MD Simulation of Vector–Receptor Pharmacologic Pairs for Tumor-Specific Drug Delivery: Transfer of Boron Atoms by RGD Peptide to αvβ3 Integrin Receptor

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This manuscript presents a molecular dynamics (MD) simulation study exploring the interaction between RGD peptides and the αvβ3 integrin receptor, with the broader goal of modeling boron delivery for tumor-targeted applications such as boron neutron capture therapy (BNCT). The topic is conceptually relevant, as RGD-integrin interactions are well-established in targeted drug delivery, and computational approaches can provide atomistic insight into ligand-receptor recognition and transport mechanisms. The authors attempt to simulate both peptide binding and the association of boron atoms in single and clustered forms, and the idea of modeling a “vector-receptor” pharmacological system is potentially interesting. However, while the manuscript includes extensive MD simulations and numerous structural visualizations, the scientific rigor, methodological transparency, and depth of analysis are currently insufficient to support the strength of the conclusions. The study remains largely descriptive, with limited quantitative validation and significant overinterpretation of the computational results. Substantial revisions are required to improve clarity, reproducibility, and scientific impact.

1. The overall scientific rationale would benefit from clearer positioning and stronger justification. While the RGD-αvβ3 interaction is a well-known paradigm in targeted delivery, the novelty of simulating boron transport in this context is not convincingly articulated. The manuscript repeatedly frames the work as modeling a “vector-receptor” system, but this terminology is not clearly defined or distinguished from existing ligand-receptor modeling frameworks. In addition, the connection between the MD simulations performed and the practical implementation of BNCT remains speculative. The Introduction would be strengthened by more explicitly identifying the gap in current knowledge and clarifying how this study advances beyond prior computational or experimental work on RGD-integrin targeting.
2. The methodological section, although lengthy, lacks critical details necessary for reproducibility and proper evaluation. While the use of AMBER18, TIP3P water, and standard equilibration protocols is noted, key parameters are either missing or inconsistently described. For example, the force fields used for the protein, peptide, and especially the boron atoms or clusters are not clearly specified, which is a major concern given that boron-containing species require careful parameterization. It is also unclear how boron atoms are represented (as neutral particles, ions, or part of a defined chemical structure), which directly affects the validity of the simulations. Furthermore, the manuscript mentions both implicit and explicit solvent models being used (page 6), but does not clarify when or why each was applied. These ambiguities significantly limit the reproducibility and credibility of the computational setup.
3. The design of the simulation system itself raises conceptual concerns. The model includes two RGD peptides(one pre-bound and one freely diffusing) used to infer binding mechanisms . While this is an interesting setup, the rationale for this specific configuration is not sufficiently justified. It is unclear whether this scenario reflects a biologically realistic environment or an artificial construct. Additionally, the use of a single PDB structure (3ZE2) without discussion of receptor flexibility, conformational states, or potential missing regions limits the generalizability of the findings. The absence of replicate simulations or statistical validation further weakens confidence in the reported observations.
4. The Results section is largely descriptive and relies heavily on visual inspection of trajectories and structural snapshots (e.g., Figures 6-15 across pages 7-16). While these figures illustrate possible configurations and interactions, they are not accompanied by sufficient quantitative analysis. For example, although distance plots between peptides are shown (Figure 11, page 12), there is no comprehensive analysis of binding stability, interaction energies, hydrogen bonding patterns, or contact frequencies. Standard MD metrics such as RMSD, RMSF, radius of gyration, or binding free energy calculations (e.g., MM/PBSA) are notably absent. Without these quantitative measures, it is difficult to assess whether the observed interactions are stable, significant, or reproducible.

Author Response

Review Report Form
Reviewer-1

Open Review

( ) I would not like to sign my review report
(x) I would like to sign my review report

Quality of English Language

( ) The English could be improved to more clearly express the research.
(x) The English is fine and does not require any improvement.

 

	Yes	Can be improved	Must be improved	Not applicable
Does the introduction provide sufficient background and include all relevant references?	(x)	( )	( )	( )
Is the research design appropriate?	( )	(x)	( )	( )
Are the methods adequately described?	( )	(x)	( )	( )
Are the results clearly presented?	( )	(x)	( )	( )
Are the conclusions supported by the results?	( )	(x)	( )	( )
Are all figures and tables clear and well-presented?	(x)	( )	( )	( )

Comments and Suggestions for Authors

This manuscript presents a molecular dynamics (MD) simulation study exploring the interaction between RGD peptides and the αvβ3 integrin receptor, with the broader goal of modeling boron delivery for tumor-targeted applications such as boron neutron capture therapy (BNCT). The topic is conceptually relevant, as RGD-integrin interactions are well-established in targeted drug delivery, and computational approaches can provide atomistic insight into ligand-receptor recognition and transport mechanisms. The authors attempt to simulate both peptide binding and the association of boron atoms in single and clustered forms, and the idea of modeling a “vector-receptor” pharmacological system is potentially interesting. However, while the manuscript includes extensive MD simulations and numerous structural visualizations, the scientific rigor, methodological transparency, and depth of analysis are currently insufficient to support the strength of the conclusions. The study remains largely descriptive, with limited quantitative validation and significant overinterpretation of the computational results. Substantial revisions are required to improve clarity, reproducibility, and scientific impact.

Thanks. We really appreciate all the introductory words and comments outlined above by the Reviewer-1. Below we have answered all the questions from p.1 to p.4. But, first of all, we are very sorry that all the discussions that were done by three respected reviewers were actually conducted on the OLD version of our manuscript. Due to some technical issues and misunderstandings between the authors and the CIMB & Pharmaceutics Editorial Boards, it was not possible to upload the substantially corrected version of our manuscript with high-quality figures and text revisions. Thus, we have now downloaded the revised version of our manuscript thereby incorporating the second revisions that we have done for the Pharmaceutics into the current CIMB-version. We have now prepared several animation movies that make the things clearer for better science impact and improving reproducibility. Also, we have now used graphics software to represent the high-quality figures, instead of the previous snapshots, etc.

1. The overall scientific rationale would benefit from clearer positioning and stronger justification. While the RGD-αvβ3 interaction is a well-known paradigm in targeted delivery, the novelty of simulating boron transport in this context is not convincingly articulated. The manuscript repeatedly frames the work as modeling a “vector-receptor” system, but this terminology is not clearly defined or distinguished from existing ligand-receptor modeling frameworks. In addition, the connection between the MD simulations performed and the practical implementation of BNCT remains speculative. The Introduction would be strengthened by more explicitly identifying the gap in current knowledge and clarifying how this study advances beyond prior computational or experimental work on RGD-integrin targeting.

Thanks. We agree that the RGD-αvβ3 interaction is a well-known paradigm in targeted delivery, nevertheless, regarding on the novelty, there are few simulation works on the boron transport in this context. As for the term “vector-receptor” in modeling, we also used the more common terminology “ligand-receptor” throughout the manuscript. In order to better link the MD modeling and the practical implementation of BNCT now we have modified the text and places in the introductory part adding more descriptions and references regarding on the previous computational and experimental work in targeting RGD-integrins.

 

1. The methodological section, although lengthy, lacks critical details necessary for reproducibility and proper evaluation. While the use of AMBER18, TIP3P water, and standard equilibration protocols is noted, key parameters are either missing or inconsistently described. For example, the force fields used for the protein, peptide, and especially the boron atoms or clusters are not clearly specified, which is a major concern given that boron-containing species require careful parameterization. It is also unclear how boron atoms are represented (as neutral particles, ions, or part of a defined chemical structure), which directly affects the validity of the simulations. Furthermore, the manuscript mentions both implicit and explicit solvent models being used (page 6), but does not clarify when or why each was applied. These ambiguities significantly limit the reproducibility and credibility of the computational setup.

Thanks. Perhaps there are not much novelty regarding on the methodological part, we had just followed the standard MD simulation implementation with well-known software, that details are lot reported in the literature and also well based from the point of the reproducibility, analysis, proper evaluation. We have used the standard techniques (AMBER18 with TIP3P water and standard equilibration protocol) and the key parameters of the FF (force fields) used in this study are well reported; those are much similar to other protein, peptides, etc, modelling and virtual experiments. With regards to the boron atoms or clusters, as major concern outlined by the Referee-1, the boron-containing RGD-vector or ligand simulated with boron atoms that are represented as neutral particles. During multiple (perhaps dozens) MD simulations for RGD-2-peptide + boron atoms, it was necessary to ensure all natural vibrational and diffusion movements, but to preserve the position of boron atoms inside RGD-2-peptide. Thus, we have restrained (preserved) all the relative positions of the boron atoms bound to RGD-2-peptide during the entire dynamical and diffusional processes. So far, the relative arrangement of boron atoms bound to RGD-peptide has not been violated, since the main focus has been on the delivery of RGD-2 + boron(s) to the integrin + RGD-1 receptor. In particular, the force field used for boron atoms was approximated as boron containing ligand (RGD-2+boron), thereby disabling and modifying the prediction of the atom type and bond type, even if the initial FF parameters for the boron atom were not recognized in the running antechamber and related modules. Next, now we mention just explicit solvent model and the word “implicit” has been removed from the methodology part of the manuscript. It’s worth noting that the implicit solvent model maybe good for some systems, but it can give poor results for certain molecular shapes (boundary effects), etc. Now the additional animation movies in this study have been generated to demonstrate the “vector-receptor” (“ligand-receptor”) interactions in the explicit solvent that clarify the reproducibility and credibility of the computational setup.

