Altering the Regioselectivity of T1 Lipase from Geobacillus zalihae toward sn-3 Acylglycerol Using a Rational Design Approach

Round 1

Reviewer 1 Report

The manuscript entitled “Altering the Regioselectivity of T1 lipase from Geobacillus zalihae toward sn-3 acylglycerol by a rational design approach” by Albayati and co-authors constructed a modified regiospecific sn-3 T1 lipase (3M lipase), and this provided insight into the oxyanion hole and lid domain structure /function relationship. Authors have cloned and expressed T1 lipase. Further authors have conducted experimental and computational study to explore the information of catalytic site and regiospecificity and preference toward long-chain selectivity. I have several that need to answered prior to acceptance of the manuscript:-

1.    Authors have predicted the model of 3M. But authors have not validated the predicted model. Authors should validate the model by Ramachandra plot, VERIFY-3D, ERRAT, and ProQ etc. Authors can refer the following article for further details:- (PUBMED: 35475370)

2.    I don’t understand why authors have modelled the M3 protein using PDB ID: 2W22 which is low resolution as compared to 2DSN. Authors can just perform the MD simulation of 2DSN structure and retrieve the open conformation from MD simulation. Later, authors can mutate the amino acid residues in COOT or PyMOL. In this way, model will be more reliable and give actual results in comparison to wt-1T. Authors can refer the following article for further information about generation of open conformation and mutation: (PUBMED: 35475370)

3.    Authors have not provided the binding affinities of OCP with wt-T1 and 3M variant.

4.    Even for molecular docking, author should perform the molecular docking of OCP with wild type open confirmation and mutant. By this approach, authors can get the exact orientation and binding affinities difference between OCP docked in wild type open conformation protein and mutant. Authors can refer the following article to get the idea of comparison of molecular docking results of open conformation wild type and mutant: (PUBMED: 35475370)

5.    Authors should provide the detailed information of molecular docking like grid size, grid dimensions, kohhlman charges on protein, and gasteiger charges on ligand etc.

6.    Authors have mentioned in 2.5.2 section that MD simulation results are in Figure S6 while in supplementary file it is a Figure S5. Authors forgot to attach the Figure S5 which is superposition of wt-T1 and 3M variant structure.

7.    In Figure S6, MD analysis results showed that 3M is less stable and compact than wt-T1. Does the authors have performed the MD analysis in triplicate ?

8.    RMSF of 3M and wt-T1 are most comparable that suggest that both structures are almost same while in modeling section (2.5.1) authors have mentioned that 3M has an RMSD of 2.874 A with wt-1T. Both these results are not in agreement to each other.  

9.    Authors reported that 3M displayed the regiospecficity. So authors should show the snapshots of MD simulation retrieved at different time intervals to illustrate the orientation of OCP with 3M.

Due to my above comments, I recommend major revision of the manuscript.

Author Response

Please check the PDF file.

Author Response File: Author Response.pdf

Reviewer 2 Report

This work contains some interesting results about the regioselectivity and substrate chain-length specificity of T1 lipase from Geobacillus zalihae. Unfortunately, the low quality of the writing makes the article difficult to follow.

Abstract is clear about goal and results of the study but Introduction can be improved. In the Introduction it is not clearly described what is known about the regioselectivity and chain-length specificity of related lipases and what kind of experiments have been performed to understand/change these properties. In other words, it should be better explained what is the current state-of-the-art and what is really new here.

Results and Discussion:

Line 114: it is strange that these data are not shown. I recommend to provide the activity and regioselectivity data of the 12 variants.

Fig. 2: what is the error in the activity values? Shift in pH optimum is not very convincing when only providing U/mL specific activities (kcat and Km values are needed for a better interpretation). Furthermore, a pH interval of 1 pH unit is quite large. 

173: Does a Tm value represent the thermodynamic properties of an enzyme? Be more precise in your writing.

The explanation for the change in Tm value and change in secondary  structure is quite vague.

190: Values for specific activity (U/mg), Vmax (U/mL) and kcat (s-1) do not correspond. Please check carefully.

Fig. 3: specific activities need to presented as U/mg

Fig. 3, Fig. 4, Fig. S2, Fig. S3, Fig. S4: To better understand the effects of the lid and oxyanion hole mutations, the properties of F180G/F181S/F16 need to be presented as a reference.

Fig. S3a and S3b: Which wavelength was used to measure the change in CD signal?

Fig. S3c: The CD signals need to be presented as mean residue ellipticities. The shape of the CD spectra of Wt and M3 is the same, suggesting similar alpha-helix and beta-sheet content. The reported changes in alpha-helix and beta-sheet content need to be discussed more critically. How sure is it that the protein concentrations were the same? 

Fig. S4: very poor correlation coefficients Lineweaver-Burk plots. Please show Michaelis-Menten graphs with non-linear regression data analysis and discuss the deviation from MM-kinetics in a proper way.

What is the reliability (reproducibility) of these experiments? There is no error indicated in the data points of Fig. S4.

Fig. 7 Indicate which protein is shown here.

General: several times reference is given to Baeyer-Villiger monooxygenases or CytP450 enzymes to support the use of site-directed mutagenesis (SDM) for changing the regioselectivity of an enzyme. There is no doubt that SDM is a powerful tool for achieving this goal.  However, the mode of action of lipases is very different from BVMOs and P450s. Therefore, it would be better to give examples of enzymes that are more closely related in structure and function.

