Study of Chiral Center Effect on CaLB-Catalyzed Hydrolysis of (±)-1-(Acetoxymethyl)-3, 4, 5-methylpyrrolidin-2-ones

Abstract

Several chemical and biocatalytic methods have been described for chiral γ-lactams syntheses. However, only one biocatalytic method has been reported for γ⁴-lactam resolution, while γ²- and γ³-lactams have not been reported. On the other hand, its resolution through biocatalysts is complicated since enzymes such as ENZA-1 (Rhodococcus equi NCIB 40213) and ENZA-20 (Pseudomonas solanacearum NCIB 40249) are difficult to obtain. Therefore, in this paper, the resolution of γ-lactams 7-9 was carried out through a hydrolysis reaction using the commercially available enzyme CaLB.

1. Introduction

γ-lactams are scaffolds of biologically active chiral compounds and drugs, such as the proteasome inhibitor Lactacystin [1], the antidepressant drug (R)-Rolipram [2], and the respiratory stimulant Doxapram [3] (Figure 1).

Furthermore, its ring-open derivatives (Figure 2) are necessary for protein synthesis, so its deficiency is associated with diseases [4]. Deficiency of the γ-amino acid (GABA) is associated with several neurological disorders, such as schizophrenia [5], attention-deficit/hyperactivity disorder [6], and major depressive disorder [7]. Likewise, they are key intermediaries to access a variety of natural products and drug molecules; for example, (S)-Pregabalin is an anticonvulsant with anxiolytic and analgesic properties [8], while (R)-Baclofen is a muscle relaxant [9].

Due to these properties, tremendous efforts have been dedicated to the preparation of chiral γ-lactams, through chemocatalysis as well as biocatalysis, among others [10]. For example, within chemocatalysis, the Pd-catalyzed asymmetric conjugate addition of arylboronic acids to α,β-unsaturated γ-lactams, represents a simple and efficient strategy for obtaining chiral γ-lactams [11]. Zhang et al. [12] used a Rh catalytic system for the asymmetric synthesis of chiral β-substituted γ-lactams through the asymmetric hydrogenation of α,β-substituted β-lactams. Another method is reductive amination of ketoacids/esters followed by intramolecular cyclization, which is considered one of the most attractive approaches to access chiral γ-lactams. Yin et al. [13] carried out an asymmetric reductive amination cascade reaction catalyzed by Ru and its cyclization, providing enantioenriched γ-lactams with up to 97% enantiomeric excess. However, this method employs heavy metals and contaminants such as Rh and Ru, which are not abundant in nature and are also expensive.

On the other hand, the methods for its chemical synthesis are generally based on the formation of C-C bonds through the Michael reaction. In this case, Zlotin et al. [14] carried out the asymmetric Michael addition reaction between nitroalkenes and 1,3-dicarbonyl compounds catalyzed by Ni(II). Another method used is the separation of the racemic mixture by liquid chromatography with a chiral stationary phase (chiral HPLC) [15]. Likewise, Biocatalytic amination is another feasible method for the synthesis of chiral γ-lactams. For example, Kroutil et al. [16] carried out the asymmetric amination of aldehydes with a library of wild-type transaminases (TA), obtaining γ-aminoesters, and subsequently carried out ring closure using basic medium to obtain chiral γ-lactams.

On the other hand, enzymatic resolution is a methodology also used to obtain chiral γ-lactams. Rousseau et al. [17] carried out the enzymatic resolution of γ-butyrolactams, through a transesterification with vinyl acetate catalyzed by porcine pancreatic lipases and Pseudomona cepacia. However, they were only able to resolve the γ⁴-lactam and not the resolution of the γ²-lactam. Also, Fülop et al. [18] carried out enzymatic resolutions for abacavir intermediate with CaLB through ring cleavage obtaining good enantiomeric selectivities. As described previously, the methods reported in the literature have several disadvantages, such as the use of metal catalysts, which are expensive with low abundance in nature, in addition to polluting the environment and having long reaction times. In the literature so far, only one method for biocatalytic γ⁴-lactam resolution has been reported, while γ²- and γ³-lactams have not been reported.

