Novel Kinetic Resolution of Thiazolo-Benzimidazolines Using MAO Enzymes

Abstract

The kinetic resolution of racemic 1H,3H-thiazolo[3,4-a]benzimidazoline (TBIM) heterocycles was achieved using E. coli whole cells expressing the MAO-N D11 enzyme. Several cosolvents were screened using TBIM 2a as the substrate. DMF was the best cosolvent, affording the pure enantiomer (+)-2a in 44% yield, 94% ee. The stereochemistry of TBIM was predicted by means of ab initio calculations of optical rotation and circular dichroism spectra. The reaction scope was investigated for 11 substituted (±) TBIM using an optimized protocol. The best yield and % ee were obtained for the nonsubstituted 2a. Among the substituted compounds, the 5-substituted-TBIM showed better % ee than the 4-substituted one. The small electron donor group (Me) led to better % ee than the electron-withdrawing groups (–NO₂ and –CO₂Et), and the bulky naphthyl group was detrimental for the kinetic resolution. Docking experiments and molecular dynamics (MD) simulations were employed to further understand the interactions between MAO-N D11 and the thiazolo-benzimidazoline substrates. For 2a, the MD showed favorable positioning and binding energy for both enantiomers, thus suggesting that this kinetic resolution is influenced not only by the active site but also by the entry tunnel. This work constitutes the first report of the enzymatic kinetic resolution applied to TBIM heterocycles.

Introduction

Heterocyclic ring systems are fundamental building blocks in medicinal chemistry and play an essential role in drug development.¹ Over 90% of new drugs contain at least one heterocyclic fragment in their structures. Among them, nitrogen-containing heterocycles are the most significant scaffolds.²

In recent years, imidazole and its reduced derivatives imidazolidines and imidazolines have been widely used as intermediate species in synthetic medicinal chemistry, leading to biologically active pharmacophores.³ In particular, 2-, 3-, and 4-imidazolines are emerging privileged scaffolds with numerous applications in homogeneous catalysis⁴ and in the surfactant industry (Figure 1).⁵ Additionally, imidazoline-containing compounds have shown a wide range of biological and pharmacological properties, playing a central role in the treatment of numerous types of diseases. Many of the pharmacological applications of imidazoline compounds include blood pressure control and cardioprotection,⁶ hyperglycemia,⁷ and psychiatric disorders treatment.⁸

Figure 1.

Structures of 2-, 3-, and 4-imidazolines and benzimidazolines.

Benzimidazolines are bicycles containing a 4-imidazoline fused with a benzene ring. This family of bicycles are widely used in organic synthesis, as reducing agents or as a hydrogen-storage material in organic reactions.⁹ Furthermore, some derivatives have shown interesting antimicrobial activity.¹⁰ Thus, benzimidazoline-containing compounds are an attractive field to be explored. Our group has recently prepared several racemic benzothiazolidines, and in order to enlarge the chemical diversity of our library, we decided to explore the synthesis of benzimidazolines.

There are a few examples of 1H,3H-thiazolo[3,4-a]benzimidazolines (TBIM) synthesis through a multicomponent reaction using o-phenylenediamine, aryl isothiocyanates, and 4-chloro-3-oxo-butanoate.¹¹ In a previous work, we prepared 1H,3H-thiazolo[3,4-b]benzothiazolines TBTZ 1 by double condensation of 2-aminothiophenol and two molecules of 1,4-dithiane-2,5-diol.¹² Similarly, in this work, we prepare new 1H,3H-thiazolo[3,4-a]benzimidazolines TBIM 2 starting from 1,2-diaminoanilines and dithiane, Figure 2. These heterocycles are obtained as a racemic mixture of the syn diastereomer as the main product, bearing both substituents on the same side of the bicycle plane.¹³

Figure 2.

Synthesis of 1H,3H-thiazolo[3,4-b]benzothiazolines TBTZ 1 and 1H,3H-thiazolo[3,4-a]benzimidazolines TBIM 2.

Since enantiomers can exhibit distinct pharmacokinetic and pharmacological properties, modern drug synthesis endeavors to produce new chiral drugs in the form of a single enantiomer.¹⁴ Therefore, the development of efficient methods for obtaining enantiomerically pure compounds has attracted significant attention.

Several biocatalytic approaches have been described for resolving chiral amines,¹⁵⁻¹⁷ including ω-transaminases,^18,19 amine oxidases,^20,21 P450 monooxygenases,^22,23 amine dehydrogenases,^24,25 imine reductases (IREDs),^26,27 and reductive aminases.²⁸ In recent years, the resolution of racemates containing amines by amino oxidase enzymes (AOs) has received particular attention for the synthesis of chiral amines in industrial biotechnology. Among these biocatalytic processes, monoamine oxidases (MAOs) have emerged as the most widely employed enzymes in biocatalysis. The multienzymatic synthesis of benzylisoquinoline alkaloids was facilitated using MAO enzymes from Micrococcus luteus.²⁹ Nevertheless, MAO enzymes from Aspergillus niger (MAO-N) and its variants have become the most commonly used for such applications.

MAO enzymes facilitate the oxidation of amines to imines using molecular oxygen as the oxidizing agent, as illustrated in Figure 3.^30,31 The native MAO enzymes show poor enantioselectivity, rendering them ineffective in resolving chiral amines. However, by utilizing directed mutagenesis, Turner et al. have developed a set of MAO variants from A. niger, named D3, D5, D9, D10, and D11, which have proven effective for resolving a wide range of chiral primary, secondary, and tertiary amines, Figure 4.²⁰ When a chemical reducing agent such as BH₃–NH₃ is employed, the imines formed during catalysis can be reduced, leading to oxidation–reduction cycles with the enzyme, and thus achieving a dynamic kinetic resolution or deracemization, with yields as high as 100%, as depicted in Figure 3.^30,32

Figure 3.

General scheme for dynamic kinetic resolution or deracemization: MAO-N catalyzes the oxidation of primary, secondary, and tertiary amines, followed by chemical nonselective reduction of imines by BH₃–NH₃.

Figure 4.

Examples of the dynamic kinetic resolution of different substrates using MAO-N variants.

In order to explore a suitable methodology to obtain the enantiomerically pure heterocycles TBTZ 1 and TBIM 2 under mild conditions, we envisioned that the use of MAO enzymes could be a useful approach. To the best of our knowledge, this methodology has not been used so far for the kinetic resolution of benzimidazolines nor benzothiazolines.

Results and Discussion

Synthesis of TBTZ 1 and TBIM 2

Racemic TBTZ 1a and TBIM 2a–g were obtained in a single step by double cyclization of 2-aminothiophenol 3a or 2-aminoanilines 4a–g, respectively, and 1,4-dithiane-2,5-diol 5 in yields ranging from 49 to 85%, Table 1.

Table 1. Synthesis of (±)-TBTZ 1 and (±)-TBIM 2, Starting from 2-Aminothiophenol 3a or 2-Aminoanilines 4a–g, Respectively, and 1,4 Dithiane-2,5-Diol 5.

entry	compound	X	R	total yield (%)a	ratio-4/-5b
1	1a	S	H	8512
2	2a	NH	H	58
3	2b	NH	4-5-naphthyl	71
4	4-2c	NH	–CH3	58	6/4
5	5-2c	NH	–CH3
6	4-2d	NH	–CO2Et	49	3/7
7	5-2d	NH	–CO2Et
8	4-2e	NH	–NO2	83	3/7
9	5-2e	NH	–NO2
10	4-2f	NH	–F	75	7/3
11	5-2f	NH	–F
12	4-2g	NH	–Br	69	7/3
13	5-2g	NH	–Br

^aYields correspond to purified compounds by column chromatography on flash SiO₂.

