Resolution of Racemic Aryloxy-Propan-2-yl Acetates via Lipase-Catalyzed Hydrolysis: Preparation of Enantiomerically Pure/Enantioenriched Mexiletine Intermediates and Analogs

Round 1

Reviewer 1 Report

In this work, the lipase kinetic resolution (KR) of aryloxy-propan-2-yl acetates was achieved by hydrolysis to produce enantiomerically pure/enantioenriched mexiletine intermediates and analogs. In the reaction, racemic acetates rac-1-(2,6-dimethylphenoxy) propan-2-yl acetate (rac-5a), rac-1-(2,4-dimethylphenoxy) propan-2-yl acetate (rac-5b), rac-1-(o-tolyloxy) propan-2-yl acetate (rac-5c) and rac-1-(naphthalen-1-yloxy)propan-2-yl acetate (rac-5d) were employed as substrates. The twelve lipases were chosen as the catalysts to accomplish the kinetic resolution using sing acetate 5a as substrate and a preliminary screening (24 h, phosphate buffer pH 7.0 with 20% acetonitrile as co-solvent, 30 °C and enzyme: substrate ratio of 2:1, m:m) was achieved. Their results confirm that the Amano AK lipase from Pseudomonas fluorescens and lipase from Thermomyces lanuginosus (TLL) immobilized on Immobead 150 present better catalytic ability as compared with the other lipases. Under above-mentioned conditions, both the (R)-1-(2,6-dimethylphenoxy)propan-2-ol [(R)-4a] and the remaining (S)-1-(2,6-dimethylphenoxy)propan-2-yl acetate [(S)-5a] were obtained with enantiomeric excess (ee) > 99%, 50% conversion and enantiomeric ratio (E) > 200. The kinetic resolution of racemic acetates 5b-d was surveyed, leading to the corresponding chiral remaining acetates with ≥ 95% ee, and the alcohols 4b-d with ≥ 98% ee, and conversion values close to 50%. The experimental data confirm that lipase from P. fluorescens or TLL immobilized on Immobead 150 present the best catalytic ability when the kinetic resolution reaction was induced for 24 or 48 h at 30 °C. It is demonstrate that these intermediates had their absolute configurations determined using 1H NMR spectroscopy (Mosher's method), showing that the KRs of these acetates obeyed the Kazlauskas’rule. Moreover, molecular docking studies corroborated the experimental results, indicating a preference for the hydrolysis of (R)-5a-d.

In this work, lipase kinetic resolution was carried out at 30 ^oC. The four lipase derivatives from Thermomyces lanuginosus (TLL) were adopted to achieve the kinetic resolution reaction. Maybe, the four lipase derivatives from Thermomyces lanuginosus (TLL) shall be employed in the thermophilic reaction. Please explain the reason why the four lipase derivatives from Thermomyces lanuginosus (TLL) were used to accomplish a kinetic resolution reaction 30 ^oC.

Author Response

We are grateful for reviewer 1's comments and suggestions.

In this work, lipase kinetic resolution was carried out at 30 ^oC. The four lipase derivatives from Thermomyces lanuginosus (TLL) were adopted to achieve the kinetic resolution reaction. Maybe, the four lipase derivatives from Thermomyces lanuginosus (TLL) shall be employed in the thermophilic reaction. Please explain the reason why the four lipase derivatives from Thermomyces lanuginosus (TLL) were used to accomplish a kinetic resolution reaction 30 ^oC.

 

Response: Initially, we had commercial TLL immobilized on Immobead 150 (macroporous support) available in our laboratory. As Professor Luciana R. B. Gonçalves' group was immobilizing TLL in iron nanoparticles, we were interested in evaluating the performance of these derivatives in the kinetic resolution of acetate rac-5a. Immobilization on iron nanoparticles allows the enzyme to be immobilized on the surface, which could reduce diffusion restrictions. However, the performance of these TLL derivatives immobilized on iron nanoparticles was inferior when compared to commercial TLL immobilized on Immobead 150.

 

Author Response File: Author Response.pdf

Reviewer 2 Report

The paper presents a well-designed and nicely presented study on kinetic resolution. I like the combination of the applied approach combined with docking study, which explains the performance of lipases. I mostly have technical comments, which are attached to the review. After the introduction of those corrections, I think the paper should be published without further delay. 

I also suggest adding to the SI representative PDBs of the docked structures - it would help in any follow-up of the study. 

Comments for author File: Comments.pdf

Author Response

Reviewer 2:

We are grateful for reviewer 2's comments and suggestions.

1. Table 1: …in that case please indicate which is which- especially as the conversion is built from this…”

Response: As suggested by reviewer 2, the following information was inserted at the bottom of Table 1: ee_s: enantiomeric excess of the remaining substrate (S)-5a; ee_p: enantiomeric excess of product (R)-4a;

 

2. 7, line 248: Is the reaction time?

Response: "15 min" was replaced by "15 min of reaction time".