 

1. The design of the simulation system itself raises conceptual concerns. The model includes two RGD peptides(one pre-bound and one freely diffusing) used to infer binding mechanisms . While this is an interesting setup, the rationale for this specific configuration is not sufficiently justified. It is unclear whether this scenario reflects a biologically realistic environment or an artificial construct. Additionally, the use of a single PDB structure (3ZE2) without discussion of receptor flexibility, conformational states, or potential missing regions limits the generalizability of the findings. The absence of replicate simulations or statistical validation further weakens confidence in the reported observations.

Thank you. Now we have added more descriptions of the purpose in yellow for the revised manuscript, and to answer question 3 correctly, we have now added several animated movies as Supplementary Material. Perhaps we have modeled both situations described by the respected Referee-1: an integrin receptor having no RGD (integrin with an RGD-free), as well as an RGD-1-bound integrin (an integrin associated with RGD as in the original PDB file). Yes, we used PDB structure 3ZE2 but the standard minimization and relaxation procedure are well established to ensure the recovering of the receptor flexibility and conformational states (on the 10-20 ns range, prior to large-scale 50-100 ns equilibrations stages). But the replicate (multiple) simulations were realized on each MD models, it’s wrong or misunderstanding by Reviewer-1 on the absence of replicate simulations. The MD results illustrate the accumulation of RGD-2 peptide in a specific integrin region, knowing that there are probably several more binding sites and more mechanisms for the existence of different binding sites between the vector and the receptor. However, the MD results clearly demonstrate one of the most likely binding mechanisms between RGD-2 and the receptor surface, although even the avß3 integrin receptor with or without RGD-1 were used. In the case of an integrin that does not contain RGD-1, the RGD-2 + boron(s) would definitely find much more room for incorporation into the receptor. Of the two RGD peptides that were modeled, RGD-1 remains embedded in the protein pocket during the entire simulation period of 50-100 nanoseconds, and the Animations and Figures illustrate the specific binding of RGD-2 in a natural way, since the binding residues of avß3 integrin to RGD-1 completely coincide with the initial experimental data, the experimentally X-ray measured PDB structure. Multiple (perhaps dozens) MD calculations of the distance between the vector (RGD-2-peptide + boron atoms) and the receptor (avß3-integrin) clearly demonstrate that the entire fragment of RGD-2+boron(s) is inserted into the receptor naturally. It is worth noting that we aimed at binding the entire RGD-2+boron fragment(s), rather than specifically binding selected residues and the receptor. The mechanism of dynamical interactions and diffusional movements between the RGD-2-peptide carrier and the avß3-integrin receptor is illustrated with an accuracy of millions of time steps and multiple configuration analysis.

1. The Results section is largely descriptive and relies heavily on visual inspection of trajectories and structural snapshots (e.g., Figures 6-15 across pages 7-16). While these figures illustrate possible configurations and interactions, they are not accompanied by sufficient quantitative analysis. For example, although distance plots between peptides are shown (Figure 11, page 12), there is no comprehensive analysis of binding stability, interaction energies, hydrogen bonding patterns, or contact frequencies. Standard MD metrics such as RMSD, RMSF, radius of gyration, or binding free energy calculations (e.g., MM/PBSA) are notably absent. Without these quantitative measures, it is difficult to assess whether the observed interactions are stable, significant, or reproducible.

Thanks for the comment. Now we have replaced Figures 5, 6a-6b-6c-6d, 7, 8a-8b-8c-8d, 9a-9b-9c-9d, 10, thereby essentially improved the figures for the best visualization, using the VMD (Visual Molecular Dynamics) and other graphical software. We have tried not to overload the content of the manuscript (say, mentioned by a respected Reviewer-1 in relation to the standard RMSD RMSF or other calculations), although the most likely interaction and dynamical effects have now been demonstrated through additional new Animated movies and Figures. Thus, during multiple (perhaps dozens) MD simulations for RGD-2-peptide + boron atoms, it was necessary to ensure all natural vibrational and diffusion movements, but to preserve the position of boron atoms inside RGD-2-peptide. Thus, we have preserved all the relative positions of the boron atoms bound to RGD-2-peptide during the entire dynamical and diffusional processes. So far, the relative arrangement of boron atoms bound to RGD-peptide has not been violated, since the main focus has been on the delivery of RGD-2 + boron(s) to the integrin + RGD-1 receptor. In particular, the force field used for boron atoms was approximated as neutral particles inside boron containing a ligand (RGD-2+boron), thereby disabling and modifying the prediction of the atom type and bond type, even if the initial FF parameters for the boron atom were not recognized in the running antechamber and related modules. 

 

Submission Date

16 March 2026

Date of this review

28 Mar 2026 21:36:35

 

Reviewer 2 Report

Comments and Suggestions for Authors

I have several suggestions and questions for the authors.

The table 1 is the same as the first part of the table 2. So, it is better to remove the table 1, as it provides no extra information.

The Figure 1 and some other figures are distorted, where the spheres look like ellipsoids in those figures. The authors can choose the right aspect ratios for those figures.

In line 154, the authors mentioned “tleap's jumprc.constph”, “jumprc.conste”, and “Sander”,…, which are amber files and programs. Not everyone is familiar with Amber package, so it is better to briefly explain those details in the manuscript.

The authors did not include the timestep that they used in their simulations in this manuscript (at least I have not found the time step in their manuscript). Since the time step is critical for their simulations, the authors must include the timestep in the section 2 of this manuscript.

The figures 6-10 look very strange to me. The resolution is low, the background color is weird. The authors should improve the image quality. Or, maybe there is something wrong with their pdf file conversion.

It is better to include an extra figure of potential energy vs. simulation time in the manuscript.

And one last thing, in line 155, the authors mentioned "a target of 303 K". Why? Why not just choose the  body temperature?

Author Response

Review Report Form
Reviewer-2

Open Review

( ) I would not like to sign my review report
(x) I would like to sign my review report

Quality of English Language

(x) The English could be improved to more clearly express the research.
( ) The English is fine and does not require any improvement.

 

	Yes	Can be improved	Must be improved	Not applicable
Does the introduction provide sufficient background and include all relevant references?	( )	(x)	( )	( )
Is the research design appropriate?	( )	(x)	( )	( )
Are the methods adequately described?	( )	(x)	( )	( )
Are the results clearly presented?	( )	(x)	( )	( )
Are the conclusions supported by the results?	( )	( )	(x)	( )
Are all figures and tables clear and well-presented?	( )	(x)	( )	( )

Comments and Suggestions for Authors

The article is devoted to molecular dynamics (MD) simulations to explore the interaction of the RGD peptide with the avb3 integrin receptor for targeted drug delivery to tumors.

Thanks. We really appreciate all comments outlined above by the Reviewer-2. Below we have answered all the questions, but, first of all, we are very sorry that all the discussions that were done by three respected reviewers were actually conducted on the OLD version of our manuscript. Due to some technical issues and misunderstandings between the authors and the CIMB & Pharmaceutics Editorial Boards, it was not possible to upload the substantially corrected version of our manuscript with high-quality figures and text revisions. Thus, we have now downloaded the revised version of our manuscript thereby incorporating the second revisions that we have done for the Pharmaceutics into the current CIMB-version. We have now prepared several animation movies that make the things clearer for better science impact and improving reproducibility. Also, we have now used graphics software to represent the high-quality figures, instead of the previous snapshots, etc.

Some remarks:

The role of boron is not disclosed in the introduction.

Thanks. Now we have improved the introduction and the entire text.

The introduction does not sufficiently disclose this RGD peptide vector system for specific delivery to the avb3 integrin receptor and describes the results and examples of targeted drug delivery through this receptor.