^bThe ratio was calculated by ¹H NMR.

When using 5-substituted-2-aminoanilines, two possible isomers at the 4- and 5-positions are obtained. After the first cyclization, a benzimidazoline intermediate is formed, bearing two nitrogen atoms able to react with a second molecule of aldehyde. The nature of the R substituent leads to a difference in the nucleophilicity between N¹ and N², determining the 4/5-TBIM ratio. Electron-withdrawing groups decrease the nucleophilicity of N² with respect to N¹, giving the 5-substituted compound (5-TBIM) as the main product, see Table 1, entries 6–9. On the other hand, electron-donating groups increase the nucleophilicity of N² compared to N¹, affording the compound 4-TBIM as the main product, see Table 1, entries 4 and 5.

Analytical-Scale Screening Using E. coli Whole Cells

In order to evaluate the enantioselectivity of the MAO enzyme toward the resolution of TBTZ 1 and TBIM 2, the simplest unsubstituted substrates 1a and 2a were selected as a model to study the optimal deracemization conditions. We hypothesize the formation of the imine oxidation product and the enantiopure TBTZ or TBIM, as shown in Scheme 1.

Scheme 1. Proposal for the Resolution of Heterocycles 1a and 2a Using MAO-N Variants.

Thus, the racemic mixtures were subjected to a deracemization protocol using E. coli whole cells expressing three different MAO-N variants (D5, D9, and D11) to compare the activity and enantioselectivity. Reactions were carried out in phosphate buffer (0.1 M, pH = 7) and the substrate at 4 mM concentration, using BH₃–NH₃ (4 equiv) as the reducing agent. Aliquots were taken at different times and diluted for chiral HPLC analysis. Yields and enantiomeric excess (% ee) were calculated, Table 2. According to literature, the expected isolated enantiomer should be (S,R)-1a and (R,R)-2a since the substrates are voluminous.³⁰

Table 2. Analytical-Scale Screening for the Enzymatic Resolution of TBTZ 1a and TBIM 2a Using Whole Cells (E. coli) Expressing MAO-N Variants: D5, D9, and D11^a.

substrate	time	4 h		48 h
	MAO-N variant	% eeb	yield (%)c	% eeb	yield (%)c
(S,R)-1a	D5	0	46	0	26
	D9	0	47	0	22
	D11	0	42	0	27
(S,S)-2a	D5	14	42	56	11
	D9	28	37	18	9
	D11	90	31	64	2

^aConditions: cells 1 g/mL, phosphate buffer 0.1 M pH 7, [substrate] = 4 mM, BH₃NH₃, DMSO 5%, 37 °C.

^bCalculated by chiral HPLC.

^cCalculated by chiral HPLC using an external standard.

Despite no resolution was found for TBTZ 1a using the three variants, all of them resulted to be enantioselective for TBIM 2a. In particular, the D11 variant showed the best result, providing the optically enriched product in 31% yield with excellent enantiomeric excess (90%) after 4 h, Table 2. A longer reaction time (48 h) decreased both the yield and % ee. This result could be explained if the oxidation rate for (S,S)-2a is higher than that for its enantiomer (R,R)-2a. Over time, when the preferred enantiomer is consumed, the enzyme continues oxidizing the remaining enantiomer, decreasing the % ee.

Attempts to scale-up the resolution reaction with whole cells led to lower % ee, so we decided to employ the purified enzyme MAO-N D11³³ for the resolution of TBIM. Although cleaner reactions were carried out, the optically enriched product was obtained with 46% ee. Thus, we decided to continue optimizing the reaction employing E. coli whole cells since enzyme purification is a time-consuming approach and did not improve the % ee or the yield of the reaction.

Solvent Screening

We hypothesized that the unsuccessful scale up could be due to the slight solubility of the compounds in the reaction media; the organic cosolvent used could have a strong influence in both the yield and the % ee; thus, its effect was studied. The reaction was performed with 2a and E. coli whole cells expressing MAO-N D11 as a biocatalyst, and several organic cosolvents at different concentrations were screened. The results (yield and % ee) are shown in Table 3.

Table 3. Analytical-Scale Screening of Cosolvents for the Enzymatic Resolution of TBIM 2a Using E. coli Whole Cells Expressing MAO-N D11^a.

entry	solvent	time (h)	% eeb	yield (%)c
1	DMSO 5%	2	82	37
2		4	86	23
3		6	84	9
4	DMSO 10%	2	83	34
5		4	85	19
6		6	80	11
7	DMF 5%	2	94	44
8		4	94	26
9		6	94	18
10	MeOH 5%	2	68	41
11		4	82	23
12		6	80	15
13	PhMe 5%	2	68	40
14		4	60	23
15		6	68	12

^aConditions: cells 1 g/mL, phosphate buffer (0.1 M, pH 7), [syn-2a] = 4 mM, co-solvent 5–10%, BH₃NH₃, 37 °C.

^bCalculated by chiral HPLC.

^cCalculated by chiral HPLC using an external standard.

The results indicated a strong influence of the cosolvent on the % ee. The best results were achieved using 5% DMF, affording 2a in 44% yield, 94% ee after 2 h (Table 3, entry 7 and Figure S15A,B). This may be due to a higher solubility of the compound in DMF, increasing the availability of the compound in the reaction medium. Even though longer reaction times improved the % ee, the reaction yields decreased notoriously (Table 3, entries 8 and 9), probably due to chemical oxidation or heterocycle ring-opening in the reaction media.

The lowest % ee were obtained using toluene as cosolvent, probably because a biphasic system is formed and compound 2a remains in the organic layer.

On the other hand, the effect of the additive BH₃NH₃ was evaluated in the kinetic resolution using the optimized conditions (DMF 5% as cosolvent). The results showed that its use smoothly improved the reaction yields, see Table S1. The reductive environment probably avoids thiol oxidation, which could be accelerated by H₂O₂ formation during the MAO-N-catalyzed biotransformation.

Reaction Scale-Up

Once the best reaction conditions were established, the reaction was scaled-up for the resolution of (±)-2a using E. coli whole cells expressing the MAO-N D11 enzyme in buffer at pH 7 and 5% DMF for 2 h. Under these conditions, the (+)-2a enantiomer was isolated in 35% yield and 89% ee, together with compound (−)-6, as shown in Scheme 2 and Figure S1.

Scheme 2. Scale-Up of the Kinetic Resolution Reaction of 2a Using E. coli Cells Expressing MAO-N D11 in Phosphate Buffer pH 7 and 5% DMF, BH₃NH₃, 2 h.

The 1H,3H-thiazolo[3,4-a]benzimidazole (−)-6 was identified as the oxidation product of the reaction, in which the aromatic secondary amine present in 2a is oxidized by the MAO enzyme. This result supports that benzothiazoline 1a, lacking the aromatic secondary amine, cannot be resolved using this methodology. In addition, since the reduction of an aromatic benzimidazole is challenging, a dynamic kinetic resolution using BH₃–NH₃ as a reducing agent was not possible.