 

3. 8, lines 260-261: This is a very interesting result. The enzyme activity decreases but also its enantioselectivity. Does the author have any hypotesis on what is going on (especially taking into consideration that the enzyme is immobilized).

Response: Although the TLL used is an immobilized enzyme, we observed that with each cycle of reuse the beads slightly changed their shape, indicating a possible degree of leaching, which could explain the change in enantioselectivity of the enzyme.

 

4. 11, line 366: poses or conformations? In docking studies we usually refer to conformation of the receptor and poses of the ligand to avoid confusion.

Response: Thank you for this observation. When dealing with docking results, the word “conformation(s)” was replaced by “pose(s)” along the text.

 

5. 11, line 369: please indicate which atoms were evaluate in the distance criterium - I presume it was O of serine and C of carbonyl atom in but this must be apparent to the reader - this should not be decribed only in the table footnote. “…and angles of 84-110° (same here)...”

Response: This sentence was rephrased to: “The NAC cutoff parameters represent the ideal distances and angles for the hydrolysis in the lipase catalytic site and were defined based on literature as: atom distance between catalytic serine (Ser-O) and the electrophilic carbon of the acyl group (Carbonyl-C) ≤ 4 Å, and attack angles between Ser-O/carbonyl-C/carbonyl-O in the range of 84-110° [43,44,51,52,53].”

 

6. 11, line 373: so the NAC refereed to contact in oxyanin hole? you see I am confused as NAC usually refers to the residue which catalyzes the actual atom transfer... I suggest rethinkig this paragraph.

Response: This sentence was rephrased to: “Interestingly, among selected results, some docking poses ex-hibited a oxyanion suitable for the stabilization of the substrate in the active site of the lipases, forming hydrogen bond with the carbonyl oxygen, and contributing to the decrease in the activation energy involved in the formation of the enzyme-substrate complex.”

 

7. 13, figure 3: the figs are in a bit too low resolutions and labes are not clear. There is no point in showing values of distances to the third decimal point as crystalographic structures have Res of approx 1-2 A. Increase the font size and try to produce higher density.

 

Response: The quality of this figure was improved; decimal point was fixed and font size increased.

 

8. 13, figure 3: I also wander how the fact that Ser is protonated in the docking studies influenses the result. It seems that not in all poses presented here Ser is making H-bond with His and instead it forms H-bond with the substrate (which would render the enzyme inhibited as Ser will.

Response:  We agree with reviewer 2's comment. A hydrogen bond between the serine amino acid residue and the substrate could lead to an inhibition of the enzyme. The experimental results show that the kinetic resolution of rac-5d only reached values close to 50% after 48 h of reaction. Therefore, the observed hydrogen bonding between serine from TLL and acetate rac-5d could be the reason for slower catalysis.

9. 14, lines 435-436: Lack of verb - consider correcting; please provide some additional information on the type of lipase used (either collection number of T. lanuginosus or sequence codes of lipases if they were identified.

Response: We acknowledge the reviewer comment. Indeed, the sentence lacks a verb and more details on the enzyme. New sentence in the manuscript: “Lipase from Thermomyces lanuginosus (Lipozyme® TL 100 L, Novozymes, Denmark) was immobilized on superparamagnetic iron oxide (Fe₃O₄) nanoparticles (SPMN)”.

10. 15, line 455: Daicel company should be mentioned…. and..Although I do not think Daicel is producing chiralpak with different bead radius this should be mentioned here for future reference (if they decide to introduce different sizes still retaining size of the column).

Response: As suggested by reviewer 2, the Daicel company name has been entered, as well as the bead radius: “…chiral column Chiralpak^® IA from Daicel company (150 mm x 4.6 mm x 5μm).”

11. 15, lines 456-461: Please add information on temperature of the separation and injection volume; as no chromatograms are shown in the main manuscript consider adding resolution value for the separated enantiomer pairs (here and in GC-FID as well).

Response: As requested by the reviewer, the temperature used in the HPLC was 35°C and the injection volume was 20 μL. Concerning the GC/FID analysis, the temperatures at which the enantiomer separations occurred were provided, as well as the injection volume (1 μL). Additionally, the resolution values (Rs) for the separation of each pair of enantiomers in both manuscripts and support information have been provided.