Also, the resolution of γ-lactams through biocatalysts is complicated since enzymes such as ENZA-1 (Rhodococcus equi NCIB 40213) and ENZA-20 (Pseudomonas solanacearum NCIB 40249) [19] are difficult to obtain. In this context, Candida antarctica lipase B (CaLB) is commercially available at a low cost, both in its free and immobilized form, with the latter being easy to use and recycle. It has also been the most used lipase in recent years, due to its characteristics such as catalytic promiscuity, thermostability, high activity in organic solvents, as well as its regio-, chemo-, and enantioselectivity [20]. For us, it is of utmost importance to corroborate whether the CaLB enzyme is capable of resolving γ²-, γ³-, and γ⁴-lactams, which is why in this work the resolution was carried out through a hydrolysis reaction (Figure 3).

2. Results and Discussion

2.1. γ-Lactams Syntheses

γ²-, γ³, and γ⁴-lactams (±)-1-3, were obtained using the methodology described by us [21]. Subsequently, a hydroxymethylenation reaction was carried out using the methodology described by Giacomini et al. [22], with 0.04 eq of potassium bicarbonate, 3 eq of paraformaldehyde, and water refluxed in THF over 3 h, obtaining the N-substituted γ-lactams (±)-4-6, with yields of 60, 96, and 95%, respectively. Subsequently, its acylation was carried out using 1.2 eq of acetic anhydride and 1 eq of DMAP in toluene at reflux over 1 h, obtaining the products (±)-7-9 with yields of 90, 30, and 97%, respectively (Scheme 1).

2.2. Enzymatic Resolution of γ-Lactams (±)-7-9

Once γ-lactams (±)-7-9 were synthesized, γ⁴-lactam (±)-7 was used as a model for enzymatic resolution, because it was expected to show greater recognition by the enzyme as the stereogenic center would be closer to the catalytic site. Therefore, the optimal conditions for its resolution were sought. First, an enzymatic hydrolysis reaction was carried out using the CaLB, water as the nucleophile, and 2-methyl-2-butanol (2M2B) as a solvent (Scheme 2, entry 1,

Table 1. Optimization of the enzymatic resolution for compound (±)-7.

Entry	CaLB	H2O (eq)	Temp (°C)	Time (h)	NMR Conversion 1	(S-)-4 (%)	e.e 2	E 3
1	5%	3	40	2	25	30	74	8
2	5%	3	40	4	44	48	81	18
3	10%	3	40	2	40	40	82	17
4	5%	3	27	4	32	33	86	20
5	5%	1.5	27	4	29	29	85	17

¹ Obtained after enzyme filtration and solvent distillation. ² Chiral HPLC, OJ-H chiral column, elution system Hex:i-PrOH 98:2, flux 1.0 mL min⁻¹, λ = 210 nm. ³ Calculated using $E=ln\left[1-c\left(1+{ee}_p\right)\right]/ln\left[1-c\left(1-{ee}_p\right)\right]$.

). Variations in temperature, nucleophile equivalents, as well as enzyme concentration results, are shown in Table 1.

The ratio between the product and starting material was determined by ¹H NMR of crude oil (Figure 4). Crude oil was purified by a gradient chromatographic column to obtain the isolated compound. Optical rotation was taken as +7.1°, thus allowing us to correlate it with (S) enantiomer, which is reported in the literature [17]. Likewise, enantiomeric excess was determined by chiral HPLC.

As can be seen from Table 1, the best conditions to obtain γ⁴-lactam (S)-4 with high e.e. are those from entry 4, in which the reaction benefits from a longer reaction time and higher water equivalents.

Once the optimal conditions for the resolution of the γ⁴-lactam (±)-7 were found, they were used to carry out enzymatic resolution on γ³-lactam (±)-8. However, the enzyme had poor selectivity according to the enantiomeric excess calculated (Scheme 3,

Table 2. Enzymatic resolution for γ³-lactam (±)-8.