The synthesis of racemic thiazolo-benzimidazoles similar to 6 is described in literature using o-phenylenediamines, 2-mercaptoacetic acid, and different aldehydes.³⁴⁻³⁶ Interestingly, some thiazolo-benzimidazoles derivatives were reported as potent inhibitors of human immunodeficiency virus replication in a variety of human cell lines.³⁷

Attempts to obtain compound 6 by chemical oxidation of TBIM 2a using MnO₂ in MeOH failed, and decomposition of the starting material was observed. The biocatalytic methodology herein described allowed us to isolate the enantiopure thiazolo-benzimidazole (−)-6. Further optimization of the reaction conditions could provide a useful approach for the enantioselective preparation of this type of compounds.

Mechanism Insights: Oxidation Mechanism Proposal

The oxidation mechanisms of monoamine oxidases (MAOs) have been widely studied for the MAO-A and MAO-B isoenzymes, commonly found in mammals. In the literature, several mechanisms have been proposed, depending on the enzyme variant and the substrate,^38,39 including radical mechanisms and hydride transfers between the FAD cofactor and the substrate.⁴⁰

Recently, Turner et al. reported the oxidation of indolines to indoles using MAO-N D11 via a mechanism that involved a transfer of hydrides from the substrate to FAD.⁴¹ Based on this observation, we suggest a similar mechanism for the oxidation of TBIM using MAO-N D11, as shown in Figure 5.

Figure 5.

Proposed oxidation mechanism of TBIM 2a by MAO-N.

In the proposed mechanism, the FAD cofactor of the enzyme abstracts a hydride from the acetalic position of TBIM 2a (H_A), forming a TBIM-FAD adduct. Further details regarding the intermediate steps of this stage have yet to be elucidated. Subsequently, deprotonation of H₂O by the flavin is followed by protonation of a second H₂O molecule from the active site, leading to the donation of the H_B proton by N and the formation of the C=N bond (Figure 5, adduct 2a-FAD). This results in the release of TBIM oxidation product 6 and the formation of FADH₂, which is reconverted by O₂ into FAD, allowing a new catalytic cycle.

Determination of the Absolute Configuration

Once the pure enantiomer (+)-2a was obtained, different methodologies were explored to determine its absolute configuration. Since compound 2a is an oil at room temperature, we attempted to derivatize it by preparing a thioester to obtain a crystalline solid whose structure could be resolved by X-ray diffraction.

First, the thioester 7 was synthesized from the racemic mixture (±)-2a, using p-chlorobenzoic acid and HBTU as coupling agent, in 42% yield, Scheme 3A.

Scheme 3. (A) Synthesis of Thioester 7. (B) ORTEP: Molecular Structure of the Racemic Mixture (±)-7.

The crystal structure of 7 was obtained after the slow evaporation of CH₂Cl₂/n-hexanes (1:9) at room temperature and then determined by single-crystal X-ray diffraction methods. The molecular structure allowed us to confirm the relative syn configuration of 2a. The ORTEP view of the compound is shown in Scheme 3B.

Then, enantiomer (+)-2a was derivatized using the aforementioned procedure. The derivative was subjected to similar crystallization conditions using slow evaporation of CH₂Cl₂/n-hexanes (1:9) at room temperature. However, crystals could not be isolated. This may be due to differences in the crystal morphology and crystal growth between the pure enantiomer and the racemic mixture.⁴²

Computational Prediction of Chiroptical Properties

In order to explore an alternative methodology for the determination of the absolute configuration of the (+)-2a isomer, theoretical calculation studies were performed. The absolute configuration was predicted from the comparison of theoretical and experimental optical rotation values (OR) and the electronic circular dichroism (ECD) spectrum.^43,44

Optical Rotation

First, the theoretical α_D value of both enantiomers (R,R)-2a and (S,S)-2a was calculated. After a conformational search for each enantiomer, the most favorable conformations were submitted to Gaussian16 for DFT optimization and optical rotation prediction (see the Experimental Section for details). The values for the different conformations of each enantiomer were combined through Boltzmann averaging using the Gibbs free energy calculated during the DFT optimization procedure.

The theoretical values of α_D were +46 for (R,R)-2a and −61 for the (S,S)-2a isomer. The isolated isomer (+)-2a has an experimental α_D = +90°; thus, the positive value suggests that (+)-2a is the (R,R)-enantiomer.

Electronic Circular Dichroism

To obtain additional structural information and confirm these results, the theoretical and experimental ECD spectra of the isolated isomer (+)-2a were determined.

The experimental ECD spectrum was recorded in the range 230–400 nm, and the values of Δε were calculated; see Supporting Information, Table S2. For ECD spectra calculations, a time-dependent DFT calculation was performed over the three most favorable conformations of each enantiomer with the same functional and basis set used for optical rotation. Thirty excited states and their electronic transition intensities and rotatory strengths were predicted. Boltzmann averaging was used to combine the results of the three conformations.

The comparison between the experimental and theoretical spectra of both enantiomers is shown in Figure 6. Both in the experimental ECD (black) and in the theoretical spectrum for the (R,R) isomer (green), a negative peak close to 230 nm and a positive peak close to 270 nm are present. These results, in accordance with the optical rotation studies, also suggest that the isolated enantiomer is the (R,R) isomer.

Figure 6.

ECD spectra: experimental (+)-2a (black); calculated (R,R)-2a (green); and calculated (S,S)-2a (pink).

Kinetic Resolution Reaction Scope

Once the reaction conditions were optimized, the substrate scope was explored. A series of TBIM bearing different substituents on the aromatic ring were subjected to a kinetic resolution with MAO-N D11 under the optimized conditions. The results are shown in Table 4 (Supporting Information, Figure S16).

Table 4. Analytical-Scale Screening for the Enzymatic Resolution of TBIM 2b–g Using E. coli Whole Cells Expressing MAO-N D11^a.

		time (h)
		1		2		4		8
TBIM	R	yield (R,R) %b	% eec	yield (R,R) %b	% eec	yield (R,R) %b	% eec	yield (R,R) %b	% eec
2a	H	39	92	28	94	19	94	4	92
2b	4-5-naphthyl	47	0	45	0	44	0	43	0
4-2c	CH3	31	70	15	72	8	78	2	64
5-2c		26	72	10	76	5	74	3	40
4-2d	–CO2Et	43	0	33	0	23	0	16	0
5-2d		41	0	35	0	28	0	22	0
4-2e	–NO2	48	0	45	0	42	0	31	0
5-2e		49	8	48	10	46	16	37	14
4-2f	–F	36	50	23	44	14	44	2	46
5-2f		15	56	0		0		0
4-2g	–Br	42	0	37	0	18	0	11	0
5-2g		22	26	11	30	6	90	0

^aConditions: cells 1 g/mL, phosphate buffer 0.1 M pH 7, BH₃NH₃, DMF 5%, 37 °C.

^bCalculated using chiral HPLC using an external standard.

^cCalculated using chiral HPLC.

The results of the study indicated that the nonsubstituted TBIM 2a achieved the best results, obtaining (R,R)-2a in 44% yield and 94% ee after 2 h. On the other hand, substitutions at the aromatic ring in TBIM 2b–g led to a decrease in the % ee.

Overall, the 5-substituted TBIM gave better % ee than the 4-TBIM, probably due to steric hindrance or electronic effects.

The electronic or steric nature of the substituent present in the benzene ring of TBIM seems to have an impact on the % ee. The best results were obtained for compounds 4-/5-2c bearing a small electron donor group (Me) with 72 and 76% ee, respectively, after 2 h. The presence of the bulky naphthyl group and the electron-withdrawing groups (–CO₂Et and –NO₂) were deleterious for the kinetic resolution reaction, and no oxidation product was detected. Degradation products were not detected under the chromatographic conditions used.

