzed reaction of donor 1 with nitrostyrene derivatives 2b–l as potential acceptors (Scheme 1) was examined. In separate analytical‐scale experiments, 2b–l and 1 were incubated with 4‐OT in NaH₂PO₄ buffer containing an appropriate co‐solvent (Table S1 in the Supporting Information). The reactions were monitored by recording the change in the absorbance at λ _max of 2b–l by UV spectroscopy. During all reactions, a decrease in the absorbance of 2b–l was observed over the course of time, which indicated depletion of these nitro olefins (Figures S1–S10). Incubation of 1 with 2b–l under the same conditions but in the absence of 4‐OT showed a negligible decrease in the absorbance at λ _max of 2b–l, which indicated that 4‐OT was responsible for the catalysis (Figures S1–S10).

Next, preparative‐scale experiments were performed to allow product identification by NMR spectroscopy and hence to confirm that the 4‐OT‐catalyzed additions of 1 to 2b–l gave Michael‐type addition adducts 3b–l. In addition, the enzymatic preparation of 3m from 2m and 1 was investigated. In separate experiments, 2b–m, 1, and 4‐OT were incubated in NaH₂PO₄ buffer with an appropriate co‐solvent (Table 1, Table S2). The progress of each reaction was monitored by UV spectroscopy (Figures S11–S14). Upon completion of the reaction, workup and purification procedures were performed to isolate the enzymatic products. Analysis of the purified products by NMR spectroscopy confirmed the formation of γ‐nitroaldehydes 3c–m (Figures S16–S25). Intriguingly, the expected γ‐nitroaldehyde 3b (from 1 and 2b) was not observed. Instead, the formation of 4‐(nitromethyl)chroman‐2‐ol (5) as a result of hemiacetalization of 3b was established by NMR spectroscopy (Figure S15) and comparison with literature data.15 4‐Substituted chroman‐2‐ol 5 is a synthetically valuable intermediate that can be easily converted into chromans and their derivatives with antimicrobial and anticancer properties.16 Good to excellent yields (up to 96 %) were achieved for products 3c–f, 3h–i, 3m, and 5 (Table 1). The lower yields of 3j–l (31–37 %) were caused by the low conversion rates with nitrostyrene derivatives 2j–l and the non‐enzymatic hydration of 2j–l in aqueous reaction media.

Enzymatically obtained γ‐nitroaldehydes 3c–m were reduced to the corresponding alcohols 4c–m for determination of their enantiomeric excess (ee) values by chiral stationary phase HPLC. Good to excellent ee values (between 82 and 97 %) were achieved for 3c–m, which indicated that 4‐OT was highly stereoselective during the catalytic process (Table 1; Figures S46–S55). The absolute configurations of 3c–m were determined by optical rotation measurements. Interestingly, comparison with literature data revealed that the meta‐ and para‐substituted γ‐nitroaldehydes (i.e., compounds 3d, 3f, 3g, 3i, and 3j) had the (S) configuration, whereas ortho‐substituted γ‐nitroaldehydes 3e and 3k had the (R) configuration (Table 1; Table S3). The observed negative optical rotation of meta‐substituted γ‐nitroaldehydes 3c, 3l, and 3m could not be compared with literature data, as, to the best of our knowledge, these compounds have so far not been reported. We assume that the major enantiomers of 3c, 3l, and 3m have the (S) configuration, because they showed the same negative optical rotation and elution order in the HPLC chromatograms as those observed with meta‐substituted products 3f and 3i. Similarly, the absolute configuration of the major enantiomer of 3h was tentatively assigned as (R) by comparing its optical rotation data with those of 3e and 3k.

Kinetic studies of the 4‐OT‐catalyzed addition of 1 to 2a–o (2o: R = para‐fluoro) were performed to elucidate the influence of substituents on the kinetic parameters. To address the influence of the co‐solvent and pH on the kinetic parameters, the kinetic assays of the 4‐OT‐catalyzed reaction with (E)‐β‐nitrostyrene (2n) were performed at different pH values and in different solvent systems (Table 2, Entries 1–4). We previously reported that lowering of the pH of the reaction media from 7.3 to 5.5 gave an approximately fourfold increase in both the turnover rate (k _cat) and Michaelis constant (K _M), which resulted in an unchanged catalytic efficiency (k _cat/K _M) (Table 2, Entries 1 and 2). Interestingly, switching the co‐solvent of the reaction medium from EtOH (10 %, v/v) to DMSO (40 %, v/v) also resulted in a slight increase (up to twofold) in both k _cat and K _M (Table 2, compare Entries 3 and 4 with Entries 1 and 2). The kinetic data obtained with 2a in two different solvent systems showed similar changes in the k _cat and K _M values (Table 2, Entries 5 and 6), which suggests that this solvent effect on the kinetic parameters is consistent for the different nitro olefins. This effect may be caused by a small conformational change in the structure of 4‐OT induced by the DMSO co‐solvent.17

Table 2.