Entry	Time (h)	NMR Conversion 1	(±)-8 (%)	(±)-5 (%)	e.e 2	E 3
1	1	37	60	36	1	1
2	1.17	42	55	40	1	1
3	1.34	62	37	60	0.4	1

¹ Obtained after enzyme filtration and solvent distillation. ² Chiral HPLC, OJ-H chiral column, elution system Hex:i-PrOH 98:2, flux 1.0 mL min⁻¹, λ = 210 nm. ³ Calculated using $E=ln\left[1-c\left(1+{ee}_p\right)\right]/ln\left[1-c\left(1-{ee}_p\right)\right]$.

).

On the other hand, the resolution of γ²-lactam (±)-9 was carried out, observing moderate recognition by the enzyme. The optical rotation was taken as −10.7°, and the enantiomeric excess and enantiomeric ratio (E) were calculated (Scheme 4,

Table 3. Enzymatic resolution for compound (±)-9.

Entry	Time (h)	NMR Conversion 1	(-)-9 (%)	(-)-6 (%)	e.e 2	E 3
1	0.75	30	68	28	65	6
2	1	37	61	36	57	1

¹ Obtained after enzyme filtration and solvent distillation. ² Chiral HPLC, OJ-H chiral column, elution system Hex:i-PrOH 98:2, flux 1.0 mL min⁻¹, λ = 210 nm. ³ Calculated using $E=ln\left[1-c\left(1+{ee}_p\right)\right]/ln\left[1-c\left(1-{ee}_p\right)\right]$.

).

Table 4. Enzymatic resolution for γ-lactams (±)-9.

Lactam	Temp. (°C)	Time (h)	NMR Conversion	Acetoxy- (%)	Hidroxy- (%)	e.e	E
γ4	27	4	32	67	33	86	20
γ3	27	1	37	60	36	1	1
γ2	27	1	30	61	36	65	6

compares the best results obtained for each N-substituted γ²-, γ³-, and γ⁴-lactams (±)-7-9. As can be seen, a change in the position of the methyl group is important because it affects enzymatic recognition. Furthermore, these results are of great importance since the resolution by hydrolysis of γ-lactams has not been reported so far.

To explain these results, in silico calculations were performed employing Molegro Virtual Docker (See Supplementary Materials). In these calculations, it is possible to know the affinity enzyme-substrate through interactions with amino acid residues, and substrate orientation within the catalytic site. For γ⁴-lactam (±)-7, enantiomers (S) and (R) were compared. The in silico results show that γ-lactam (R) has unfavorable bumps with Ser105 because of the orientation occupied by the acyl group in the catalytic site (Figure 5A), while γ-lactam (S) does not present unfavorable bumps since the acyl group adopts a better conformation inside the catalytic site (Figure 5B). The orientation of both carbonyl groups seems to play an important role in recognition by the enzyme. In the case of (S), both carbonyls are oriented towards the interior of the active site, generating hydrogen bonds with Gln157, Thr40, Ser105, and His224, which are amino acids of the pocket acyl and the catalytic triad, thus enabling better recognition.

For γ³-lactam (±)-8 enantiomers with the enzyme, it can be seen that the (R) enantiomer has a slightly better recognition because it forms hydrogen bonds with Thr40, His224, and Ser105, and does not have unfavorable bumps (Figure 6A).

In contrast, its (S) enantiomer presents hydrogen bonds with His224, and Ser105 also presents an unfavorable bump with Gln157 (Figure 6B). However, the interaction difference is not very significant, and experimentally, low selectivity is observed.

On the other hand, in γ²-lactam (±)-9, the (S) enantiomer has better recognition due to the presence of a greater number of favorable interactions as hydrogen bondings with His224, Ser105, and Thr40 (Figure 7B). In contrast, the (R) enantiomer presents unfavorable bumps with Ile189 and Ile285 and no hydrogen bonding (Figure 7A), which results in better selectivity towards the (S) enantiomer than the (R).

These in silico results are consistent with the experimental results obtained.