Thank you. We have added more descriptions for the revised manuscript; thereby we have now added several more Animated movies and high-quality Figures as Supplementary Materials. Perhaps we have modeled both situations: an integrin receptor having no RGD (integrin with an RGD-free), as well as an RGD-1-bound integrin (an integrin associated with RGD as in the original PDB file). The MD results illustrate the accumulation of RGD-2 peptide in a specific integrin region, knowing that there are probably several more binding sites and more mechanisms for the existence of different binding sites between the vector and the receptor. However, the MD results clearly demonstrate one of the most likely binding mechanisms between RGD-2 and the receptor surface, although even the avß3 integrin receptor with or without RGD-1 were used.

The results and examples of targeted drug delivery through this receptor and  the mechanism of howRGD-peptides bind to avb3 integrin receptor has already been described in the literature references [1-6]. There is no emphasis on what the authors have introduced in comparison with previous works.

Thanks. We agree that the RGD-αvβ3 interaction is a well-known paradigm in targeted delivery, nevertheless, regarding on the novelty, there are few simulation works on the boron transport in this context. In order to better link the MD modeling and the practical implementation now we have modified the text and places in the introductory and discussion parts adding more descriptions and references regarding on the previous computational and experimental work in literature.

Discussion: There is not enough overview here are not enough generalizing discussions , comparisons with literature results, and it is not entirely clear what was found out, there is no clear result and no clear conclusion of this study

It would be nice to have some experimental data to confirm or link to literary works and compare them with literary works. The links are largely to Russian works, it is advisable to add others.

Thank you. The links to Russian works has been reduced now to one cited paper and adding others. And now we have added more descriptions for the revised manuscript, and generated several Animated movies as Supplementary Material. Perhaps as like as in the experimental setup we have modeled both situations: an integrin receptor having no RGD (integrin with an RGD-free), as well as an RGD-1-bound integrin (an integrin associated with RGD as in the original PDB file). Yes, we used the experimental X-ray defined PDB structure 3ZE2 with the standard minimization and relaxation procedure are well established the receptor flexibility and conformational states (on the 10-20 ns range, prior to large-scale 50-100 ns equilibrations stages). We have performed replicate (multiple) simulations on each MD models. The MD results illustrate the accumulation of RGD-2 peptide in a specific integrin region, knowing that there are probably several more binding sites and more mechanisms for the existence of different binding sites between the vector and the receptor. However, the MD results clearly demonstrate one of the most likely binding mechanisms between RGD-2 and the receptor surface, although even the avß3 integrin receptor with or without RGD-1 were used. In the case of an integrin that does not contain RGD-1, the RGD-2 + boron(s) would definitely find much more room for incorporation into the receptor. Of the two RGD peptides that were modeled, RGD-1 remains embedded in the protein pocket during the entire simulation period of 50-100 nanoseconds, and the Animations and Figures illustrate the specific binding of RGD-2 in a natural way, since the binding residues of avß3 integrin to RGD-1 completely coincide with the initial experimental data, the experimentally X-ray measured PDB structure. Multiple (perhaps dozens) MD calculations of the distance between the vector (RGD-2-peptide + boron atoms) and the receptor (avß3-integrin) clearly demonstrate that the entire fragment of RGD-2+boron(s) is inserted into the receptor naturally. It is worth noting that we aimed at binding the entire RGD-2+boron fragment(s), rather than specifically binding selected residues and the receptor. The mechanism of dynamical interactions and diffusional movements between the RGD-2-peptide carrier and the avß3-integrin receptor is illustrated with an accuracy of millions of time steps and multiple configuration analysis.

 

Comments on the Quality of English Language

The English could be improved to more clearly express the research.

Submission Date

16 March 2026

Date of this review

30 Mar 2026 19:31:19

 

Reviewer 3 Report

Comments and Suggestions for Authors

The article is devoted to molecular dynamics (MD) simulations to explore the interaction of the RGD peptide with the avb3 integrin receptor for targeted drug delivery to tumors.

Some remarks:

The role of boron is not disclosed in the introduction.

The introduction does not sufficiently disclose this RGD peptide vector system for specific delivery to the avb3 integrin receptor and describes the results and examples of targeted drug delivery through this receptor.

The results and examples of targeted drug delivery through this receptor and  the mechanism of how   RGD-peptides bind to avb3 integrin receptor has already been described in the literature references [1-6]. There is no emphasis on what the authors have introduced in comparison with previous works.

Discussion: There is not enough overview here are not enough generalizing discussions , comparisons with literature results, and it is not entirely clear what was found out, there is no clear result and no clear conclusion of this study

It would be nice to have some experimental data to confirm or link to literary works and compare them with literary works. The links are largely to Russian works, it is advisable to add others.

 

Comments on the Quality of English Language

The English could be improved to more clearly express the research.

Author Response

Review Report Form

Reviewer-3

Open Review

(x) I would not like to sign my review report
( ) I would like to sign my review report

Quality of English Language

(x) The English could be improved to more clearly express the research.
( ) The English is fine and does not require any improvement.

 

	Yes	Can be improved	Must be improved	Not applicable
Does the introduction provide sufficient background and include all relevant references?	( )	(x)	( )	( )
Is the research design appropriate?	(x)	( )	( )	( )
Are the methods adequately described?	(x)	( )	( )	( )
Are the results clearly presented?	(x)	( )	( )	( )
Are the conclusions supported by the results?	( )	(x)	( )	( )
Are all figures and tables clear and well-presented?	(x)	( )	( )	( )

Comments and Suggestions for Authors

Line 57, Could the authors clarify what specific challenges they are referring to, and how molecular dynamics simulations help address them?

Thanks. Yes, this study employed molecular dynamics (MD) simulations to investigate the binding of RGD peptides to the αvβ3 integrin receptor, a critical target for drug delivery in cancer. We have simulated the experimentally defined X-ray structures on the vector-receptor RGD-peptide-integrin systems targeting integrin-expressing tumor cells. The RGD peptide was selected due to its compact size, specificity, and the αvβ3 integrin's significant involvement in tumor angiogenesis, proliferation, and metastasis. These computational efforts aim to model the transport of boron atoms via RGD peptides for targeted boron neutron capture therapy (BNCT), a novel strategy for anti-cancer therapeutic delivery. Peptide-binding receptors are vital in cancer therapy, offering specific and selective targets for drug delivery, enabling selective targeting of cancer cells while sparing healthy tissue. .

Line 224, Did the authors perform additional analyses, such as binding energy calculations or hydrogen bond analysis, to support the claim that RGD-2 is tightly associated with RGD-1, beyond the distance distribution data?

Thanks. Additionally to previous analysis data now we have generated several Animated movies and added more descriptions for the revised manuscript. Perhaps as like as in the experimental setup we have modeled both situations: an integrin receptor having no RGD (integrin with an RGD-free), as well as an RGD-1-bound integrin (an integrin associated with RGD as in the original PDB file).

Line 236, For clarity, could the authors provide a few specific examples of ion–peptide interactions discussed here?

In the introductory part several specific examples, similar to the ion-peptides mentioned by respected Referee-3, have been outlined.  For practical applications, two primary strategies are suggested for directing boron-containing nanoparticles to their targets utilizing a peptide with the RGD sequence, known for its specific affinity for integrin αvβ3 (a receptor fre-quently overexpressed in tumors); employing folic acid, which targets the FR-α receptor, also commonly found in abundance on tumor cells. Thus, the goal of these computational MD calculations and experiments is to model the interaction processes of the pharmacological pair "VECTOR-RECEPTOR", Pair RGD - integrin αvβ3; Vector            Receptor RGD (peptide containing amino acid sequence – L-arginine, glycine, L-aspartic acid) and Integrin αvβ3.

 

 

Line 237, How does the use of RGD motif peptides as targeting agents compare with other established delivery systems in BNCT in terms of specificity and effectiveness?

This study employed molecular dynamics (MD) simulations to investigate the binding of RGD peptides to the αvβ3 integrin receptor, a critical target for drug delivery in cancer. Over 100-ns simulations of two RGD peptides revealed spontaneous binding and diffusional processes, mirroring the experimental observation of RGD peptide conjugates specifically targeting integrin-expressing tumor cells. The RGD peptide was selected due to its compact size, specificity, and the αvβ3 integrin's significant involvement in tumor angiogenesis, proliferation, and metastasis. These computational efforts aim to model the transport of boron atoms via RGD peptides for targeted boron neutron capture therapy (BNCT), a novel strategy for anti-cancer therapeutic delivery. Peptide-binding receptors are vital in cancer therapy, offering specific and selective targets for drug delivery, enabling selective targeting of cancer cells while sparing healthy tissue. .

Line 239, This sentence appears incomplete. Please revise it for clarity, even if it is intended to reference a citation.

Thanks. Done.

Line 246-248, The authors may consider revising this sentence to improve clarity and coherence.

Thanks. Done.

Line 248, Can the authors comment on the biological plausibility of simulating two RGD peptides simultaneously positioned inside and outside the αvβ3 integrin receptor, given its known binding specificity?