Compounds 4/5-2f (F) afforded moderate results; pure enantiomers were obtained in 50 and 56% ee, respectively, after 1 h of reaction. Notably, in the case of compound 5-2f (F), neither enantiomer was observed after 4 h, likely due to the rapid chemical degradation of the compound in the reaction medium.

No resolution was achieved for compound 4-2g (Br), while compound 5-2g (Br) was obtained with an ee of 90% but in a low yield (6%) after 4 h of the reaction. Although a good enantioselectivity was obtained at longer reaction times, probably the instability of the compounds in the reaction medium was the cause of the low yields.

Docking Experiments

In order to gain deeper insights into the observed enantioselectivity of MAO-N D-11, a multistep in silico analysis was performed over both enantiomers of substrate (+) and (−) syn-2a, which included flexible docking, molecular dynamics (MD) simulations, and free energy calculations. The most favorable docking positions were selected based on mechanistic considerations and were then subjected to 100 ns of free and steered MD simulations. Finally, MM-GBSA and MM-PBSA free energy calculations were performed over the clusterized trajectory of the MDs to provide a comprehensive understanding of the results.

The results indicate that both enantiomers (R,R)-2a and (S,S)-2a are optimally positioned for the reaction and have negative binding energies of −39.0 ± 3.2 and −31.9 ± 3.1 kcal/mol (MM-GBSA) and suggest that the two stereocomplementary molecules have a high binding affinity. Only a slight advantage was observed for the SS enantiomer regarding the positioning angle of the H_A hydrogen in relation to the nitrogen of the FAD. In addition, the RR enantiomer shows a subtle steric hindrance at the thiol chain with the amino acid PHE382, something that does not occur with the SS enantiomer (Supporting Information, Figure S17). However, experimentally, it was observed that one of the enantiomers reacted more slowly. From this analysis, it can be concluded that the observed enantioselectivity at 4 and 24 h of the reaction may be due to the transition speed of the enantiomers through the entry tunnel of the active site. The steric hindrance toward the RR enantiomer may also affect the behavior of the molecule inside the active site, reducing its binding capacity and increasing the reaction time.

The MAO-N entry tunnel has been mutated in several studies, proving to be a limiting factor in substrate specificity. In this case, the results obtained from the MD simulations and energy calculations seem to indicate that the physicochemical properties of the tunnel and not the active site might be crucial for the observed reaction rate. This observation supplies an interesting target to improve substrate specificity in the future with this type of compounds.

Flexible docking analyses for all the derivatives 2b–g were performed. Docking analysis of 2b (naphthyl) and 2d (CO₂Et) showed that clusters with the negative energy of binding did not provide a suitable structure from a mechanistic point of view, which correlates with the lack of experimental activity. Additionally, a clear increase in energy binding was observed in the entry regions of the tunnel, indicating steric hindrance that could prevent entry into the active site.

Conclusions

A series of 12 new 1H,3H-thiazolo[3,4-a]benzimidazolines TBIM 2a–g compounds were synthesized through the double condensation of 1,2-aminoanilines and dithiane, with yields ranging from 49 to 85%. In order to resolve the racemic mixture, the use of MAO enzymes was investigated.

Multiple conditions were evaluated, including different MAO-N variants, reaction times, and cosolvents. The optimal resolution conditions were achieved using the MAO-N D11 variant and 5% DMF as cosolvent after 2 h of reaction. This approach was effective for the enantiomeric resolution of TBIM 2a, but not for the sulfur analogue TBTZ 1a.

The reaction scope was studied using different substituted TBIM 2a–g. The best result was achieved for the nonsubstituted TBIM 2a, yielding the pure enantiomer (+)-2a with 94% ee and 44% yield at 4 mM. Additionally, the enantiomerically pure 1H,3H-thiazolo[3,4-a]benzimidazole (−)-6 was isolated as the oxidation product, providing a new methodology for its enantioselective preparation, to be optimized. The formation of this benzimidazole avoided the dynamic resolution of the racemic mixture. This result also explains why benzothiazoline 1a could not be resolved using MAO enzymes.

The absolute configuration was determined by measuring the α_D value and the ECD spectrum. The results were compared to those obtained by theoretical calculations, indicating that the absolute configuration of the isolated enantiomer is (R,R). Additionally, docking studies demonstrate that the SS enantiomer has only a slight advantage in positioning in the active site, indicating that the entry tunnel must account for the observed differences in reaction rates.

In summary, a new application of MAO-N D11 for the resolution of 1H,3H-thiazolo[3,4-a]benzimidazoline was achieved.

Experimental Section

General Procedure for the Synthesis of 1H,3H-Thiazolo[3,4-a]benzimidazolines (TBIM 2)

To a stirred solution of 2-diaminoaniline (1 mmol) in EtOH (10 mL) were added 1,4-dithiane-2,5-dithiol (1.2 mmol) and catalytic p-TsOH acid (0.1 mmol). The mixture was heated at reflux for 1 h. The solvent was then removed under reduced pressure. The crude was poured into water (30 mL), extracted with EtOAc (3 × 50 mL), dried (Na₂SO₄), and filtered. The organic layer was evaporated under reduced pressure. The crude reaction mixture was purified by column chromatography on flash silica gel using n-hexanes/EtOAc, to obtain the TBIM compounds.

(3a,4-Dihydro-1H,3H-benzo[4,5]imidazo[1,2-c]thiazol-1-yl)methanethiol (2a)

Prepared starting from o-phenylenediamine (200 mg, 0.925 mmol) and purified with n-hexanes/EtOAc 8:2 to give 2a (112 mg, 58%) as a colorless oil: ¹H NMR: δ 6.81 (m, 3H), 6.71 (dd, J = 6.1, 2.6 Hz, 1H), 5.35 (dd, J = 7.5, 5.4 Hz, 1H), 4.98 (dd, J = 8.9, 4.7 Hz, 1H), 4.18 (br, 1H), 3.16 (dd, J = 10.6, 5.4 Hz, 1H), 2.91 (ddd, J = 13.8, 8.9, 6.3 Hz, 1H), 2.83 (dd, J = 10.6, 7.5 Hz, 1H), 2.70 (ddd, J = 13.6, 10.6, 4.7 Hz, 1H), 1.90 (dd, J = 10.6, 6.3 Hz, 1H); ¹³C RMN: δ 140.8, 138.0, 121.9, 121.8, 111.9, 110.0, 83.1, 71.5, 38.9, 33.9; IR (NaCl) ν: 1620, 1495, 1406, 1292, 1250, 1115, 740, 695 cm^–1; HRMS (ESI) m/z: calcd for C₁₀H₁₂N₂S₂ [M + Na]⁺ 247.0340; found, 247.0341.