3. Materials and Methods

The lipase Novozym^® 435 (Novozymes, Mexico City, Mexico) was used as the catalyst for all enzymatic reactions. Novozym^® 435 is the commercially available form of the recombinant Candida antarctica lipase B, immobilized by adsorption in a hydrophobic microporous acrylic resin. Reactions were monitored by thin-layer chromatography (TLC) and carried out on 0.25 mm Merck (Rahway, NJ, USA) silica gel plates (60F-254) using UV light as the visualizing agent. Column chromatography was performed on Merck silica gel 60 (0.040–0.063 mm). ¹H and ¹³C NMR spectra were recorded on Varian Gemini 200 MHz equipment. Chemical shift δ values (ppm) are reported relative to the internal tetramethylsilane (TMS) reference (δ = 0.0 ppm) for determinations in CDCl₃. Microwave reactions were performed in sealed vessels in a CEM (Matthews, NC, USA) Discover apparatus.

Molecular Docking.

Structures were created in Spartan 14 [23] and then submitted to conformational analysis within molecular mechanics level theory and an MMFF force field. Conformers of minimum energy were selected and optimized within DFT level theory using B3LYP [24,25] and the 6-31G* basis set. Molecular docking calculations were carried out in Molegro Virtual Docker (MVD) [26], using the Cal-B structure PDB: 1LBS [27]. All docking images were made with Discovery Studio Visualizer [28]. See Supplementary Materials for more details.

Syntheses of γ-lactams (±)-1-3; General Procedure 1 (GP1).

γ-lactams (±)-1-3 were obtained by preparing the γ-nitroaliphatic methyl esters under microwave irradiation, as has been reported [20]. Then, the reduction of the nitro group using Pd/C [29] yielded the corresponding γ-lactams (±)-1-3, resulting from a tandem cyclization of the free amines.

γ-Nitroaliphatic methyl esters synthesis.

A solution of α,β-unsaturated ester (methyl methacrylate, methyl crotonate, and methyl acrylate), and the corresponding nitroalkane (nitromethane or nitroethane) in DBU or TMG were added to a 10 mL glass microwave reaction vessel containing a stir bar. The reaction vessel was sealed with a cap and placed into the microwave cavity. The microwave unit was programmed to 70–75 °C with a power of 50 Watts over 5 min. After the reaction was completed and the vessel was cooled to below 50 °C using a compressed air flow, the crude material was purified by column chromatography to give the nitro methyl esters.

Nitro group reduction.

Ammonium formate (8 eq) was added to a solution of nitro methyl esters (1 eq), with Pd/C in MeOH and the mixture was stirred overnight. The solution is then filtered with a vacuum and Pd/C is washed with absolute methanol (10 mL). γ-lactams (±)-1-3 were purified by flash column chromatography.

Synthesis of N-hydroxymethylated γ-lactams (±)-4-6; General Procedure 2 (GP2).

In a 50 mL round flask equipped with a magnetic stirrer, the corresponding γ-lactam dissolved in THF was placed. Subsequently, potassium bicarbonate (0.04 eq), paraformaldehyde (3 eq), and HPLC water (20.1 mmol) were added. The reaction was refluxed over 3 h. After obtaining the corresponding N-hydroxymethylated γ-lactam, it was purified by column chromatography.

Synthesis of N-Acetoxymethyl γ-lactams (±)-7-9; General Procedure 3 (GP3).

In a 50 mL round flask equipped with a magnetic stirrer, the corresponding N-hydroxymethylated γ-lactam dissolved in toluene was placed. Subsequently, acetic anhydride (1.2 eq) and DMAP (1 eq) were added, and the reaction mixture was brought to reflux for 1 h. The corresponding N-substituted γ-lactam was purified by column chromatography.