Thanks. Considering the biological plausibility in simulating two RGD-peptides, perhaps as like as in the experimental setup, we have modeled both situations: an integrin receptor having no RGD (integrin with an RGD-free), as well as an RGD-1-bound integrin (an integrin associated with RGD as in the original PDB file). Herein, we have used the experimental X-ray defined PDB structure 3ZE2 and have performed replicate (multiple) simulations on each MD models. The MD results illustrate the accumulation of RGD-2 peptide in a specific integrin region, knowing that there are probably several more binding sites and more mechanisms for the existence of different binding sites between the vector and the receptor. However, the MD results clearly demonstrate one of the most likely binding mechanisms between RGD-2 and the receptor surface, although even the avß3 integrin receptor with or without RGD-1 were used. In the case of an integrin that does not contain RGD-1, the RGD-2 + boron(s) would definitely find much more room for incorporation into the receptor. Of the two RGD peptides that were modeled, RGD-1 remains embedded in the protein pocket during the entire simulation period of 50-100 nanoseconds, and the Animations and Figures illustrate the specific binding of RGD-2 in a natural way, since the binding residues of avß3 integrin to RGD-1 completely coincide with the initial experimental data, the experimentally X-ray measured PDB structure. Multiple (perhaps dozens) MD calculations of the distance between the vector (RGD-2-peptide + boron atoms) and the receptor (avß3-integrin) clearly demonstrate that the entire fragment of RGD-2+boron(s) is inserted into the receptor naturally. It is worth noting that we aimed at binding the entire RGD-2+boron fragment(s), rather than specifically binding selected residues and the receptor. The mechanism of dynamical interactions and diffusional movements between the RGD-2-peptide carrier and the avß3-integrin receptor is illustrated with an accuracy of millions of time steps and multiple configuration analysis.

Line 264, How do the authors ensure that the changing conformations of RGD-2 during diffusion do not introduce non-physical interactions or simulation artifacts?

Thanks. To address this important comment outlined by the respected Reviewer-3 (regarding on the changing conformations of RGD-2 during diffusion do not introduce non-physical interactions or simulation artifacts?) it’s worth noting that we had just followed the standard MD simulation implementation with well-known software, which details are well reported in the literature and also well based from the point of the reproducibility, analysis, proper evaluation. We have used the standard techniques (AMBER18 with TIP3P water and standard equilibration protocol) and the key parameters of the FF (force fields) used in this study are well reported; those are much similar to other protein, peptides, etc, modelling and virtual experiments. And now we have generated several Animated movies and added more descriptions for the revised manuscript. Perhaps as like as in the experimental setup we have modeled both situations: an integrin receptor having no RGD (integrin with an RGD-free), as well as an RGD-1-bound integrin (an integrin associated with RGD as in the original PDB file).The molecule αvβ3, identified as an integrin and the vitronectin receptor, serves as an indicator of new blood vessel formation in tumors. This integrin is composed of two distinct subunits: integrin alpha V and integrin beta 3. It's important to recognize that RGD peptides possess the capability to attach to either of these subunits individually, or to both at the same time. At present, it remains unclear if RGD peptides exhibit varying degrees of affinity for αvβ3 integrins found in mice compared to those in humans. Therefore, the plan is to quantify the binding interaction between RGD and both murine and human αvβ3 integrins. This principle extends to other potential strategies for targeted delivery of boron to tumor sites.

Line 266, What conditions or constraints were applied to model the “free diffusion” of RGD-2, and how do these reflect realistic biological conditions?

Thanks. In “free-diffusion” modeling the MD results clearly demonstrated one of the most likely binding mechanisms between RGD-2 and the receptor surface, although even the avß3 integrin receptor with or without RGD-1 were used. Of the two RGD peptides that were modeled, RGD-1 remains embedded in the protein pocket during the entire simulation period of 50-100 nanoseconds, and the Animations and Figures illustrate the specific binding of RGD-2 in a natural way, since the binding residues of avß3 integrin to RGD-1 completely coincide with the initial experimental data, the experimentally X-ray measured PDB structure. Multiple (perhaps dozens) MD calculations of the distance between the vector (RGD-2-peptide + boron atoms) and the receptor (avß3-integrin) clearly demonstrate that the entire fragment of RGD-2+boron(s) is inserted into the receptor naturally.

Line 327, Could the authors explain the choice of a 100-nanosecond simulation time? Were multiple simulation runs performed to ensure consistency of the results?

Thanks. Perhaps the most interesting event RGD-2 – integrin (vector-receptor) binding occurs at earlier stage around 45-50-nanoseconds range. it’s worth noting that 50-100-ns time scale has rather established well from the point of the standard MD simulation modeling for similar systems (extended macromolecular objects, enzymes, proteins, etc). We had just extended the simulation time scale till 100-ns just for the more “statistical safety”. Of the two RGD peptides that were modeled, RGD-1 remains embedded in the protein pocket during the entire simulation period of 50-100 nanoseconds, and Figures illustrate the specific binding of RGD-2 in a natural way, since the binding residues of avß3 integrin to RGD-1 completely coincide with the initial experimental data, the experimentally X-ray measured PDB structure. Multiple (perhaps dozens) MD calculations of the distance between the vector (RGD-2-peptide + boron atoms) and the receptor (avß3-integrin) clearly demonstrate that the entire fragment of RGD-2+boron(s) is inserted into the receptor naturally. It is worth noting that we aimed at binding the entire RGD-2+boron fragment(s), rather than specifically binding selected residues and the receptor. The mechanism of dynamical interactions and diffusional movements between the RGD-2-peptide carrier and the avß3-integrin receptor is illustrated with an accuracy of millions of time steps and multiple configuration analysis.

Submission Date

16 March 2026

Date of this review

31 Mar 2026 03:58:31

Reviewer 4 Report

Comments and Suggestions for Authors

Line 57, Could the authors clarify what specific challenges they are referring to, and how molecular dynamics simulations help address them?

Line 224, Did the authors perform additional analyses, such as binding energy calculations or hydrogen bond analysis, to support the claim that RGD-2 is tightly associated with RGD-1, beyond the distance distribution data?

Line 236, For clarity, could the authors provide a few specific examples of ion–peptide interactions discussed here?

Line 237, How does the use of RGD motif peptides as targeting agents compare with other established delivery systems in BNCT in terms of specificity and effectiveness?

Line 239, This sentence appears incomplete. Please revise it for clarity, even if it is intended to reference a citation.

Line 246-248, The authors may consider revising this sentence to improve clarity and coherence.

Line 248, Can the authors comment on the biological plausibility of simulating two RGD peptides simultaneously positioned inside and outside the αvβ3 integrin receptor, given its known binding specificity?

Line 264, How do the authors ensure that the changing conformations of RGD-2 during diffusion do not introduce non-physical interactions or simulation artifacts?

Line 266, What conditions or constraints were applied to model the “free diffusion” of RGD-2, and how do these reflect realistic biological conditions?

Line 327, Could the authors explain the choice of a 100-nanosecond simulation time? Were multiple simulation runs performed to ensure consistency of the results?

Author Response

Review Report Form
Reviewer-1

Open Review

( ) I would not like to sign my review report
(x) I would like to sign my review report

Quality of English Language

( ) The English could be improved to more clearly express the research.
(x) The English is fine and does not require any improvement.

 

	Yes	Can be improved	Must be improved	Not applicable
Does the introduction provide sufficient background and include all relevant references?	(x)	( )	( )	( )
Is the research design appropriate?	( )	(x)	( )	( )
Are the methods adequately described?	( )	(x)	( )	( )
Are the results clearly presented?	( )	(x)	( )	( )
Are the conclusions supported by the results?	( )	(x)	( )	( )
Are all figures and tables clear and well-presented?	(x)	( )	( )	( )

Comments and Suggestions for Authors

This manuscript presents a molecular dynamics (MD) simulation study exploring the interaction between RGD peptides and the αvβ3 integrin receptor, with the broader goal of modeling boron delivery for tumor-targeted applications such as boron neutron capture therapy (BNCT). The topic is conceptually relevant, as RGD-integrin interactions are well-established in targeted drug delivery, and computational approaches can provide atomistic insight into ligand-receptor recognition and transport mechanisms. The authors attempt to simulate both peptide binding and the association of boron atoms in single and clustered forms, and the idea of modeling a “vector-receptor” pharmacological system is potentially interesting. However, while the manuscript includes extensive MD simulations and numerous structural visualizations, the scientific rigor, methodological transparency, and depth of analysis are currently insufficient to support the strength of the conclusions. The study remains largely descriptive, with limited quantitative validation and significant overinterpretation of the computational results. Substantial revisions are required to improve clarity, reproducibility, and scientific impact.