(3a,4-Dihydro-1H,3H-naphtho[2′,3′:4,5]imidazo[1,2-c]thiazol-1-yl)methanethiol (2b)

Prepared starting from 2,3-diaminonaphthalene (160 mg, 1 mmol) and purified with n-hexanes/EtOAc 8:2 to give 2b (194.5 mg, 71%); brown solid, mp 165–173 °C; ¹H NMR: δ 7.62 (m, 1H), 7.56 (m, 1H), 7.23 (m, 2H), 7.04 (s, 1H), 6.91 (s, 1H), 5.48 (dd, J = 7.6, 5.2 Hz, 1H), 5.08 (dd, J = 8.8, 4.7 Hz, 1H), 4.48 (br, 1H), 3.19 (dd, J = 10.6, 5.2 Hz, 1H), 3.00 (ddd, J = 14.9, 8.7, 6.4 Hz, 1H), 2.82 (dd, J = 10.5, 7.4 Hz, 1H), 2.78 (ddd, J = 13.8, 10.8, 5.3 Hz, 1H), 1.93 (dd, J = 10.6, 6.3 Hz, 1H); ¹³C NMR: δ 141.3, 139.1, 131.2, 130.7, 126.6, 126.2, 123.8, 123.8, 105.5, 104.7, 83.0, 70.8, 39.1, 33.8; IR (NaCl) ν: 3346, 2924, 1633, 1599, 1472, 1433, 1375, 1314, 1265, 959, 853, 741 cm^–1; HRMS (ESI) m/z: calcd for C₁₄H₁₃N₂S₂ [M – H]⁻ 273.0520; found, 273.0521.

(6-Methyl-3a,4-dihydro-1H,3H-benzo[4,5]imidazo[1,2-c]thiazol-1-yl) Methanethiol (4-2c) and (7-Methyl-3a,4-dihydro-1H,3H-benzo[4,5]imidazo[1,2-c]thiazol-1-yl)methanethiol (5-2c)

Prepared starting from 4-methyl-o-phenylenediamine (120 mg, 1 mmol) and purified with CH₂Cl₂/n-hexanes 1:1 to give a mixture of isomers 4-2c and 5-2c, 0.138 g, 58%, (7:3).

4-2c: (40.5 mg, 17%): colorless oil; ¹H NMR: δ 6.66 (m, 2H), 6.54 (t, J = 11.2 Hz, 1H), 5.33 (dd, J = 7.1, 5.5 Hz, 1H), 4.93 (dd, J = 9.0, 4.7 Hz, 1H), 4.13 (br, 1H), 3.16 (dd, J = 10.7, 5.4 Hz, 1H), 2.89 (ddd, J = 13.7, 9.0, 6.2 Hz, 1H), 2.83 (dd, J = 10.7, 7.3 Hz, 1H), 2.68 (ddd, J = 13.6, 10.6, 4.7 Hz, 1H), 2.24 (s, 3H), 1.91 (dd, J = 10.6, 6.2 Hz, 1H); ¹³C NMR: δ 138.6, 138.3, 131.7, 122.0, 112.9, 109.8, 83.4, 72.1, 39.0, 33.9, 21.2; IR (NaCl) ν: 3346, 1497, 1458, 1373, 1265, 1217, 1150, 760 cm^–1; HRMS (ESI) m/z: calcd for C₁₁H₁₄N₂NaS₂ [M – H]⁻ 261.0496; found, 261.0500.

5-2c: (27.9 mg, 12%): colorless oil; ¹H NMR: δ 6.63 (m, 3H), 5.32 (dd, J = 7.7, 5.5 Hz, 1H), 4.97 (dd, J = 9.0, 4.7 Hz, 1H), 3.14 (dd, J = 10.6, 5.4 Hz, 1H), 2.90 (ddd, J = 13.6, 9.0, 6.2 Hz, 1H), 2.81 (dd, J = 10.6, 7.7 Hz, 1H), 2.69 (ddd, J = 13.6, 10.6, 4.6 Hz, 1H), 2.28 (d, J = 0.7 Hz, 3H), 1.91 (dd, J = 10.6, 6.2 Hz, 1H); ¹³C NMR: δ 141.4, 135.6, 132.2, 122.1, 112.5, 110.9, 83.4, 77.5, 71.3, 38.7, 34.0, 21.4; IR (NaCl) ν: 3344, 1492, 1458, 1373, 1260, 1217, 1208, 1155, 760 cm^–1; HRMS (ESI) m/z: calcd for C₁₁H₁₄N₂NaS₂ [M – H]⁻ 261.0496; found, 261.0500.

Ethyl 1-(Mercaptomethyl)-3a,4-dihydro-1H,3H-benzo[4,5]imidazo[1,2-c]thiazole-6-carboxylate (4-2d) and Ethyl 1-(Mercaptomethyl)-3a,4-dihydro-1H,3H-benzo[4,5]imidazo[1,2-c]thiazole-7-carboxylate (5-2d)

Prepared starting from ethyl 3,4-diaminobenzoate (180 mg, 1 mmol) and purified with n-hexanes/EtOAc 8:2 to give a mixture of isomers 4-2d and 5-2d 0.145 g, 49%, (3:7).

4-2d: (35.6 mg, 12%), yellow oil; ¹H NMR: δ 7.61 (dd, J = 8.1, 1.6 Hz, 1H), 7.31 (d, J = 1.5 Hz, 1H), 6.76 (d, J = 8.1 Hz, 1H), 5.42 (dd, J = 7.8, 5.3 Hz, 1H), 5.00 (dd, J = 8.8, 4.6 Hz, 1H), 4.31 (q, J = 7.1 Hz, 2H), 3.16 (dd, J = 10.6, 5.3 Hz, 1H), 2.91 (ddd, J = 13.8, 8.8, 6.6 Hz, 1H), 2.80 (dd, J = 10.6, 7.8 Hz, 1H), 2.72 (ddd, J = 13.8, 10.5, 4.7 Hz, 1H), 1.86 (dd, J = 10.5, 6.5 Hz, 1H), 1.35 (t, J = 7.1 Hz, 3H); ¹³C NMR: δ 166.8, 144.7, 138.2, 125.0, 124.0, 111.9, 108.4, 83.2, 70.2, 60.7, 38.8, 33.7, 14.5; IR (NaCl) ν: 2976, 2926, 1700, 1603, 1497, 1450, 1366, 1285, 1210, 1103, 1020, 765 cm^–1; HRMS (ESI) m/z: calcd for C₁₃H₁₆N₂NaO₂S₂ [M + Na]⁺ 319.0545; found, 319.0554.

5-2d: (77.1 mg, 26%), yellow oil; ¹H NMR: δ 7.57 (dd, J = 8.0, 1.5 Hz, 1H), 7.38 (d, J = 1.6 Hz, 1H), 6.57 (d, J = 8.0 Hz, 1H), 5.48 (dd, J = 6.8, 5.3 Hz, 1H), 5.00 (dd, J = 8.4, 5.1 Hz, 1H), 4.32 (q, J = 7.1 Hz, 2H), 3.20 (dd, J = 10.7, 5.3 Hz, 1H), 2.90 (ddd, J = 13.6, 8.4, 6.7 Hz, 1H), 2.82 (dd, J = 10.7, 6.9 Hz, 1H), 2.73 (ddd, J = 13.7, 10.2, 5.1 Hz, 1H), 1.88 (dd, J = 10.2, 6.7 Hz, 1H), 1.36 (t, J = 7.1 Hz, 3H); ¹³C NMR: δ 166.8, 142.9, 139.7, 125.6, 122.8, 110.5, 108.4, 83.6, 72.0, 60.7, 39.4, 33.6, 14.6; IR (NaCl) ν: 2980, 2920, 1700, 1602, 1492, 1455, 1366, 1288, 1210, 1115, 1020, 765 cm^–1; HRMS (ESI) m/z: calcd for C₁₃H₁₆N₂NaO₂S₂ [M + Na]⁺ 319.0545; found, 319.0554.

