Structure-Based Virtual Screening and Biological Evaluation of Peptide Inhibitors for Polo-Box Domain
Round 1
Reviewer 1 Report
The author applied a virtual screening and docking to find PLK1-PBD peptide inhibitors, and apply in vitro and in vivo test for a few selected peptides.
The procedure and result are well explained.
Minors.
Figure 2, "5 peptides" should be update as "9 peptides".
Author Response
Response to Reviewer 1 Comments
Point 1: English language and style
( ) Extensive editing of English language and style required
( ) Moderate English changes required
(x) English language and style are fine/minor spell check required
( ) I don't feel qualified to judge about the English language and style
Response 1: Thank you for the suggestion. We have carefully corrected some grammatical errors in the revised manuscript.
Point 2: Minors.
Figure 2, "5 peptides" should be update as "9 peptides".
Response 2: Thank you for the suggestion. The "5 peptides" has been updated as "9 peptides" in the revised manuscript (please see “Figure 2”).
Author Response File: 
Author Response.docx
Reviewer 2 Report
This re-submitted manuscript aims to utilize pharmacophore modeling and structural searching to develop new PLK1 inhibitors targeting the PBD. While the idea and purpose of the work remain interesting and potentially impactful, as with the original submission, the authors have not addressed any of the important issues associated with Plk1 PBD ligand development. Based on the previous reviews, the authors have incorporated the known issues associated with Plk1 PBD inhibitors into their background information (ie poor cell penetration), but have not provided any additional details of how their reported inhibitors have improved upon known ligands that are very similar in structure. Moreover, the sparse additional details provided regarding the ligand databases searched using their pharmacophore model raise further questions as to the purpose of the work. Namely, the authors have validated their pharmacophore model using known PBD ligands from the literature, and cross-referenced the model against 1985 "inactive" ligands to determine accuracy. However the authors then use this model to search a database of 32,000 peptides that appear to already contain the "HSpTA" core sequence. This core sequence search is highly biased toward PBD ligands very similar to known ligands, and would very likely result in hits with at least moderate affinity and poor cellular permeability. Additional commentary regarding the purpose and goals of this research design are needed.
Additional detailed questions or comments are listed below:
Lines 40-41: Grammar
Line 52: “pharmacophore-based database search was used…”. What was searched to produce hit sequences? Were the new sequences generated de novo?
Line 84: “The 83 results indicated a good mapping of ligands on the pharmacophore model.” The crystal structures used are ligand bound forms. Was the ligand included in the pharmacophore modeling?
Line 94-95: “The quality of generated structure-based model was assessed using the GH score as a metric to search a database that consisted of 1985 inactive and 15 active molecules.” How were “active” vs “inactive” molecules selected? From what database?
Line 109-110: “…the pharmacophore model was firstly used as a 3D query to identify potential peptide inhibitors from database containing ~32,000 peptides.” What database? How were these 32,000 molecules selected? Why not screen a much larger library?
Line 120: Figure 2 shows 5 peptides selected, however the text indicates that 9 hits were chosen.
Line 172: “interference” should be “interfere”
Line 205: “1985 peptide sequences without HSpTA were selected as inactive peptides”. Selected from where? What database?
Line 215-216: “MOE was applied to build a two-dimensional (2D) database containing ~32,000 phosphorylated peptides with the HSpTA motif.” How was this database built? Why was phospho-Thr included as a primary motif of the database knowing that the resulting peptides would demonstrate poor cell permeability?
Line 240: “sores” should be “scores”
Line 257-259: “We further performed the sensitivity by synthesizing two new fluorescent probes, FITC-GPMQTSpTPKNG-OH for PLK2-PBD and FITC-GPLATSpTPKNG-OH for PLK3-PBD, respectively”. These probe structures are not “new” sequences. They were reported in several publications as fluorescent probes for their corresponding PBD isoforms.
Author Response
Response to Reviewer 2 Comments
Point 1: English language and style
( ) Extensive editing of English language and style required
( ) Moderate English changes required
(x) English language and style are fine/minor spell check required
( ) I don't feel qualified to judge about the English language and style
Response 1: Thank you for the suggestion. We have carefully corrected some grammatical errors in the revised manuscript.
Point 2: This re-submitted manuscript aims to utilize pharmacophore modeling and structural searching to develop new PLK1 inhibitors targeting the PBD. While the idea and purpose of the work remain interesting and potentially impactful, as with the original submission, the authors have not addressed any of the important issues associated with Plk1 PBD ligand development. Based on the previous reviews, the authors have incorporated the known issues associated with Plk1 PBD inhibitors into their background information (ie poor cell penetration), but have not provided any additional details of how their reported inhibitors have improved upon known ligands that are very similar in structure.