Thanks. We really appreciate all the introductory words and comments outlined above by the Reviewer-1. Below we have answered all the questions from p.1 to p.4. But, first of all, we are very sorry that all the discussions that were done by three respected reviewers were actually conducted on the OLD version of our manuscript. Due to some technical issues and misunderstandings between the authors and the CIMB & Pharmaceutics Editorial Boards, it was not possible to upload the substantially corrected version of our manuscript with high-quality figures and text revisions. Thus, we have now downloaded the revised version of our manuscript thereby incorporating the second revisions that we have done for the Pharmaceutics into the current CIMB-version. We have now prepared several animation movies that make the things clearer for better science impact and improving reproducibility. Also, we have now used graphics software to represent the high-quality figures, instead of the previous snapshots, etc.

1. The overall scientific rationale would benefit from clearer positioning and stronger justification. While the RGD-αvβ3 interaction is a well-known paradigm in targeted delivery, the novelty of simulating boron transport in this context is not convincingly articulated. The manuscript repeatedly frames the work as modeling a “vector-receptor” system, but this terminology is not clearly defined or distinguished from existing ligand-receptor modeling frameworks. In addition, the connection between the MD simulations performed and the practical implementation of BNCT remains speculative. The Introduction would be strengthened by more explicitly identifying the gap in current knowledge and clarifying how this study advances beyond prior computational or experimental work on RGD-integrin targeting.

Thanks. We agree that the RGD-αvβ3 interaction is a well-known paradigm in targeted delivery, nevertheless, regarding on the novelty, there are few simulation works on the boron transport in this context. As for the term “vector-receptor” in modeling, we also used the more common terminology “ligand-receptor” throughout the manuscript. In order to better link the MD modeling and the practical implementation of BNCT now we have modified the text and places in the introductory part adding more descriptions and references regarding on the previous computational and experimental work in targeting RGD-integrins.

 

1. The methodological section, although lengthy, lacks critical details necessary for reproducibility and proper evaluation. While the use of AMBER18, TIP3P water, and standard equilibration protocols is noted, key parameters are either missing or inconsistently described. For example, the force fields used for the protein, peptide, and especially the boron atoms or clusters are not clearly specified, which is a major concern given that boron-containing species require careful parameterization. It is also unclear how boron atoms are represented (as neutral particles, ions, or part of a defined chemical structure), which directly affects the validity of the simulations. Furthermore, the manuscript mentions both implicit and explicit solvent models being used (page 6), but does not clarify when or why each was applied. These ambiguities significantly limit the reproducibility and credibility of the computational setup.

Thanks. Perhaps there are not much novelty regarding on the methodological part, we had just followed the standard MD simulation implementation with well-known software, that details are lot reported in the literature and also well based from the point of the reproducibility, analysis, proper evaluation. We have used the standard techniques (AMBER18 with TIP3P water and standard equilibration protocol) and the key parameters of the FF (force fields) used in this study are well reported; those are much similar to other protein, peptides, etc, modelling and virtual experiments. With regards to the boron atoms or clusters, as major concern outlined by the Referee-1, the boron-containing RGD-vector or ligand simulated with boron atoms that are represented as neutral particles. During multiple (perhaps dozens) MD simulations for RGD-2-peptide + boron atoms, it was necessary to ensure all natural vibrational and diffusion movements, but to preserve the position of boron atoms inside RGD-2-peptide. Thus, we have restrained (preserved) all the relative positions of the boron atoms bound to RGD-2-peptide during the entire dynamical and diffusional processes. So far, the relative arrangement of boron atoms bound to RGD-peptide has not been violated, since the main focus has been on the delivery of RGD-2 + boron(s) to the integrin + RGD-1 receptor. In particular, the force field used for boron atoms was approximated as boron containing ligand (RGD-2+boron), thereby disabling and modifying the prediction of the atom type and bond type, even if the initial FF parameters for the boron atom were not recognized in the running antechamber and related modules. Next, now we mention just explicit solvent model and the word “implicit” has been removed from the methodology part of the manuscript. It’s worth noting that the implicit solvent model maybe good for some systems, but it can give poor results for certain molecular shapes (boundary effects), etc. Now the additional animation movies in this study have been generated to demonstrate the “vector-receptor” (“ligand-receptor”) interactions in the explicit solvent that clarify the reproducibility and credibility of the computational setup.

 

1. The design of the simulation system itself raises conceptual concerns. The model includes two RGD peptides(one pre-bound and one freely diffusing) used to infer binding mechanisms . While this is an interesting setup, the rationale for this specific configuration is not sufficiently justified. It is unclear whether this scenario reflects a biologically realistic environment or an artificial construct. Additionally, the use of a single PDB structure (3ZE2) without discussion of receptor flexibility, conformational states, or potential missing regions limits the generalizability of the findings. The absence of replicate simulations or statistical validation further weakens confidence in the reported observations.

Thank you. Now we have added more descriptions of the purpose in yellow for the revised manuscript, and to answer question 3 correctly, we have now added several animated movies as Supplementary Material. Perhaps we have modeled both situations described by the respected Referee-1: an integrin receptor having no RGD (integrin with an RGD-free), as well as an RGD-1-bound integrin (an integrin associated with RGD as in the original PDB file). Yes, we used PDB structure 3ZE2 but the standard minimization and relaxation procedure are well established to ensure the recovering of the receptor flexibility and conformational states (on the 10-20 ns range, prior to large-scale 50-100 ns equilibrations stages). But the replicate (multiple) simulations were realized on each MD models, it’s wrong or misunderstanding by Reviewer-1 on the absence of replicate simulations. The MD results illustrate the accumulation of RGD-2 peptide in a specific integrin region, knowing that there are probably several more binding sites and more mechanisms for the existence of different binding sites between the vector and the receptor. However, the MD results clearly demonstrate one of the most likely binding mechanisms between RGD-2 and the receptor surface, although even the avß3 integrin receptor with or without RGD-1 were used. In the case of an integrin that does not contain RGD-1, the RGD-2 + boron(s) would definitely find much more room for incorporation into the receptor. Of the two RGD peptides that were modeled, RGD-1 remains embedded in the protein pocket during the entire simulation period of 50-100 nanoseconds, and the Animations and Figures illustrate the specific binding of RGD-2 in a natural way, since the binding residues of avß3 integrin to RGD-1 completely coincide with the initial experimental data, the experimentally X-ray measured PDB structure. Multiple (perhaps dozens) MD calculations of the distance between the vector (RGD-2-peptide + boron atoms) and the receptor (avß3-integrin) clearly demonstrate that the entire fragment of RGD-2+boron(s) is inserted into the receptor naturally. It is worth noting that we aimed at binding the entire RGD-2+boron fragment(s), rather than specifically binding selected residues and the receptor. The mechanism of dynamical interactions and diffusional movements between the RGD-2-peptide carrier and the avß3-integrin receptor is illustrated with an accuracy of millions of time steps and multiple configuration analysis.

1. The Results section is largely descriptive and relies heavily on visual inspection of trajectories and structural snapshots (e.g., Figures 6-15 across pages 7-16). While these figures illustrate possible configurations and interactions, they are not accompanied by sufficient quantitative analysis. For example, although distance plots between peptides are shown (Figure 11, page 12), there is no comprehensive analysis of binding stability, interaction energies, hydrogen bonding patterns, or contact frequencies. Standard MD metrics such as RMSD, RMSF, radius of gyration, or binding free energy calculations (e.g., MM/PBSA) are notably absent. Without these quantitative measures, it is difficult to assess whether the observed interactions are stable, significant, or reproducible.

Thanks for the comment. Now we have replaced Figures 5, 6a-6b-6c-6d, 7, 8a-8b-8c-8d, 9a-9b-9c-9d, 10, thereby essentially improved the figures for the best visualization, using the VMD (Visual Molecular Dynamics) and other graphical software. We have tried not to overload the content of the manuscript (say, mentioned by a respected Reviewer-1 in relation to the standard RMSD RMSF or other calculations), although the most likely interaction and dynamical effects have now been demonstrated through additional new Animated movies and Figures. Thus, during multiple (perhaps dozens) MD simulations for RGD-2-peptide + boron atoms, it was necessary to ensure all natural vibrational and diffusion movements, but to preserve the position of boron atoms inside RGD-2-peptide. Thus, we have preserved all the relative positions of the boron atoms bound to RGD-2-peptide during the entire dynamical and diffusional processes. So far, the relative arrangement of boron atoms bound to RGD-peptide has not been violated, since the main focus has been on the delivery of RGD-2 + boron(s) to the integrin + RGD-1 receptor. In particular, the force field used for boron atoms was approximated as neutral particles inside boron containing a ligand (RGD-2+boron), thereby disabling and modifying the prediction of the atom type and bond type, even if the initial FF parameters for the boron atom were not recognized in the running antechamber and related modules. 