(rac)-5-Methylpyrrolidin-2-one, (±)-1. Prepared according to GP1, methyl acrylate (20 mmol, 1.81 mL) was mixed with nitroethane (25 mmol, 1.78 mL), DBU (1 mmol, 0.15 mL), and irradiated with microwaves over 5 min at 70 °C and 50 W potency. Crude was purified (Hex:AcOEt 90:10 → 70:30) to obtain a colorless oil, yield: 1.61 g (50%). Methyl nitroester (12.4 mmol, 2 g) was placed in a 250 mL round flask, and MeOH (75 mL), Pd/C (6% p/p, 0.12 g), and ammonium formate (99.2 mmol, 6.25 g) were added. The mixture was stirred overnight, crude was filtered, rotaevaporated, and after purification by flash chromatographic column (Hex:AcOEt 90:10 → 60:40), γ-lactam (±)-1 was isolated as a colorless oil. Yield: 1.03 g (84%). ¹H NMR (200 MHz, CDCl₃): δ 1.23 (d, J = 6.2 Hz, 3H, CH₃), 1.60–1.71 (m, 1H, CH₂CH₂CHN), 2.18–2.40 (m, 3H, CH₂CH₂CHN), 3.68–3.91 (m, 1H, CHN). ¹³C NMR (50 MHz, CDCl₃): 178.6, 50.4, 30.8, 29.3, 22.4. Compared with lit [30].

(rac)-4-Methylpyrrolidin-2-one, (±)-2. Prepared according to GP1, methyl crotonate (35 mmol, 3.7 mL) was mixed with nitromethane (87.5 mmol, 4.72 mL) and DBU (1.75 mmol, 0.261 mL) and irradiated with microwaves over 15 min at 75 °C and 50 W potency. Crude was purified (Hex:AcOEt 98:02 → 90:10) to obtain a colorless oil, yield: 2.98 g (53%). Methyl nitroester (12.4 mmol, 2 g) was placed in a 250 mL round flask, MeOH (75 mL), Pd/C (6% p/p, 0.12 g) and ammonium formate (99.2 mmol, 6.25 g) were added. The mixture was stirred overnight, crude was filtered, rotaevaporated, and after purification by flash chromatographic column (Hex:AcOEt 90:10 → 60:40), γ-lactam (±)-2 was isolated as a colorless oil. Yield: 0.92 g (75%). ¹H NMR (200 MHz, CDCl₃): δ 1.13 (d, J = 6.6 Hz, 3H, CH₃),1.94–2.05 (dd, J = 5.0, 15.4 Hz, 1H, CH₂CH), 2.4–2.61 (m, 2H, CH₂CH), 3.18–3.26 (dd, J = 5.8, 7.5 Hz, 1H, CH₂N), 3.69–3.78 (dd, J = 8.2 Hz, 1H, CH₂N). ¹³C NMR (50 MHz, CDCl₃): δ 170.1, 56.4, 37.1, 24.3, 20.5. Compared with lit [22].

(rac)-3-Methylpyrrolidin-2-one, (±)-3. Prepared according to GP1, methtyl methacrylate (20 mmol, 2.1 mL), was mixed with nitromethane (25 mmol, 1.35 mL) and TMG (10 mmol, 1.25 mL) and irradiated with microwaves over 20 min at 70 °C and 50 W potency. Crude was purified (Hex:AcOEt 90:10 → 70:30) to obtain a colorless oil, yield: 1.38 g (43%). Methyl nitroester (12.4 mmol, 2 g) was placed in a 250 mL round flask, and MeOH (75 mL), Pd/C (6% p/p, 0.12 g) and ammonium formate (99.2 mmol, 6.25 g) were added. The mixture was stirred overnight, crude was filtered, rotaevaporated, and after purification by flash chromatographic column (Hex:AcOEt 90:10 → 60:40). γ-lactam (±)-3 was isolated as a colorless oil. Yield: 1.1 g (90%). ¹H NMR (200 MHz, CDCl₃: δ 1.20 (d, J = 6.6 Hz, 3H, CH₃), 1.50–1.70 (m, 1H, CH₂CH), 2.22–2.37 (m, 3H, CH₂CHCH₂N), 3.70–3.86 (m, 1H, CH₂N). ¹³C NMR (50 MHz, CDCl₃): δ 180.6, 64.5, 31.7, 30.2, 30.2. Compared with lit [22].

