Toward the Identification of Potential α-Ketoamide Covalent Inhibitors for SARS-CoV-2 Main Protease: Fragment-Based Drug Design and MM-PBSA Calculations
, Mohamed Fares 2,3, Mohammed K. Abdel-Hamid Amin 4, Sara T. Al-Rashood 5,*, Amal Alharbi 5, Razan O. Eskandrani 5, Hamad M. Alkahtani 5 and Wagdy M. Eldehna 6Round 1
Reviewer 1 Report
The manuscript entitled "Toward the Identification of Potential α-ketoamide Covalent Inhibitors for SARS-COV2 Main Protease: Fragment-BasedD rug Design and MM-PBSA Calculations" by Mahmoud A. El Hassab et al. describes an interesting computer-aided approach in order to find a new anti-COVID agent acting as SARS-COV2 Mpro inhibitor.
Authors combined FBDD with SBDD methods by the use of modern computational algorithms available in the high-level software widely advised in computer-aided drug design. As result, structure of compound RMH148 has been found as potential active Mpro inhibitor promising for future therapy of COVID.
In my opinion this idea is interesting and "trendy", taking into account the present pandemic state, and the computational studies are correct in light of current CADD requirements.
Thus, I recommend this manuscript for Processes journal. Several issues, however, should be addressed and modified before publishing, as follows:
1.The FBDD was performed theoretically only, although the crystal of Mpro was obtained experimentally (PDB 6y2g). Exeperimental FBDD assays would be useful to confirm the results of molecular modelling.
2. RMH148 was indicated as promising anti-COVID agent but its structure is also purely theoretical. Is this compound obtainable via chemical or biotechnological synthesis? There are a number of computational methods with application of bases for commercially available chemicals that allow to estimate this. Authors should add such considerations when they propose this structure of Mpro inhibitor. Without this, the compound and all scientific efforts can be only hypothetical.
3. In introduction Authors have written:
"Extensive studies investigating the SARS COV-2 life cycle revealed many potential targets for drug discovery against the Covid-19 infection including the angiotensin-converting enzyme II (ACE2) entry receptor, the main protease (Mpro), and the RNA-dependent RNA polymerase (RdRp), Fig 1"
However, there is no RdRp shown in the Fig.1. The picture should be modified in this field.
4. Molecular mechanisms described in lines 72-77 (Introduction) should be supported with an appropriate scheme or figure.
Author Response
Many thanks for the reviewers’ comments which are all valuable and very helpful for revising and improving our manuscript. The manuscript was revised according to reviewers’ comments. The changes are listed below and highlighted yellow in the revised manuscript.
Reviewer #1:
- The FBDD was performed theoretically only, although the crystal of Mpro was obtained experimentally (PDB 6y2g). Experimental FBDD assays would be useful to confirm the results of molecular modelling.
Response: we agree with the reviewer opinion that the application of experimental FBDD assays would be useful to confirm the results of molecular modelling. However, the current circumstances have forced all the laboratories to shut down and thus preventing us from performing any experimental assays.
- RMH148 was indicated as promising anti-COVID agent but its structure is also purely theoretical. Is this compound obtainable via chemical or biotechnological synthesis? There are a number of computational methods with application of bases for commercially available chemicals that allow to estimate this. Authors should add such considerations when they propose this structure of Mpro inhibitor. Without this, the compound and all scientific efforts can be only hypothetical.
Response: We thank the reviewer for raising such important point, and of course our future plan is to synthesize, develop and evaluate RMH148. We also included the synthetic score for RMH148 and our future prospects for the compound.
- In introduction Authors have written: "Extensive studies investigating the SARS COV-2 life cycle revealed many potential targets for drug discovery against the Covid-19 infection including the angiotensin-converting enzyme II (ACE2) entry receptor, the main protease (Mpro), and the RNA-dependent RNA polymerase (RdRp), Fig 1"
However, there is no RdRp shown in the Fig.1. The picture should be modified in this field.
Response: Figure 1 was modified as requested by the reviewer.
- Molecular mechanisms described in lines 72-77 (Introduction) should be supported with an appropriate scheme or figure.
Response: The reviewer suggestion is highly appreciated. An appropriate scheme was provided in the supporting information as the reviewer suggested .