 

Submission Date

16 March 2026

Date of this review

28 Mar 2026 21:36:35

 

 

Review Report Form

Reviewer-2

Open Review

( ) I would not like to sign my review report
(x) I would like to sign my review report

Quality of English Language

(x) The English could be improved to more clearly express the research.
( ) The English is fine and does not require any improvement.

 

	Yes	Can be improved	Must be improved	Not applicable
Does the introduction provide sufficient background and include all relevant references?	( )	(x)	( )	( )
Is the research design appropriate?	( )	(x)	( )	( )
Are the methods adequately described?	( )	(x)	( )	( )
Are the results clearly presented?	( )	(x)	( )	( )
Are the conclusions supported by the results?	( )	( )	(x)	( )
Are all figures and tables clear and well-presented?	( )	(x)	( )	( )

Comments and Suggestions for Authors

The article is devoted to molecular dynamics (MD) simulations to explore the interaction of the RGD peptide with the avb3 integrin receptor for targeted drug delivery to tumors.

Thanks. We really appreciate all comments outlined above by the Reviewer-2. Below we have answered all the questions, but, first of all, we are very sorry that all the discussions that were done by three respected reviewers were actually conducted on the OLD version of our manuscript. Due to some technical issues and misunderstandings between the authors and the CIMB & Pharmaceutics Editorial Boards, it was not possible to upload the substantially corrected version of our manuscript with high-quality figures and text revisions. Thus, we have now downloaded the revised version of our manuscript thereby incorporating the second revisions that we have done for the Pharmaceutics into the current CIMB-version. We have now prepared several animation movies that make the things clearer for better science impact and improving reproducibility. Also, we have now used graphics software to represent the high-quality figures, instead of the previous snapshots, etc.

Some remarks:

The role of boron is not disclosed in the introduction.

Thanks. Now we have improved the introduction and the entire text.

The introduction does not sufficiently disclose this RGD peptide vector system for specific delivery to the avb3 integrin receptor and describes the results and examples of targeted drug delivery through this receptor.

Thank you. We have added more descriptions for the revised manuscript; thereby we have now added several more Animated movies and high-quality Figures as Supplementary Materials. Perhaps we have modeled both situations: an integrin receptor having no RGD (integrin with an RGD-free), as well as an RGD-1-bound integrin (an integrin associated with RGD as in the original PDB file). The MD results illustrate the accumulation of RGD-2 peptide in a specific integrin region, knowing that there are probably several more binding sites and more mechanisms for the existence of different binding sites between the vector and the receptor. However, the MD results clearly demonstrate one of the most likely binding mechanisms between RGD-2 and the receptor surface, although even the avß3 integrin receptor with or without RGD-1 were used.

The results and examples of targeted drug delivery through this receptor and  the mechanism of howRGD-peptides bind to avb3 integrin receptor has already been described in the literature references [1-6]. There is no emphasis on what the authors have introduced in comparison with previous works.

Thanks. We agree that the RGD-αvβ3 interaction is a well-known paradigm in targeted delivery, nevertheless, regarding on the novelty, there are few simulation works on the boron transport in this context. In order to better link the MD modeling and the practical implementation now we have modified the text and places in the introductory and discussion parts adding more descriptions and references regarding on the previous computational and experimental work in literature.

Discussion: There is not enough overview here are not enough generalizing discussions , comparisons with literature results, and it is not entirely clear what was found out, there is no clear result and no clear conclusion of this study

It would be nice to have some experimental data to confirm or link to literary works and compare them with literary works. The links are largely to Russian works, it is advisable to add others.

Thank you. The links to Russian works has been reduced now to one cited paper and adding others. And now we have added more descriptions for the revised manuscript, and generated several Animated movies as Supplementary Material. Perhaps as like as in the experimental setup we have modeled both situations: an integrin receptor having no RGD (integrin with an RGD-free), as well as an RGD-1-bound integrin (an integrin associated with RGD as in the original PDB file). Yes, we used the experimental X-ray defined PDB structure 3ZE2 with the standard minimization and relaxation procedure are well established the receptor flexibility and conformational states (on the 10-20 ns range, prior to large-scale 50-100 ns equilibrations stages). We have performed replicate (multiple) simulations on each MD models. The MD results illustrate the accumulation of RGD-2 peptide in a specific integrin region, knowing that there are probably several more binding sites and more mechanisms for the existence of different binding sites between the vector and the receptor. However, the MD results clearly demonstrate one of the most likely binding mechanisms between RGD-2 and the receptor surface, although even the avß3 integrin receptor with or without RGD-1 were used. In the case of an integrin that does not contain RGD-1, the RGD-2 + boron(s) would definitely find much more room for incorporation into the receptor. Of the two RGD peptides that were modeled, RGD-1 remains embedded in the protein pocket during the entire simulation period of 50-100 nanoseconds, and the Animations and Figures illustrate the specific binding of RGD-2 in a natural way, since the binding residues of avß3 integrin to RGD-1 completely coincide with the initial experimental data, the experimentally X-ray measured PDB structure. Multiple (perhaps dozens) MD calculations of the distance between the vector (RGD-2-peptide + boron atoms) and the receptor (avß3-integrin) clearly demonstrate that the entire fragment of RGD-2+boron(s) is inserted into the receptor naturally. It is worth noting that we aimed at binding the entire RGD-2+boron fragment(s), rather than specifically binding selected residues and the receptor. The mechanism of dynamical interactions and diffusional movements between the RGD-2-peptide carrier and the avß3-integrin receptor is illustrated with an accuracy of millions of time steps and multiple configuration analysis.

 

Comments on the Quality of English Language

The English could be improved to more clearly express the research.

Submission Date

16 March 2026

Date of this review

30 Mar 2026 19:31:19

 

 

Review Report Form

Reviewer-3

Open Review

(x) I would not like to sign my review report
( ) I would like to sign my review report

Quality of English Language

(x) The English could be improved to more clearly express the research.
( ) The English is fine and does not require any improvement.

 

	Yes	Can be improved	Must be improved	Not applicable
Does the introduction provide sufficient background and include all relevant references?	( )	(x)	( )	( )
Is the research design appropriate?	(x)	( )	( )	( )
Are the methods adequately described?	(x)	( )	( )	( )
Are the results clearly presented?	(x)	( )	( )	( )
Are the conclusions supported by the results?	( )	(x)	( )	( )
Are all figures and tables clear and well-presented?	(x)	( )	( )	( )

Comments and Suggestions for Authors

Line 57, Could the authors clarify what specific challenges they are referring to, and how molecular dynamics simulations help address them?

Thanks. Yes, this study employed molecular dynamics (MD) simulations to investigate the binding of RGD peptides to the αvβ3 integrin receptor, a critical target for drug delivery in cancer. We have simulated the experimentally defined X-ray structures on the vector-receptor RGD-peptide-integrin systems targeting integrin-expressing tumor cells. The RGD peptide was selected due to its compact size, specificity, and the αvβ3 integrin's significant involvement in tumor angiogenesis, proliferation, and metastasis. These computational efforts aim to model the transport of boron atoms via RGD peptides for targeted boron neutron capture therapy (BNCT), a novel strategy for anti-cancer therapeutic delivery. Peptide-binding receptors are vital in cancer therapy, offering specific and selective targets for drug delivery, enabling selective targeting of cancer cells while sparing healthy tissue. .

Line 224, Did the authors perform additional analyses, such as binding energy calculations or hydrogen bond analysis, to support the claim that RGD-2 is tightly associated with RGD-1, beyond the distance distribution data?

Thanks. Additionally to previous analysis data now we have generated several Animated movies and added more descriptions for the revised manuscript. Perhaps as like as in the experimental setup we have modeled both situations: an integrin receptor having no RGD (integrin with an RGD-free), as well as an RGD-1-bound integrin (an integrin associated with RGD as in the original PDB file).

Line 236, For clarity, could the authors provide a few specific examples of ion–peptide interactions discussed here?

In the introductory part several specific examples, similar to the ion-peptides mentioned by respected Referee-3, have been outlined.  For practical applications, two primary strategies are suggested for directing boron-containing nanoparticles to their targets utilizing a peptide with the RGD sequence, known for its specific affinity for integrin αvβ3 (a receptor fre-quently overexpressed in tumors); employing folic acid, which targets the FR-α receptor, also commonly found in abundance on tumor cells. Thus, the goal of these computational MD calculations and experiments is to model the interaction processes of the pharmacological pair "VECTOR-RECEPTOR", Pair RGD - integrin αvβ3; Vector            Receptor RGD (peptide containing amino acid sequence – L-arginine, glycine, L-aspartic acid) and Integrin αvβ3.