Author Response

Please check the PDF file.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The revised manuscript entitled “Altering the Regioselectivity of T1 lipase from Geobacillus zalihae toward sn-3 acylglycerol by a rational design approach” is improved than earlier version. Although the current version of manuscript still has several issues than need to be addressed prior to acceptance of the manuscript. I recommend a major revision of the manuscript. Here are my comments/suggestions for the manuscript:-

1.    I did not find any file having authors responses for my comments.

2.    In Section 2.5.1, authors written that “The open structure of lipases (3M and wt-T1) was developed using available crystal structure of BTL2 lipase (PDB ID: 2W22) as a template.” While, in section 3.7, authors written that “The 3D structure of T1 lipase (PDB ID: 2DSN) deposited in the RSB Protein Data Bank (PDB) database is in close conformation. The wt-T1 and 3M lipases were remodelled using BTL2 lipase (PDB ID: 2W22) as a template with an open conformation and sequence identity of more than 96% with wt-T1 lipase.” So, it is confusing in manuscript that whether authors have considered 2W22 or 2DSN as a template to predict the 3M and wt-T1.

3.    In section 3.7, it is not clear how authors predict the models, as authors written that “FASTA file of T1 lipase (2DSN) was obtained from 687 PDB, and the 2W22.pdb file was uploaded and modelled using the command file 688 hm_build.mcr.” How authors can predict the model from 2DSN just by consideration of its sequence.

4.    In revised version also figures still required improvement. For example, in figure 6, amino acid residues labelling is not clear.

5.    From Figure 8 captions, it is very difficult to understand comparison of MD simulation trajectories of wt-T1 and 3M variant. It is very hard to understand whether authors have compared the MD simulation results of wt-T1 or 3M variant. Authors can just superpose MD simulaton trajectories of wt-T1 with model of wt-T1 while in other figures compare simulation trajectories of 3M variant with model of 3M variant.

6.    In the revised manuscript also, authors have not provided the binding affinities of OCP with wt-T1 and 3M variant.

Author Response

Reviewer 1

The revised manuscript entitled “Altering the Regioselectivity of T1 lipase from Geobacillus zalihae toward sn-3 acylglycerol by a rational design approach” is improved than earlier version. Although the current version of manuscript still has several issues than need to be addressed prior to acceptance of the manuscript. I recommend a major revision of the manuscript. Here are my comments/suggestions for the manuscript:-

1. I did not find any file having authors responses for my comments.

Response:

Dear reviewer. Due to the technical issue arise during the submission of the response file, here, we attached again the response to the first round.

2. In Section 2.5.1, authors written that “The open structure of lipases (3M and wt-T1) was developed using available crystal structure of BTL2 lipase (PDB ID: 2W22) as a template.” While, in section 3.7, authors written that “The 3D structure of T1 lipase (PDB ID: 2DSN) deposited in the RSB Protein Data Bank (PDB) database is in close conformation. The wt-T1 and 3M lipases were remodelled using BTL2 lipase (PDB ID: 2W22) as a template with an open conformation and sequence identity of more than 96% with wt-T1 lipase.” So, it is confusing in manuscript that whether authors have considered 2W22 or 2DSN as a template to predict the 3M and wt-T1.

Response: Line 601

‘The 3D structure of T1 lipase (PDB ID: 2DSN) deposited in the RSB Protein Data Bank (PDB) database is in close conformation.’ This sentence was deleted from the text to avoid confusion. Since we only used 2W22 to predict the structure, we decided to take out the 2DSN from the text, since they are no comparison carry out between close and open structure of T1 lipase.

3. In section 3.7, it is not clear how authors predict the models, as authors written that “FASTA file of T1 lipase (2DSN) was obtained from 687 PDB, and the 2W22.pdb file was uploaded and modelled using the command file 688 hm_build.mcr.” How authors can predict the model from 2DSN just by consideration of its sequence.

Response: line 601 - 608

Thank you for your valuable comment. We revised this section to avoid confusion. We are using YASARA software to run the structure modelling, docking analysis and molecular dynamics simulation. YASARA Structure features a complete homology modeling module that fully automatically takes all the steps from an amino acid sequence to a refined high-resolution model using a CASP (Critical Assessment of Structure Prediction) approved protocol. CASP which organized by the Prediction Center since 1994, is a biennial evaluation of today's many approaches to protein structure prediction. The predictions are thus real 'blind predictions', which makes CASP a unique opportunity.

4. In revised version also figures still required improvement. For example, in figure 6, amino acid residues labelling is not clear.

Response

The figure 6 has been revised as suggested

5. From Figure 8 captions, it is very difficult to understand comparison of MD simulation trajectories of wt-T1 and 3M variant. It is very hard to understand whether authors have compared the MD simulation results of wt-T1 or 3M variant. Authors can just superpose MD simulaton trajectories of wt-T1 with model of wt-T1 while in other figures compare simulation trajectories of 3M variant with model of 3M variant.

Response

Thank you for the comment but I find it is difficult to understand this question. In this study we conducted the experiment to compare wt-T1 with ligand and 3M with ligand. We compare both wt-T1 and 3M lipase with the OCP ligand. We did not perform simulation of wt-T1 closed structure since our objective is to study the MD simulation on open structure of wt-T1_OCP and 3M_OCP. So, there is only 2 MD simulation here, wt-T1_OCP and 3M_OCP. We re-write this section for clearer explanation and the superimposed of closed and open conformations were included in Figure 8.

6. In the revised manuscript also, authors have not provided the binding affinities of OCP with wt-T1 and 3M variant.

Response

The binding affinity was included in Table S3 as requested.

Reviewer 2 Report

The manuscript has improved.  English grammar still needs careful check.

Throughout the manuscript, K_cat needs to be changed into k_cat, because we are dealing with rate constants, not equilibrium constants.

Author Response

Reviewer 2

The manuscript has improved.  English grammar still needs careful check.

English editing

Throughout the manuscript, K_cat needs to be changed into k_cat, because we are dealing with rate constants, not equilibrium constants.

Response:

K_cat has been changed to k_cat