(rac)-1-(Hydroxymethyl)-5-methylpyrrolidin-2-one (±)-4. Prepared according to GP2, (±)-1 (5.04 mmol, 500 mg), potassium bicarbonate (0.2 mmol, 27.3 mg), paraformaldehyde (15.13 mmol, 454 mg), and water (20.1 mmol, 0.362 mL) were refluxed in THF (20 mL) over 3 h. After chromatographic column (Hex:AcOEt 90:10 → 0:100), γ-lactam (±)-4 was isolated as a colorless oil. Yield: 0.39 g (60%). ¹H NMR (200 MHz, CDCl₃): δ 4.99 (dd, J = 10.9, 6.6 Hz, 1H CH₂OH), 4.62 (dd, J = 8.5, 10.9 Hz, 1H CH₂OH), 3.81 (m, 2H, CHCH₃, OH), 2.42 (m, 2H, CH₂CO), 2.20 (m, 1H, CH₂CH), 1.60 (m, 1H, CH₂CH), 1.29 (d, J = 6.3 Hz, 3H, CH₃). ¹³C NMR (50 MHz, CDCl₃): δ 176.6, 65.0, 53.2, 30.7, 27.1, 20.4.

(rac)-1-(Hydroxymethyl)-4-methylpyrrolidin-2-one (±)-5. Prepared according to GP2, (±)-2 (11 mmol, 1.09 g) potassium bicarbonate (0.4 mmol, 60.6 mg), paraformaldehyde (33 mmol, 0.99 g), and water (40.2 mmol, 0.788 mL) were refluxed in THF (40 mL) over 3 h. After chromatographic column (Hex:AcOEt 90:10 → 0:100), γ-lactam (±)-5 was isolated as a colorless oil. Yield: 1.24 g (96%). ¹H NMR (200 MHz, CDCl₃): δ 4.74 (AB, J = 10.6 Hz, 2H, CH₂OH), 3.68 (dd, J = 7.7, 9.6 Hz, 1H, CH₂NH), 3.16 (dd, J = 5.9, 9.6 Hz, 1H, CH₂NH), 2.50 (m, 2H, CHCH₃, CH₂CO), 2.03 (m, 1H, CH₂CO), 1.14 (d, J = 6.6 Hz, 3H, CH₃). ¹³C NMR (50 MHz, CDCl₃): δ 175.8, 66.3, 53.4, 37.0, 39.7, 26.5, 19.6.

(rac)-1-(Hydroxymethyl)-3-methylpyrrolidin-2-one (±)-6. Prepared according to GP2, (±)-3 (1 mmol, 100 mg), potassium bicarbonate (0.4 mmol, 5.6 mg), paraformaldehyde (3 mmol, 90.8 mg), and water (72.5 µL) were refluxed in THF (12.5 mL) over 3 h. After chromatographic column (Hex:AcOEt 90:10 → 0:100), γ-lactam (±)-6 was isolated as a colorless oil. Yield: 0.12 g (95%). ¹H NMR (200 MHz, CDCl3): δ 4.8 (d, J = 10.7 Hz, 1H CH₂OH), 4.68 (d, J = 10.7 Hz, 1H CH₂OH), 3.49 (dd, J = 8.6, 1.5 Hz, 1H CH₂CH₂CH), 3.46 (d, J = 8.5 Hz, 1H CH₂CH₂CH), 2.46 (m, 1H, CH₂CH), 2.23 (m, 1H, CHCH₃), 1.62 (m, 1H, CH₂CH), 1.16 (d, J = 7.1 Hz, 3H, CH₃). ¹³C NMR (50 MHz, CDCl₃): δ 178.8, 66.8, 44.4, 37.3, 27.3, 16.1.