 

 

Line 237, How does the use of RGD motif peptides as targeting agents compare with other established delivery systems in BNCT in terms of specificity and effectiveness?

This study employed molecular dynamics (MD) simulations to investigate the binding of RGD peptides to the αvβ3 integrin receptor, a critical target for drug delivery in cancer. Over 100-ns simulations of two RGD peptides revealed spontaneous binding and diffusional processes, mirroring the experimental observation of RGD peptide conjugates specifically targeting integrin-expressing tumor cells. The RGD peptide was selected due to its compact size, specificity, and the αvβ3 integrin's significant involvement in tumor angiogenesis, proliferation, and metastasis. These computational efforts aim to model the transport of boron atoms via RGD peptides for targeted boron neutron capture therapy (BNCT), a novel strategy for anti-cancer therapeutic delivery. Peptide-binding receptors are vital in cancer therapy, offering specific and selective targets for drug delivery, enabling selective targeting of cancer cells while sparing healthy tissue. .

Line 239, This sentence appears incomplete. Please revise it for clarity, even if it is intended to reference a citation.

Thanks. Done.

Line 246-248, The authors may consider revising this sentence to improve clarity and coherence.

Thanks. Done.

Line 248, Can the authors comment on the biological plausibility of simulating two RGD peptides simultaneously positioned inside and outside the αvβ3 integrin receptor, given its known binding specificity?

Thanks. Considering the biological plausibility in simulating two RGD-peptides, perhaps as like as in the experimental setup, we have modeled both situations: an integrin receptor having no RGD (integrin with an RGD-free), as well as an RGD-1-bound integrin (an integrin associated with RGD as in the original PDB file). Herein, we have used the experimental X-ray defined PDB structure 3ZE2 and have performed replicate (multiple) simulations on each MD models. The MD results illustrate the accumulation of RGD-2 peptide in a specific integrin region, knowing that there are probably several more binding sites and more mechanisms for the existence of different binding sites between the vector and the receptor. However, the MD results clearly demonstrate one of the most likely binding mechanisms between RGD-2 and the receptor surface, although even the avß3 integrin receptor with or without RGD-1 were used. In the case of an integrin that does not contain RGD-1, the RGD-2 + boron(s) would definitely find much more room for incorporation into the receptor. Of the two RGD peptides that were modeled, RGD-1 remains embedded in the protein pocket during the entire simulation period of 50-100 nanoseconds, and the Animations and Figures illustrate the specific binding of RGD-2 in a natural way, since the binding residues of avß3 integrin to RGD-1 completely coincide with the initial experimental data, the experimentally X-ray measured PDB structure. Multiple (perhaps dozens) MD calculations of the distance between the vector (RGD-2-peptide + boron atoms) and the receptor (avß3-integrin) clearly demonstrate that the entire fragment of RGD-2+boron(s) is inserted into the receptor naturally. It is worth noting that we aimed at binding the entire RGD-2+boron fragment(s), rather than specifically binding selected residues and the receptor. The mechanism of dynamical interactions and diffusional movements between the RGD-2-peptide carrier and the avß3-integrin receptor is illustrated with an accuracy of millions of time steps and multiple configuration analysis.

Line 264, How do the authors ensure that the changing conformations of RGD-2 during diffusion do not introduce non-physical interactions or simulation artifacts?

Thanks. To address this important comment outlined by the respected Reviewer-3 (regarding on the changing conformations of RGD-2 during diffusion do not introduce non-physical interactions or simulation artifacts?) it’s worth noting that we had just followed the standard MD simulation implementation with well-known software, which details are well reported in the literature and also well based from the point of the reproducibility, analysis, proper evaluation. We have used the standard techniques (AMBER18 with TIP3P water and standard equilibration protocol) and the key parameters of the FF (force fields) used in this study are well reported; those are much similar to other protein, peptides, etc, modelling and virtual experiments. And now we have generated several Animated movies and added more descriptions for the revised manuscript. Perhaps as like as in the experimental setup we have modeled both situations: an integrin receptor having no RGD (integrin with an RGD-free), as well as an RGD-1-bound integrin (an integrin associated with RGD as in the original PDB file).The molecule αvβ3, identified as an integrin and the vitronectin receptor, serves as an indicator of new blood vessel formation in tumors. This integrin is composed of two distinct subunits: integrin alpha V and integrin beta 3. It's important to recognize that RGD peptides possess the capability to attach to either of these subunits individually, or to both at the same time. At present, it remains unclear if RGD peptides exhibit varying degrees of affinity for αvβ3 integrins found in mice compared to those in humans. Therefore, the plan is to quantify the binding interaction between RGD and both murine and human αvβ3 integrins. This principle extends to other potential strategies for targeted delivery of boron to tumor sites.

Line 266, What conditions or constraints were applied to model the “free diffusion” of RGD-2, and how do these reflect realistic biological conditions?

Thanks. In “free-diffusion” modeling the MD results clearly demonstrated one of the most likely binding mechanisms between RGD-2 and the receptor surface, although even the avß3 integrin receptor with or without RGD-1 were used. Of the two RGD peptides that were modeled, RGD-1 remains embedded in the protein pocket during the entire simulation period of 50-100 nanoseconds, and the Animations and Figures illustrate the specific binding of RGD-2 in a natural way, since the binding residues of avß3 integrin to RGD-1 completely coincide with the initial experimental data, the experimentally X-ray measured PDB structure. Multiple (perhaps dozens) MD calculations of the distance between the vector (RGD-2-peptide + boron atoms) and the receptor (avß3-integrin) clearly demonstrate that the entire fragment of RGD-2+boron(s) is inserted into the receptor naturally.

Line 327, Could the authors explain the choice of a 100-nanosecond simulation time? Were multiple simulation runs performed to ensure consistency of the results?

Thanks. Perhaps the most interesting event RGD-2 – integrin (vector-receptor) binding occurs at earlier stage around 45-50-nanoseconds range. it’s worth noting that 50-100-ns time scale has rather established well from the point of the standard MD simulation modeling for similar systems (extended macromolecular objects, enzymes, proteins, etc). We had just extended the simulation time scale till 100-ns just for the more “statistical safety”. Of the two RGD peptides that were modeled, RGD-1 remains embedded in the protein pocket during the entire simulation period of 50-100 nanoseconds, and Figures illustrate the specific binding of RGD-2 in a natural way, since the binding residues of avß3 integrin to RGD-1 completely coincide with the initial experimental data, the experimentally X-ray measured PDB structure. Multiple (perhaps dozens) MD calculations of the distance between the vector (RGD-2-peptide + boron atoms) and the receptor (avß3-integrin) clearly demonstrate that the entire fragment of RGD-2+boron(s) is inserted into the receptor naturally. It is worth noting that we aimed at binding the entire RGD-2+boron fragment(s), rather than specifically binding selected residues and the receptor. The mechanism of dynamical interactions and diffusional movements between the RGD-2-peptide carrier and the avß3-integrin receptor is illustrated with an accuracy of millions of time steps and multiple configuration analysis.

Submission Date

16 March 2026

Date of this review

31 Mar 2026 03:58:31

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The image resolution is improved. But I still cannot find the time step for each run.

I have no more questions for the authors.

Author Response

Comments and Suggestions for Authors

The image resolution is improved. But I still cannot find the time step for each run.

 

THANKS. We highly appreciate all Reviewer-2' valuable comments, indeed. Now we have cleared the time step for each run in the text.

I have no more questions for the authors.

THANKS AGAIN!    Submission Date 16 March 2026 Date of this review 02 Apr 2026 10:40:09

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors I did not see the answers to the reviewer's comments, nor the answers to the questions: no changes were found in the text - (only the response for the  Reviewer-3 is presented). The Comments and Suggestions for Authors (report 1)

The article is devoted to molecular dynamics (MD) simulations to explore the interaction of the RGD peptide with the avb3 integrin receptor for targeted drug delivery to tumors.

Some remarks:

1) The role of boron is not disclosed in the introduction (only in Conclusion section there is some info) 

2)The introduction does not sufficiently disclose the literature data for  RGD peptide vector system for specific delivery to the avb3 integrin receptor and not sufficiently description of the results and examples of targeted drug delivery through this receptor.

3)The results and examples of targeted drug delivery through this receptor and  the mechanism of how   RGD-peptides bind to avb3 integrin receptor has already been described in the literature references [1-6]. 

There is no emphasis on what the authors have introduced in comparison with previous works.