(6-Nitro-3a,4-dihydro-1H,3H-benzo[4,5]imidazo[1,2-c]thiazol-1-yl)methanethiol (4-2e) and (7-Nitro-3a,4-dihydro-1H,3H-benzo[4,5]imidazo[1,2-c]thiazol-1-yl)methanethiol (5-2e)

Prepared starting from 4-nitro-o-phenylenediamine (150 mg, 1 mmol) and purified with n-hexanes/EtOAc 8:2 to give a mixture of isomers 4-2e and 5-2e 0.223 g, 83%, (3:7).

4-2e: (59.2 mg, 22%): red solid; mp 206–208 °C; ¹H NMR 7.80 (dd, J = 8.5, 2.2 Hz, 1H), 7.43 (d, J = 2.2 Hz, 1H), 6.74 (d, J = 8.5 Hz, 2H), 5.54 (t, J = 6.0 Hz, 1H), 4.99 (dd, J = 8.8, 4.6 Hz, 1H), 4.59 (br, 1H), 3.22 (dd, J = 10.6, 5.2 Hz, 1H), 2.90 (dt, J = 8.8, 7.0 Hz, 1H), 2.84 (dd, J = 10.6, 7.9 Hz, 1H), 2.75 (ddd, J = 14.0, 10.5, 4.6 Hz, 1H), 1.83 (dd, J = 10.5, 6.8 Hz, 1H); ¹³C NMR: δ 146.0, 142.6, 138.9, 119.5, 107.3, 105.0, 83.4, 69.4, 38.9, 33.4; IR (NaCl) ν: 1603, 1489, 1447, 1400, 1364, 1317, 1281, 1130, 1072, 858, 741 cm^–1; HRMS (ESI) m/z: calcd for C₁₀H₁₂N₃NaO₂S₂ [M + H]⁺ 270.0365; found, 270.0366.

5-2e: (132 mg, 49%): orange solid; mp 128–134 °C; ¹H NMR: δ 7.81 (dd, J = 8.5, 2.2 Hz, 1H), 7.58 (d, J = 2.2 Hz, 1H), 6.51 (d, J = 8.5 Hz, 1H), 5.62 (ddd, J = 6.7, 5.1, 1.5 Hz, 1H), 4.97 (dd, J = 8.5, 4.9 Hz, 1H), 4.92 (br, 1H), 3.28 (dd, J = 10.8, 5.2 Hz, 1H), 2.90 (m, 2H), 2.77 (ddd, J = 13.9, 10.2, 4.9 Hz, 1H), 1.87 (dd, J = 10.2, 6.9 Hz, 1H); ¹³C NMR: δ 144.8, 141.5, 140.0, 120.8, 106.4, 105.0, 84.0, 71.9, 39.5, 33.4; IR (NaCl) ν: 1600, 1490, 1446, 1400, 1362, 1318, 1280, 1130, 1075, 858, 740 cm^–1; HRMS (ESI) m/z: calcd for C₁₀H₁₂N₃NaO₂S₂ [M + H]⁺ 270.0365; found, 270.0366.

(6-Fluoro-3a,4-dihydro-1H,3H-benzo[4,5]imidazo[1,2-c]thiazol-1-yl)methanethiol (4-2f) and (7-Fluoro-3a,4-dihydro-1H,3H-benzo[4,5]imidazo[1,2-c]thiazol-1-yl)methanethiol (5-2f)

Prepared starting from 4-fluoro-1,2-phenylenediamine (126 mg, 1 mmol) and purified with n-hexanes/EtOAc 8:2 to give a mixture of isomers 4-2f and 5-2f, 0.182 g, 75%, (7:3).

4-2f: (94.5 mg, 39%): red solid; mp 82–85 °C; ¹H NMR: δ 6.66 (dd, J = 8.4, 4.7 Hz, 1H), 6.47 (td, J = 9.4, 2.4 Hz, 1H), 6.38 (dd, J = 8.8, 2.4 Hz, 1H), 5.39 (t, J = 6.0 Hz, 1H), 4.86 (dd, J = 9.0, 4.7 Hz, 1H), 4.28 (br, 1H), 3.18 (dd, J = 10.8, 5.3 Hz, 1H), 2.85 (m, 2H), 2.68 (ddd, J = 13.7, 10.4, 4.7 Hz, 1H), 1.90 (dd, J = 10.4, 6.4 Hz, 1H); ¹³C NMR: δ 159.17 (d, J = 237.2 Hz), 139.51 (d, J = 12.0 Hz), 136.56 (d, J = 1.7 Hz), 109.84 (d, J = 9.8 Hz), 106.40 (d, J = 23.6 Hz), 99.30 (d, J = 28.0 Hz), 84.0, 72.7, 39.2, 33.7; IR (NaCl) ν: 1614, 1491, 1391, 1354, 1310, 1273, 1255, 1134, 833, 795 cm^–1; HRMS (ESI) m/z: calcd for C₁₀H₁₂FN₂S₂ [M + H]⁺ 243.0426; found, 243.0404.

5-2f: (60.6 mg, 25%): orange oil; ¹H NMR: δ 6.62 (dd, J = 8.3, 4.7 Hz, 1H), 6.54 (dd, J = 9.0, 2.4 Hz, 1H), 6.46 (ddd, J = 9.3, 8.4, 2.5 Hz, 1H), 5.35 (dd, J = 7.9, 5.5 Hz, 1H), 4.89 (dd, J = 9.0, 4.5 Hz, 1H), 3.14 (dd, J = 10.7, 5.4 Hz, 1H), 2.88 (ddd, J = 13.2, 8.7, 6.2 Hz, 1H), 2.82 (dd, J = 10.6, 7.8 Hz, 1H), 2.70 (ddd, J = 13.8, 10.5, 4.5 Hz, 1H), 1.86 (dd, J = 10.5, 6.4 Hz, 1H); ¹³C NMR: δ 159.6 (d, J = 238.0 Hz), 142.9 (d, J = 11.7 Hz), 133.8 (d, J = 1.8 Hz), 112.7 (d, J = 9.7 Hz), 107.0 (d, J = 23.5 Hz), 98.6 (d, J = 28.5 Hz), 83.9, 70.9, 38.5, 33.7; IR (NaCl) ν: 1610, 1490, 1390, 1314, 1272, 1255, 1130, 834, 785 cm^–1; HRMS (ESI) m/z: calcd for C₁₀H₁₂FN₂S₂ [M + H]⁺ 243.0426; found, 243.0404.

(7-Bromo-3a,4-dihydro-1H,3H-benzo[4,5]imidazo[1,2-c]thiazol-1-yl)methanethiol (4-2g) and (6-Bromo-3a,4-dihydro-1H,3H-benzo[4,5]imidazo[1,2-c]thiazol-1-yl)methanethiol (5-2g)

Prepared starting from 4-bromo-1,2-diaminobenzene (190 mg, 1 mmol) and purified with n-hexanes/EtOAc 8:2 to give a mixture of isomers 4-2g and 5-2g 0.209 g, 69%, (7:3).