Apparent kinetic parameters for the 4‐OT‐catalyzed Michael‐type addition of acetaldehyde (1) to para‐, meta‐, and ortho‐substituted β‐nitrostyrenes 2a–o

Previously reported data; assays were performed in EtOH/phosphate buffer (10:90, v/v) at pH 7.3.[13a]

Previously reported data; assays were performed in EtOH/phosphate buffer (10:90, v/v) at pH 5.5.[13c]

Assays were performed in DMSO/phosphate buffer (40:60, v/v) at pH 7.3.

Assays were performed in DMSO/phosphate buffer (40:60, v/v) at pH 5.5.

Assays were performed in DMSO/phosphate buffer (40:60, v/v) at pH 5.5 to prevent nonenzymatic hydration of nitrostyrene derivatives 2b, 2e, 2h, and 2k in the aqueous solvent system.

The kinetic parameters of the 4‐OT‐catalyzed additions of 1 to 2a–o were determined and compared with those of 2n (R = H) measured under the same solvent and pH conditions (Table 2). In the case of para‐substituted nitrostyrenes, an increase in k _cat for substrates carrying electron‐donating groups (e.g., hydroxy and methoxy groups) and a decrease in k _cat for the substrate with a strongly electron‐withdrawing group (e.g., nitro group) were observed (Table 2, compare Entries 6–10 with Entry 3). The k _cat value seemed not to be significantly affected by fluoro and chloro substituents. The obtained K _M values suggest that the affinity of 4‐OT for para‐substituted nitrostyrenes is affected by the hydrophilicity of the substituents. Substituents with high hydrophilicity generally resulted in increased K _M values, which indicates a lower affinity of 4‐OT towards the substrate (Table 2, compare Entries 6, 8, and 9 with Entry 3). The highest K _M values were obtained with hydroxy‐ and fluoro‐substituted nitrostyrenes (Table 2, Entries 6 and 8).

Results obtained with meta‐substituted nitrostyrenes showed similar electronic effects caused by the substituents. The presence of electron‐donating groups (e.g., hydroxy and methoxy groups) enhanced the k _cat values, whereas the presence of a strongly electron‐withdrawing group (e.g., nitro group) showed the opposite effect. Furthermore, the affinity of 4‐OT towards meta‐substituted nitrostyrenes seemed to be influenced by the hydrophilicity of the substituents as well, as the bromo‐ and chloro‐substituted nitrostyrenes showed the lowest K _M values (Table 2, compare Entries 13 and 14 with Entry 3), whereas a more hydrophilic substituent (e.g., hydroxy group) at the meta position led to an increased K _M value (Table 2, compare Entry 11 with Entry 3). The K _M values obtained with meta‐substituted nitrostyrenes were generally somewhat lower than those determined for the corresponding para‐substituted substrates, which suggests that nitrostyrene derivatives with a meta substitution fit better to the active site of 4‐OT. The relatively low K _M values obtained with meta‐substituted nitrostyrenes also resulted in higher k _cat/K _M values than that obtained with 2n (Table 2, compare Entries 11–15 with Entry 3).

The results of the kinetic assays with ortho‐substituted substrates show that introducing a methoxy group into the ortho position of nitrostyrene significantly increased the value of k _cat, whereas the presence of a hydroxy group at this position resulted in a significant decrease in k _cat (Table 2, compare Entries 16 and 17 to Entry 4). The different K _M values observed for the different ortho‐substituted nitrostyrenes may be related to steric effects.

Conclusion

4‐OT was shown to catalyze the asymmetric Michael‐type addition of acetaldehyde to various ortho‐, meta‐, and para‐substituted β‐nitrostyrenes. For most of the γ‐nitroaldehydes, excellent yields (up to 96 %) and high optical purities (ee values up to 97 %) were achieved. Enzymatic access to (R)‐γ‐nitroaldehydes was achieved with ortho‐substituted nitrostyrenes, whereas (S)‐γ‐nitroaldehydes were obtained with para‐ and meta‐substituted nitrostyrenes. Apparently, attaching substituents to the ortho position of the aromatic substrate induced steric effects, which caused either substrate repositioning in the active site of 4‐OT or a stereofacial shielding effect. Introducing electron‐donating groups at the meta and para positions of the nitrostyrene derivatives significantly enhanced the catalytic rates (up to fivefold improvement in k _cat). Together with the ongoing enzyme engineering studies in our group, this “substrate engineering” work is an important step towards our aim of developing novel “Michaelases” for carbon–carbon bond‐forming Michael‐type addition reactions with high efficiency and stereoselectivity. Finally, the 4‐OT‐catalyzed reaction between 1 and 2b did not give γ‐nitroaldehyde 3b as the final product; instead, it yielded 4‐(nitromethyl)chroman‐2‐ol (5) as a result of hemiacetalization between the aldehyde and hydroxy groups of presumed intermediate 3b. This Michael‐type addition/cyclization cascade reaction provides exciting options for the enzymatic synthesis of precursors for chromans and derivatives, which are valuable scaffolds for preparing biologically active natural products.

Supporting Information (see footnote on the first page of this article): Experimental procedures and characterization of the compounds.

Supporting information

Supporting Information