Reviewer 2 Report
The manuscript by El Hassab et al reports a computer aided drug design investigation on the SARS-CoV-2 Main protease. There are some important aspects of the description of the methods (MD) that must be improved. But most importantly, the authors do not provide enough evidence and justification of the importance of their findings. Therefore, these aspects have to be improved much more for a favorable recommendation. In addition, there are a number of minor aspects that should be addressed.
For all these reasons, my recommendation currently is reconsideration after major revision.
Major aspects:
- The major issue is that the authors do not transmit the importance of their findings. What is the contribution of their research in regards to other ligands that have been proposed for the SARS-CoV-2 MPro? The authors should compare the binding mode and binding energy with other reported inhibitors. Is their ligand binding much more strongly than other covalent ligands so that the probability for the covalent inhibition are larger?
- Can the authors use some ADMET prediction on their ligand? Also compare it to the binding affinity and ADMET properties of other experimentally tested inhibitors of SARS-CoV-2. Is the RMH148 showing binding and ADMET properties that make it more promising than the existing ligands?
- There is little information extracted from the MD simulations. It would be interesting to include in the main article a minimal description of the behavior of the ligand. Are all the interactions observed in the docked complex maintained during the dynamics? Is there any relaxation of the protein or ligand structure? Any relevant interactions with the solvent?
- It would be also interesting to characterize what are the structural differences between the inhibitor-free and inhibitor-bound proteins that imply such large RMSD values.
- There are a number of deficiencies in the description of the MD simulations:
- The protocol of the system building of the free protein, and protein-ligand compounds must be included in the manuscript. The authors indicate a reference to other publication instead. Furthermore, the authors indicate that they used the GROMACS software and the referenced publication is about a different MD software (NAMD).
- Specify clearly which is the force field used for the ligands and how they were parametrized.
- Indicate which ions were added (cations, anions and which type)
- It seems that the authors only neutralized the system rather than using the proper concentration of ions to reproduce the physiological ionic strength. This is a common but important mistake, as the presence of ions can cause major structure and dynamical changes in proteins.
- Indicate the size of the simulation box and the number of water molecules employed in the simulations.
- Please indicate which was the pressure employed in the NPT and production simulations.
- Indicate and cite properly the thermostat and barostat employed to maintain constant T and P in the MD simulations.
Minor aspects:
- It is not clear whether the MD simulations were performed on the MPro dimer or the monomer.
- (P7) The authors indicate that the high RMSD value observed for the inhibitor free Mpro indicates high flexibility of the ligand. This is not necessarily true. High RMSD values imply significant difference between the reference and the aligned structure. Even if the structure is not very dynamic, it will show large RMSD if it is very different. Instead, the flexibility understood in a dynamical sense should be measured in terms of RMSD fluctuations, or the RMSF root mean square fluctuations (i.e. oscillations around an equilibrium value no matter this is large or small). Please correct.
- The quality of Figure 5 should be improved. There is no need for a key in each plot, but the magnitude and units (RMSD (Å)) should be indicated in the y-axes of the plots.
- Please, report the MM-PBSA energies (all terms) in Table 1 with only 1 significant figure for the standard deviation and report the correspondent energy terms with coherent significant figures. For example, change -420.73 ±33 to (-4.2±0.2)·10^2 or -37.47±2.28 to -37±2.
- Figure 4 is not clear. It is difficult to identify the residues and to tell them apart from the interaction labels. Please improve.
Author Response
Reviewer #2:
The authors are grateful for reviewer2 for his/her accurate revision and professional suggestions to enhance our manuscript.
Major aspects:
- The major issue is that the authors do not transmit the importance of their findings. What is the contribution of their research in regards to other ligands that have been proposed for the SARS-CoV-2 MPro? The authors should compare the binding mode and binding energy with other reported inhibitors. Is their ligand binding much more strongly than other covalent ligands so that the probability for the covalent inhibition are larger?
Response: the author included a separate section to analyze the advantages and disadvantages of RMH 148 in comparison to other reported inhibitors.
- Can the authors use some ADMET prediction on their ligand? Also compare it to the binding affinity and ADMET properties of other experimentally tested inhibitors of SARS-CoV-2. Is the RMH148 showing binding and ADMET properties that make it more promising than the existing ligands?
Response: ADMET prediction was provided as the reviewer requested.
- There is little information extracted from the MD simulations. It would be interesting to include in the main article a minimal description of the behavior of the ligand. Are all the interactions observed in the docked complex maintained during the dynamics? Is there any relaxation of the protein or ligand structure? Any relevant interactions with the solvent?