4-2g: 109.2 mg; 36%, white solid; mp 146–152 °C; ¹H NMR: δ 6.90 (dd, J = 8.1, 1.8 Hz, 1H), 6.75 (d, J = 1.8 Hz, 1H), 6.63 (d, J = 8.1 Hz, 1H), 5.37 (dd, J = 7.1, 5.4 Hz, 1H), 4.88 (dd, J = 8.9, 4.6 Hz, 1H), 4.26 (br, 1H), 3.17 (dd, J = 10.7, 5.3 Hz, 1H), 2.87 (ddd, J = 13.0, 8.5, 6.0 Hz, 1H), 2.82 (dd, J = 10.8, 7.2 Hz, 1H), 2.68 (ddd, J = 13.7, 10.5, 4.7 Hz, 1H), 1.87 (dd, J = 10.5, 6.4 Hz, 1H); ¹³C NMR: δ 139.8, 139.8, 123.9, 114.0, 113.7, 110.9, 83.5, 71.7, 39.1, 33.7; IR (NaCl) ν: 3362, 2930, 2543, 1603, 1482, 1422, 1401, 1297, 1246, 1218, 1060, 1001, 964, 907, 847, 797, 758, 696, 619 cm^–1; HRMS (ESI) m/z: calcd for C₁₀H₁₂BrN₂S₂ [M + H]⁺ 302.9625; found, 302.9483.

5-2g: 57.6 mg; 19%; white solid; mp 102–105 °C; ¹H NMR: δ 6.88 (m, 2H), 6.52 (m, 1H), 5.37 (dd, J = 7.5, 5.4 Hz, 1H), 4.89 (dd, J = 8.9, 4.6 Hz, 1H), 4.17 (br, 1H), 3.16 (dd, J = 10.7, 5.3 Hz, 1H), 2.88 (ddd, J = 14.1, 9.1, 6.6 Hz, 1H), 2.82 (dd, J = 10.7, 7.6 Hz, 1H), 2.69 (ddd, J = 13.8, 10.6, 4.6 Hz, 1H), 1.86 (dd, J = 10.5, 6.5 Hz, 1H); ¹³C NMR: δ 142.4, 137.5, 124.3, 113.7, 113.1, 112.5, 83.5, 71.2, 38.9, 33.7; IR (NaCl) ν: 3360, 2920, 2540, 1630, 1480, 1420, 1408, 1298, 1240, 1215, 1063, 1010, 961, 908, 847, 797, 760, 696, 620 cm^–1; HRMS (ESI) m/z: calcd for C₁₀H₁₂BrN₂S₂ [M + H]⁺ 302.9625; found, 302.9483.

Synthesis of S-(((1R,3aR)-3a,4-Dihydro-1H,3H-benzo[4,5]imidazo[1,2-c]thiazol-1-yl)methyl) 4-Chlorobenzothioate (7)

To a stirred solution of p-chlorobenzoic acid (62 mg, 0.4 mmol) in CH₂Cl₂ dry (8 mL) were added HBTU (163 mg, 0.43 mmol), DIPEA (189 μL, 1.08 mmol), and 4-DMAP (4 mg, 0.036 mmol) y (±)-2a (80 mg, 0.36 mmol) at 0 °C. The mixture was stirred at rt for 2 h. The crude was poured into water (30 mL), extracted with EtOAc (3 × 50 mL), dried (Na₂SO₄), and filtered. The organic layer was evaporated under reduced pressure. The crude reaction mixture was purified by column chromatography on flash silica gel using n-hexanes/EtOAc (7:3) to give 7 (55 mg, 42%): green solid; mp 121–126 °C; ¹H NMR: δ 7.94 (m, 2H), 7.43 (m, 2H), 6.78 (m, 2H), 6.66 (m, 2H), 5.43 (ddd, J = 7.3, 5.2, 1.8 Hz, 1H), 5.05 (dd, J = 8.5, 5.7 Hz, 1H), 3.44 (m, 2H), 3.22 (ddd, J = 10.8, 5.5, 1.5 Hz, 1H), 2.85 (ddd, J = 11.0, 7.2, 1.4 Hz, 1H); ¹³C RMN: δ 190.3, 140.7, 140.2, 138.2, 135.2, 129.1, 128.9, 121.9, 111.9, 109.9, 83.3, 68.5, 39.2, 37.9. IR (NaCl) ν: 3053, 2928, 1665, 1485, 1203, 916 cm^–1

Biocatalytic Assays

General Procedure for the Analytical-Scale Screening

In a 15 mL Falcon tube was added TBIM 2a (0.05 mmol) dissolved in an appropriate solvent (500 μL), BH₃NH₃ (0.2 mmol, 4 equiv), and potassium phosphate buffer (10 mL, 0.1 M, pH = 7). Cell pellet (500 mg) from E. coli cultures expressing a MAO-N, corresponding to the D5, D9, and D11 variants were expressed as previously reported,³² was added to the solution. The tube was placed in a shaking incubator and shaken at 37 °C and 250 rpm. The samples were prepared as follows: aqueous NaCl solution was added to a 1000 μL sample of the reaction mixture in a test tube, followed by 1 mL of EtOAc. After vigorous mixing by a vortex mixer, the organic phase was separated, dried with MgSO₄, and concentrated. Then, the residue was taken in DMSO (50 μL), iPrOH (450 μL), and n-hexanes (500 μL) and analyzed by chiral HPLC (DAICEL Chiralpak IF, 250 mm × 4.6 mm, 5 μm column; flow rate 1 mL/min; λ 306 nm; eluent = n-hexanes/iPrOH 8:2; Rt [(R,R)] = 7.3 min, Rt [(S,S)] = 9.4 min).

2b: n-hexanes/iPrOH 8:2; λ 306 nm, t₁ = 14.8 min, and t₂ = 17.8 min.

4-2c: n-hexanes/iPrOH 8:2; λ 312 nm, t₁ = 8.0 min, and t₂ = 9.6 min.

5-2c: n-hexanes/iPrOH 8:2; λ 312 nm, t₁ = 8.1 min, and t₂ = 9.8 min.

4-2d: n-hexanes/iPrOH 6:4; λ 330 nm, t₁ = 6.7 min, and t₂ = 19.0 min.

5-2d: n-hexanes/iPrOH 6:4; λ 330 nm, t₁ = 6.8 min, and t₂ = 10.6 min.

4-2e: n-hexanes/iPrOH 6:4; λ 306 nm, t₁ = 14.2 min, and t₂ = 21.3 min.

5-2e: n-hexanes/iPrOH 6:4; λ 306 nm, t₁ = 9.0 min, and t₂ = 13.5 min.

4-2f: n-hexanes/iPrOH 8:2; λ 313 nm, t₁ = 7.9 min, and t₂ = 8.6 min.

5-2f: n-hexanes/iPrOH 8:2; λ 315 nm, t₁ = 8.6 min, and t₂ = 11.5 min.

4-2g: n-hexanes/iPrOH 8:2; λ 318 nm, t₁ = 8.4 min, and t₂ = 10.1 min.

5-2g: n-hexanes/iPrOH 8:2; λ 318 nm, t₁ = 9.1 min, and t₂ = 15.1 min.

Preparative Procedure

In a 250 mL Erlenmeyer flask was added TBIM 2a (0.45 mmol) dissolved in DMF (5 mL), BH₃NH₃ (0.2 mmol, 4 equiv), and potassium phosphate buffer (10 mL, 0.1 M, pH = 7). Cell pellets (500 mg) from E. coli cultures containing MAO-N D11 were added to the solution. The tube was placed in a shaking incubator and shaken at 37 °C and 250 rpm. The crude was poured into a saturated NaCl solution (1 mL) and extracted with EtOAc (3 × 1 mL). The organic layer was evaporated under reduced pressure and purified by SiO₂ column chromatography, eluent n-hexanes/EtOAc (8:2) to give (+)-2a (35 mg, 35%, 89% ee, α_D = +90°) and (−)-6 (5 mg, 5%).