4) Discussion: There is not enough overview, not enough generalizing discussions , no enough comparisons with literature results, and it is not entirely clear what was found out, there is no clear result and no clear conclusion of this study

5)Conclusion: "the computer designed RGD-2-peptide + boron(s) and receptor + 
RGD-1-peptide that MD modeled above are fully corresponding to their structural be-
havior in vivo". - What are the arguments for this statement?

6)It would be nice to have some experimental data to confirm or link to literary works and compare them with literary works.

7) The references are mainly to Russian works, it is advisable to add others for to ensure proper balance.

Comments on the Quality of English Language

The English could be improved to more clearly express the research.

Author Response

Review Report Form
Reviewer-3 (Round 2)

Open Review

(x) I would not like to sign my review report
( ) I would like to sign my review report

Quality of English Language

(x) The English could be improved to more clearly express the research.
( ) The English is fine and does not require any improvement.

 

	Yes	Can be improved	Must be improved	Not applicable
Does the introduction provide sufficient background and include all relevant references?	( )	(x)	( )	( )
Is the research design appropriate?	( )	(x)	( )	( )
Are the methods adequately described?	( )	(x)	( )	( )
Are the results clearly presented?	( )	(x)	( )	( )
Are the conclusions supported by the results?	( )	( )	(x)	( )
Are all figures and tables clear and well-presented?	( )	(x)	( )	( )

Comments and Suggestions for Authors

I did not see the answers to the reviewer's comments, nor the answers to the questions: no changes were found in the text - (only the response for the  Reviewer-3 is presented). The Comments and Suggestions for Authors (report 1)

THANKS! Yes, we agree that, we have previously just responded the Reviewer-3’ questions and answered comments. But now we added those answers as changes in the text. 

The article is devoted to molecular dynamics (MD) simulations to explore the interaction of the RGD peptide with the avb3 integrin receptor for targeted drug delivery to tumors.

Some remarks:

1) The role of boron is not disclosed in the introduction (only in Conclusion section there is some info) 

Thanks for the comment. Now the following changes could be found in the text:

“It’s worth noting that the role of boron in BNCT is characterized through the selective destruction of malignant tumor cells by accumulating a stable boron-10 isotope in them and subsequent irradiation with epithermal neutrons. The BNCT (boron-neutron capture therapy) is a promising method of treating malignant tumors and this method involves the targeted destruction of only cancer cells without surgical intervention. In BNCT method, the key importance is not so much the "take-off length" of the neutrons themselves as the depth of their penetration into tissues until the moment of boron-10 capture. Neutrons used in BNCT which is based on the epithermal ones (neutrons with energies from 0.5 eV to 10-20 keV) and penetration depth is about 8-10 cm (The epithermal neutrons, slowing down to thermal energies, are able to penetrate body tissues to a depth of 8-10 cm which makes it possible to treat deep-lying tumors; for example, glioblastomas of the brain, while thermal neutrons are absorbed by the skin). But regarding the mileage of secondary particles, after neutron capture by the boron-10 nucleus, a nuclear reaction occurs, and the resulting alpha particle and lithium-7 nucleus have a low free path of only 5-9 micrometers (approximately the diameter of one cell). This ensures the selective destruction of only the cancer cell containing boron, so thus, neutrons travel centimeters in tissues, but the destructive reaction occurs at the micrometer level”.

2)The introduction does not sufficiently disclose the literature data for  RGD peptide vector system for specific delivery to the avb3 integrin receptor and not sufficiently description of the results and examples of targeted drug delivery through this receptor.

Thanks for the comment. Now we have added more citations in the Introduction and conclusion parts:

 “The literature data for the RGD-peptide vector system in specific delivery to the avb3 integrin receptor and description of the results and examples of targeted drug delivery through this receptor are well as the molecular basis for the targeted binding of RGD-containing peptide to integrin αvβ3 and novel linear peptides with high affinity to αvβ3 integrin in precise tumor identification could be also found in Refs. [2, 7, 9-11, 15, 19-28]”.

3)The results and examples of targeted drug delivery through this receptor and  the mechanism of how   RGD-peptides bind to avb3 integrin receptor has already been described in the literature references [1-6]. 

There is no emphasis on what the authors have introduced in comparison with previous works.

Thanks. We agree that the RGD-peptide - αvβ3 receptor interaction is a well-known paradigm in targeted delivery, nevertheless, regarding on the novelty, there are few simulation works on the mechanism of the RGD-peptides binding to avb3 integrin receptor as well as a boron transport in this context. In order to better link the MD modeling and the practical implementation now we have modified the text and places in the introductory and discussion parts adding more descriptions and references regarding on the previous computational and experimental work in literature.

4) Discussion: There is not enough overview, not enough generalizing discussions , no enough comparisons with literature results, and it is not entirely clear what was found out, there is no clear result and no clear conclusion of this study

Thank you. We mostly enlighten in this simulation work the atomic / molecular scale events on the RGD-peptide - αvβ3 integrin receptor binding mechanism. But there are currently not a much experimental data regarding the above mechanism as well as there are few simulation works on the mechanism of the RGD-peptides binding to avb3 integrin receptor in this context. In order to better link the MD modeling and the practical implementation now we have modified the text and places in the introductory and discussion parts adding more descriptions and references regarding on the previous computational and experimental work in literature.

5)Conclusion: "the computer designed RGD-2-peptide + boron(s) and receptor + 
RGD-1-peptide that MD modeled above are fully corresponding to their structural be-
havior in vivo". - What are the arguments for this statement?

The above sentence means that all the initial structural data for RGD-2-peptide, αvβ3 integrin receptor, RGD-1-peptide, etc. were completely obtained experimentally using X-rays and those are available from PDB (Protein Data Base), that files fully correspond to the structural forms in vivo. We have just modeled the phase transition from crystal-to-liquid, and the structural behavior of the modeled multiple RGD-2-peptide, receptor, and RGD-1-peptide systems in an aqueous medium now corresponds to their more adequate physiological liquid state.

6)It would be nice to have some experimental data to confirm or link to literary works and compare them with literary works.

Thank you. Again we mostly are being concentrated in current study on the simulation aspect of the atomic / molecular scale events of the RGD-peptide - αvβ3 integrin receptor binding mechanism. We believe that for today there are not a much experimental data regarding the above mechanism as well as there are few simulation works on the mechanism of the RGD-peptides binding to avb3 integrin receptor in this context. Now we have added more descriptions for the revised manuscript, and generated several Animated movies as Supplementary Material. Perhaps as like as in the experimental setup we have modeled both situations: an integrin receptor having no RGD (integrin with an RGD-free), as well as an RGD-1-bound integrin (an integrin associated with RGD as in the original PDB file). Yes, we used the experimental X-ray defined PDB structure 3ZE2 with the standard minimization and relaxation procedure are well established the receptor flexibility and conformational states (on the 10-20 ns range, prior to large-scale 50-100 ns equilibrations stages). We have performed replicate (multiple) simulations on each MD models. The MD results illustrate the accumulation of RGD-2 peptide in a specific integrin region, knowing that there are probably several more binding sites and more mechanisms for the existence of different binding sites between the vector and the receptor. However, the MD results clearly demonstrate one of the most likely binding mechanisms between RGD-2 and the receptor surface, although even the avß3 integrin receptor with or without RGD-1 were used. In the case of an integrin that does not contain RGD-1, the RGD-2 + boron(s) would definitely find much more room for incorporation into the receptor. Of the two RGD peptides that were modeled, RGD-1 remains embedded in the protein pocket during the entire simulation period of 50-100 nanoseconds, and the Animations and Figures illustrate the specific binding of RGD-2 in a natural way, since the binding residues of avß3 integrin to RGD-1 completely coincide with the initial experimental data, the experimentally X-ray measured PDB structure. Multiple (perhaps dozens) MD calculations of the distance between the vector (RGD-2-peptide + boron atoms) and the receptor (avß3-integrin) clearly demonstrate that the entire fragment of RGD-2+boron(s) is inserted into the receptor naturally. It is worth noting that we aimed at binding the entire RGD-2+boron fragment(s), rather than specifically binding selected residues and the receptor. The mechanism of dynamical interactions and diffusional movements between the RGD-2-peptide carrier and the avß3-integrin receptor is illustrated with an accuracy of millions of time steps and multiple configuration analysis.

7) The references are mainly to Russian works, it is advisable to add others for to ensure proper balance.

The links to Russian works has been reduced now to one cited paper and adding others.

 

Comments on the Quality of English Language

The English could be improved to more clearly express the research.

Submission Date

16 March 2026

Date of this review

04 Apr 2026 20:25:20

 

Author Response File: Author Response.pdf

Round 3

Reviewer 3 Report

Comments and Suggestions for Authors

The authors responded to the comments and reinforced the text and explanations.

 

Comments on the Quality of English Language

The English is fine and does not require any improvement.