Response: all the requested data was provided in the main article.
- It would be also interesting to characterize what are the structural differences between the inhibitor-free and inhibitor-bound proteins that imply such large RMSD values.
Response: five conformational sampling for the inhibitor-free and inhibitor-bound proteins were provided to characterize the structural differences between for the inhibitor-free and inhibitor-bound proteins
- There are a number of deficiencies in the description of the MD simulations:
- The protocol of the system building of the free protein, and protein-ligand compounds must be included in the manuscript. The authors indicate a reference to other publication instead. Furthermore, the authors indicate that they used the GROMACS software and the referenced publication is about a different MD software (NAMD).
Response: the protocol of the system building was provided as well as a proper reference.
- Specify clearly which is the force field used for the ligands and how they were parametrized.
Response: all the requested data was provided in the main article.
- Indicate which ions were added (cations, anions and which type)
Response: all the requested data was provided in the main article.
- It seems that the authors only neutralized the system rather than using the proper concentration of ions to reproduce the physiological ionic strength. This is a common but important mistake, as the presence of ions can cause major structure and dynamical changes in proteins
Response: in the GROMACS simulation, adding proper ions is done by standard commands that apply for all types of simulations especially when no specific conditions is reported.
- Indicate the size of the simulation box and the number of water molecules employed in the simulations.
Response: all the requested data was provided in the main article.
- Please indicate which was the pressure employed in the NPT and production simulations.
Response: all the requested data was provided in the main article.
- Indicate and cite properly the thermostat and barostat employed to maintain constant T and P in the MD simulations.
Response: all the requested data was provided in the main article.
Minor aspects:
- It is not clear whether the MD simulations were performed on the MPro dimer or the monomer.
Response: MD simulations were performed on the MPro monomer.
- (P7) The authors indicate that the high RMSD value observed for the inhibitor free Mpro indicates high flexibility of the ligand. This is not necessarily true. High RMSD values imply significant difference between the reference and the aligned structure. Even if the structure is not very dynamic, it will show large RMSD if it is very different. Instead, the flexibility understood in a dynamical sense should be measured in terms of RMSD fluctuations, or the RMSF root mean square fluctuations (i.e. oscillations around an equilibrium value no matter this is large or small). Please correct.
Response: The reviewer suggestion is highly appreciated. RMSF root mean square fluctuations were provided for all the aligned structures.
- The quality of Figure 5 should be improved. There is no need for a key in each plot, but the magnitude and units (RMSD (Å)) should be indicated in the y-axes of the plots.
Response: the magnitude and units (RMSD (Å)) was indicated in the y-axes of the plots
- Please, report the MM-PBSA energies (all terms) in Table 1 with only 1 significant figure for the standard deviation and report the correspondent energy terms with coherent significant figures. For example, change -420.73 ±33 to (-4.2±0.2)·10^2 or -37.47±2.28 to -37±2.
Response: MM-PBSA energies (all terms) in Table 1 was corrected as the reviewer suggested.
- Figure 4 is not clear. It is difficult to identify the residues and to tell them apart from the interaction labels. Please improve.
Response: Figure 4 was improved as the reviewer requested.
Reviewer 3 Report
In this manuscript the authors describe the application of a structure based drug design protocol for the design of new potential SARS-COV2 Mpro inhibitor. In particular, the authors have applied fragment-based drug design (FBDD), covalent docking and molecular dynamics (MD) simulations towards the identification of the α-ketoamide Mpro inhibitor RMH148.
Overall, the manuscript is clear and well presented. The applied methodology is well described and the results are interesting. Based on these, the authors claim that compound RMH148 binds to the Mpro and forms a more stable complex in comparison with the crystal α-ketoamide inhibitor O6K. The whole study could be considered as the starting point of an extensive campaign for the discovery of new Mpro inhibitors which will incorporate structural features of compound RMH148.
Given that this is a computational study, I wonder, if the authors plan the synthesis of the identified compound RMH148 and its preliminary biological evaluation as a Mpro inhibitor.
Some minor corrections are indicated in the attached pdf file.
On this basis, I recommend publication of the present after a minor revision.
Author Response
Reviewer #3:
The authors are grateful to reviewer3 for his/her accurate revision and supportive suggestions for our manuscript.