“High-performance liquid chromatography (HPLC) analyses were carried out in a Shimadzu chromatograph model LC solution 20A using a chiral column Chiralpak^® IA from Daicel company (150 mm x 4.6 mm x 5μm) for rac-4b and rac-5b with oven set at 35 °C, injection volume: 20 μL. Reaction time related to flow was 0.3 mL/min using hexane/IPA (98/2) as eluent and UV detector at 254 nm. Resolution value (Rs) for rac-4b: 1.5 and Rs for rac-5b: 2.3. Retention times were: (R)-4b 22.2 min; (S)-4b 25.3 min; (R)-5b 7.1 min; (S)-5b 8.4 min. Chiral column Chiralpak^® AS-H from Daicel company (150 mm x 4.6 mm x 5μm) for rac-4d and rac-5d with oven set at 35 °C, injection volume: 20 μL. Rs for rac-4d: 1.6 and Rs for rac-5d: 4.7. Reaction time related to flow was 0.3 mL/min using hexane/IPA (98/2) and UV detector at 254 nm. Retention times were: (R)-4d 33.3 min; (S)-4d 29.3 min; (R)-5d 12.0 min; (S)-5d 13.5 min. Gas chromatography (GC) analysis was carried out with a Shimadzu chromatograph model GC 2010 with a flame ionization detector, using a CP-Chirasil-Dex chiral column (25 m × 0.25 mm × 0.25 μm, 0.5 bar N₂). Injection volume: 1 μL. Rs for rac-4a: 2.0 and Rs for rac-5a: 3.3. For monitoring the time courses of reactions: rac-4a: 100 °C; 0.5 °C/min; 130 °C (hold 15 min); 5 °C/min 140 °C (hold 5 min). Retention times were: (R)-4a 62.8 min; (S)-4a 62.3 min (temperature of separation: 130 °C); rac-5a: 100 °C; 0.5 °C/min; 130 °C (hold 15 min); 5 °C/min 140 °C (hold 5 min). Retention times were: (R)-5a 73.6 min; (S)-5a 71.0 min (temperature of separation: 130 °C). For rac-4c: 120 °C; 1.0 °C/min; 135 °C (hold 5 min); 0.5 °C/min 150 °C (hold 5 min). Rs for rac-4c: 2.7 and Rs for rac-5c: 4.5. Retention times were: (R)-4c 14.8 min (temperature of separation: 134.8 °C); (S)-4c 14.6 min (temperature of separation: 134.6 °C). For rac-5c: 120 °C; 1.0 °C/min; 135 °C (hold 5 min); 0.5 °C/min 150 °C (hold 5 min). Retention times were: (R)-5c 19.0 min; (S)-5c 17.1 min (temperature of separation: 135 °C).”

12. 15, lines 475-479: It is not clear to me if the provided procedure is just a short version of that delivered in 54 or a new one just for covalent immobilization on functionalized SPMN with GA. If the latter please expand it to indicate the whole procedure, with endcapping (if done) and washing. Here the procedure is descried in a very concise manner. I may be mistaken but it seems not all details are provided.

Response: In fact, the TLL immobilization protocol was exactly the one described in reference 54. To make it clear in the text, we introduced the following sentence: “Lipase from Thermomyces lanuginosus (TLL) was immobilized onto superparamagnetic nanoparticle (SPMN) by ionic interaction or covalent binding, according to the protocol previously described [54]. Briefly, both the adsorbed and covalent immobilized biocatalysts were obtained by contacting an enzyme solution and the activated support (3 mg of protein per gram of SPMN) at pH 7 and 25 °C, for 1h. For the covalent immobilization, 0.01% (v/v) CTAB was added to solution containing lipase to avoid adsorption.”

13. Item 3.5, p.16, line 535: First of all be precise on the type of phosphate buffer (Na, K, mixed). Secondly, PO₄ ^3- no t 2^- not ^-2 as well.

Response: As requested by reviewer 2, corrections were made to the text:”… potassium phosphate buffer (PO₄^3-)…”

14. Item 3.9, p.16, line 536: Please indicate if it was liophilized lipase or liquid (and if latter what concetration - I presume it was a liophilized as mass was provided) However as the immoblized emzes were used as well the simple mass is not a very precise parameter. Later I can see there is a procedure for Immobead150 supported enzyme but the reader may have difficulty in comparing those experiments. I would compare relative activity of those catalyst in the same reaction to let the reader compare those reactors.

Response: In fact, we used an enzyme/substrate mass ratio. In all cases, either the lipase was lyophilized or immobilized.

15. 17, line 565: binder???

Response:  The word “binders” was replaced by “any small molecule present in the crystallographic structures”.

 

16. 17, lines 566-571: What pH was selected; how authors treated His residues - if some arbitrary decisions on protonation of His were undertaken please list them here.

 

Response: This sentence was rephrased to: “Both structures were protonated at pH 7.0, using the PROPKA code in the PDB2PQR webserver [56], which suggested that the ND1 atom of His312 is protonated at indicated pH. The enzymatic mechanism for TLL and P. fluorescens lipase involves a catalytic triad formed by Ser146, Asp201, His258 and Ser206, Asp254, His312, respectively. For histidine residues, the NE2 atom forms a hydrogen bond with the side chain hydroxyl group of the catalytic serine, which can be both a hydrogen bond acceptor and donor. The ND1 atom of this same residue makes a hydrogen bond with OD2 of Aspartate.”

 

17, line 576: I understand that water has not penetrated the active site as no MD was used.”

Response: In both crystallographic structures, there is no water in the active site. In these simulations, water was used to best mimic the environment during energy minimization. No changes were made in the main text.

 

Author Response File: Author Response.pdf