(rac)-1-(Acetoxymethyl)-5-methylpyrrolidin-2-one (±)-7. Prepared according to GP3, (±)-4 (2.5 mmol, 300 mg), acetic anhydride (2.82 mmol, 0.26 mL), and DMAP (2.35 mmol, 287 mg) were refluxed in toluene (15 mL) over 1 h. After chromatographic column (Hex:AcOEt 90:10 → 0:100), γ-lactam (±)-7 was isolated as a colorless oil. Yield: 0.36 g (90%). ¹H NMR (200 MHz, CDCl₃): δ 5.34 (AB, J = 10.5 Hz, 2H, CH₂O), 3.80 (m, 1H, CHCH₃), 2.42 (m, 2H, CH₂CO), 2.24 (m, 1H, CH₂CH), 2.07 (s, 3H, CH₃CO), 1.62 (m, 1H, CH₂CH), 1.29 (d, J = 6.3 Hz, 3H, CH₃CH). ¹³C NMR (50 MHz, CDCl₃): δ 176.26, 170.8, 64.7, 53.1, 29.9, 26.8, 20.8, 19.9.

(rac)-1-(Acetoxymethyl)-4-methylpyrrolidin-2-one (±)-8. Prepared according to GP3, (±)-5 (3.93 mmol, 390 mg), acetic anhydride (11.93 mmol, 1.12 mL), and DMAP (9.9 mmol, 1.21 g) were refluxed in toluene (30 mL) over 1 h. After chromatographic column (Hex:AcOEt 90:10 → 0:100), γ-lactam (±)-8 was isolated as a colorless oil. Yield: 0.15 g (30%). ¹H NMR (200 MHz, CDCl₃): δ 5.27 (AB, J = 10.0 Hz, 2H, CH₂O), 3.62 (dd, J = 9.6, 7.7 Hz, 1H, CH₂N), 3.08 (dd, J = 9.6, 6.3 Hz, 1H, CH₂N), 2.48 (m, 2H, CHCH₃, CH₂CO), 2.05 (s, 3H, CH₃CO), 1.96 (m, 1H, CH₂CO), 1.11 (d, J = 6.6 Hz, 3H, CH₃CH). ¹³C NMR (50 MHz, CDCl₃): δ 175.9, 171.0, 66.6, 54.4, 39.1, 26.8, 20.8, 19.48.

(rac)-1-(Acetoxymethyl)-3-methylpyrrolidin-2-one (±)-9. Prepared according to GP3, (±)-6 (0.82 mmol, 106 mg), acetic anhydride (0.986 mmol, 93 µL), and DMAP (0.82 mmol, 100 mg) were refluxed in toluene (12 mL) over 1 h. After chromatographic column (Hex:AcOEt 90:10 → 0:100) γ-lactam (±)-9 was isolated as a colorless oil. Yield: 0.13 g (97%). ¹H NMR (200 MHz, CDCl₃): δ 5.33 (AB, J = 10.3 Hz, 2H, CH₂O), 3.45 (m, 2H, CH₂CH₂CH), 2.51 (m, 1H, CH₂CH), 2.27 (m, 1H, CHCH₃), 2.08 (s, 3H, CH₃CO), 1.67 (m, 1H, CH₂CH), 1.22 (d, J = 7 Hz, 3H, CH₃CH). ¹³C NMR (50 MHz, CDCl₃): δ 178.47, 170.9, 66.9, 45.0, 36.5, 27.3, 20.7, 15.8.

4. Conclusions

A synthetic approach to γ², γ³, and γ⁴-lactam resolution employing CaL-B has been described. In this methodology, the substituent position is important, as the farther the stereogenic center is from the catalytic site, the lower the recognition is expected. γ⁴-lactams have good recognition and the product can be obtained enantio-enriched, while γ²-lactams can be obtained at moderated enantiomeric ratios. On the other hand, γ³-lactams do not have good selectivity and are obtained as racemic mixtures. Since resolution is performed under innocuous reagents, this approach is an eco-friendly alternative to access enantio-enriched γ⁴-lactams. Chemical correlation allowed us to assign the enantiomer recognized by CaLB as (R), which was supported by the in silico results.