(S)-(1H,3H-Benzo[4,5]imidazo[1,2-c]thiazol-1-yl)methanethiol (−)-6

¹H RMN: δ 7.73 (dd, J = 6.1, 2.2 Hz, 1H), 7.35 (dd, J = 5.5, 2.2 Hz, 1H), 7.28 (m, 2H), 5.78 (ddd, J = 5.5, 3.1, 1.9 Hz, 1H), 4.44 (dd, J = 14.6, 1.9 Hz, 1H), 4.16 (d, J = 14.5 Hz, 1H), 3.23 (m, 2H), 1.50 (t, J = 8.9 Hz, 1H); ¹³C RMN: δ 158.2, 149.1, 131.6, 122.8, 122.7, 120.5, 109.5, 62.0, 32.2, 27.9. [α]²⁰_D = −55.6 (CH₂Cl_2,c = 0.21); IR (NaCl) ν: 2922, 2555, 1616, 1525, 1454, 741 cm^–1 HRMS (ESI) m/z: calcd for C₁₀H₁₁N₂S₂ [M + H]⁺ 223.0364; found, 223.0366.

Optical Rotation and ECD Spectra Computational Predictions

SS and RR enantiomer structures were built in MOE2019 (Chemical Computing Group ULC). A stochastic conformational search was performed in the MOE over both enantiomers. Dichloromethane was used as a simulated solvent for the conformational search through a generalized Born implicit solvation model (ε = 8.93). All conformations within an energy window of 7 kcal/mol, as estimated with (AMBER14:EHT force field⁴⁵), and RMSD > 0.25 Å were retained. Conformational analysis showed that the structure is very rigid, with the thiol bond free rotation being the only variation among the different conformations. The three conformations with lowest energy for each enantiomer were used as input for density functional theory (DFT) calculations.⁴⁶ Optimization and frequency calculations were performed in Gaussian16 using the ub3lyp functional⁴⁷ and 6-31++g(d,p) basis set, with an ultrafine integration grid and no symmetry constraints. Implicit solvation (dichloromethane for optical rotation and chloroform for ECD spectra, to match the experimental conditions) was simulated by an IEFPCM method, with radii and non-electrostatic terms from Truhlar and co-workers’ SMD solvation model.⁴⁸ Optical rotation was calculated at 589.3 nm from DFT-optimized structures.⁴⁹ For ECD spectra calculations, a time-dependent DFT calculation⁵⁰ was performed with the same functional and basis set in order to get 30 excited states and their electronic transition intensities and rotatory strengths.

HPLC Method

A chiral phase separation column was used (DAICEL Chiralpak IF, 250 mm × 4.6 mm, 5 μm). Measurements were carried out under isocratic conditions. The eluent consisted of n-hexanes (mobile phase A) and isopropanol (mobile phase B) at a flow rate of 1.0 mL/min. The injection volume was 20 μL. Data and chromatograms were collected and analyzed using the Empower System program from Waters Corporation, 2002. Retention times and isocratic conditions are available in the Supporting Information.

% Yields

Yields were calculated using an external standard (Std 1) and corrected using a recovering factor, calculated as follows:

• (a)External standard preparation (Std 1): TBIM (50 μL of a solution 90 mM in DMSO) was diluted in iPrOH (450 μL) and n-hexanes (500 μL). The solution was filtered and analyzed by chiral HPLC.
• (b)Standard inoculated into inactivated cells (Std 2): TBIM (50 μL of a solution 90 mM in DMSO) was suspended in EtOAc (1 mL) and it was added an E. coli culture containing a MAO-N D11 suspension (1 mL). After vigorous mixing by a vortex mixer, the organic phase was separated, dried with MgSO₄, and concentrated. Then, the residue was taken in DMSO (50 μL) and added isOH (450 μL) and n-hexanes (500 μL) and analyzed by chiral HPLC.
• (c)Recovering factor (F) = A_(Std2)/A_(Std1).

Recovering factors are available in Table S3.

ECD Experiments

In a 2 mL vial, was dissolved (+)-2a in CHCl₃ (300 μL), and then diluted with CHCl₃ (final concentration = 9,75 × 10^–6 M). The ECD spectrum was recorded in the range 230–400 nm using CHCl₃ as the reference. The data obtained (θ, mdeg) were converted into ΔA according to⁴⁴

Then, the Δε was calculated according to

where c = 1 cm and [2a] = 9,75 × 10^–6 M.

Data available in the Supporting Information.

In Silico Reconstruction and Relaxation of Missing Structural Regions in MAO-N D11

The amino acid sequence of the full MAO-N D11 (495 residues) was aligned to the crystal structure (PDB code 3zdn, chains A and B) and a 3D model was created with MODELLER⁵¹ to rebuild the regions not defined in the electron density map. Protonation states at pH 7.8 were assigned by the PDB2PQR server.⁵² The FAD cofactor was added as found in the crystal structure; the model was neutralized with 12 K⁺ ions, solvated with explicit waters and 1 M KCl, and minimized in solution. Then, it was thermalized to 310 K at NVT and simulated during 100 ns by employing MD at NPT (P = 1 bar). To treat the protein the ff19SB force field⁵³ was used, and the entire system was surrounded by a truncated octahedral box of TIP3P water molecules,⁵⁴ applying Dang’s parameters on ions.⁵⁵ The FAD cofactor was treated with the parameters found in the RED database.⁵⁶ All systems were simulated by using the Langevin algorithm to control the temperature and the pressure, with a coupling constant of 5 ps. SHAKE was used to keep all bonds involving hydrogen at their equilibrium values, which allowed us to use a 2 fs step for the integration of Newtons equations of motion. Long-range electrostatic interactions were accounted for using the Particle Mesh Ewald method with standard defaults. All simulations were carried out using the PMEMD CUDA code module of AMBER18 and analyzed with CPPTRAJ.⁵⁷ Data of RMSD and RMSF are available in Figures S18 and S19.

Docking of Substrate 1 and MD Simulations

Ligands were initially prepared using YASARA’s molecular modeling software. Solvation of the ligands was performed in an explicit solvent shell at pH 7.4 and a 0.9% NaCl concentration prior to energy minimization. Energy minimization was performed to refine the bond lengths and angles of the ligands. The minimized conformation with the lowest energy for each ligand was used in the docking experiments.

Flexible docking was performed using AutoDock VINA software with default parameters but with the number of VINA runs increased to 250. The size of the docking box was set to 5 Å around the amino acids in the active site of each protein structure. Clusters were generated based on positions with a root-mean-square deviation (RMSD) lower than 2.5 Å and were represented by the position with the lowest free energy of binding.

Every cluster generated by AutoDock VINA for each substrate used in amine oxidation was analyzed. The binding energy, distance for hydride transfer from FAD, and interactions were evaluated.

The best scoring docking structures for both the SS and RR enantiomers were subjected to two different 100 ns MD simulations in equal conditions as the above indicated, but one of them using a harmonic potential of 1.5 kcal/mol/Ų to steer the substrate toward the FAD cofactor to reach the reaction distance. Both simulations were clustered and MM-GBSA/PBSA calculations were performed. Data are available in Table S5.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.3c03223.

• Full spectroscopic data and copies of ¹H NMR and ¹³C NMR spectra and crystal structure of compound 7 (PDF)

The authors declare no competing financial interest.