Response 2: Thank you for the suggestion. It should be noted that the ligand-binding site of PLK1-PBD consist of a hydrophobic pocket and a positively charged binding pocket. The most commonly reported peptide inhibitors including “HSpTA” motif can be only bound to the positively charged binding pocket of PLK1-PBD (Nat. Struct. Mol. Biol.. 2009, 16, 876). By comparison, as shown in Figure 3 and Figure S3, the N-terminal 3,4-dichlorophenylalanine of peptide 5 was bound to the hydrophobic pocket of PLK1-PBD. The MOE docking results of the peptide 5 suggested that there were two major interactions between peptide 5 and the PLK1-PBD active site (Figure 3 and 4): (i) The C-terminal phosphorylated threonine bound to the positively charged binding pocket forms multiple hydrogen-bonding interactions with Lys540 and water molecules that were indispensable for the ligand binding of the PLK1-PBD (Nat. Struct. Mol. Biol.. 2009, 16, 876); (ii) The N-terminal 3,4-dichlorophenylalanine bound to the hydrophobic pocket of PLK1-PBD was engaged in a strong hydrophobic interaction with some amino acids, including Tyr481, Tyr421, Phe482 and Tyr485, indicated that it plays key role in stabilizing peptide 5 in the hydrophobic pocket (Figure S2). The related statements have been added to the revised manuscript (please see “2.3. PLK1-PBD inhibition assay”).
Point 3: Moreover, the sparse additional details provided regarding the ligand databases searched using their pharmacophore model raise further questions as to the purpose of the work. Namely, the authors have validated their pharmacophore model using known PBD ligands from the literature, and cross-referenced the model against 1985 "inactive" ligands to determine accuracy. However the authors then use this model to search a database of 32,000 peptides that appear to already contain the "HSpTA" core sequence. This core sequence search is highly biased toward PBD ligands very similar to known ligands, and would very likely result in hits with at least moderate affinity and poor cellular permeability.
Response 3: Thank you for the suggestion. Previous studies showed that the PBD of PLK1 has a crucial role in proper subcellular localization and mitotic functions of PLK1 by interacting with threonine or serine-phosphorylated peptides with the invariable serine residue at the 1 position (S-p-S/T motif) (Nat. Struct. Mol. Biol.. 2009, 16, 876; Science 2003, 299, 1228–1231; J. Biol. Chem. 2002, 277, 32282–32293). At present, a large number of studies have proved that a phosphopeptide sequence including the HSpTA motif is required for targeting PLK1-PBD (Nat. Struct. Mol. Biol. 2009, 16, 876; Angew. Chem. 2012, 124, 7800-7803; Bioorg. Med. Chem. 2018, 26, 3429-3437; Angew. Chem. 2012, 51, 10078-10081; Angew. Chem. 2011, 50, 4003-4006; Adv. Funct. Mater. 2019, 1904969), thus these reported peptide inhibitors have high affinity and highly selective ligands of the PLK1-PBD. However, one of the major limitations of using these phosphopeptides as an anticancer agents is membrane impermeability (Adv. Funct. Mater. 2019, 1904969: DOI: 10.1002/adfm.201904969; Biomaterials, 2012, 33, 6915-6925). Consequently, the anticancer effect has been achieved by directly microinjecting the peptide into HeLa cells (Nat. Struct Mo.l Biol. 2009, 16, 876), but if a new inhibitor based on a “HSpTA” motif can be developed that is capable of penetrating the membranes of cancer cells, it would have good potential as an antitumor agent. In addition, previous reviews also pointed out that “However, very few have advanced as therapeutics due to the requirement of the anionic pSer/pThr residue and its resulting poor cellular penetration. This needs to be address in the paper”, which is consistent with previous studies (Adv. Funct. Mater. 2019, 1904969: DOI: 10.1002/adfm.201904969; Biomaterials, 2012, 33, 6915-6925). Therefore, there is a pressing need to develop cell-permeable PLK1-PBD inhibitors that effectively interfere with the function of PLK1-PBD during cell signaling. In previous work, we found that the incorporation of hydrophobic 3,4-dichlorophenylalanine into a peptide may improve its cellular penetration (Bioorg. Med. Chem. 2018, 26, 3429-3437). In this study, we utilized the computational tools to design the PLK1-PBD inhibitors. Among the 9 selected peptides with submicromolar activities, the peptide 5, as the most potent inhibitor (IC50 = 0.07 μM), showed an approximately 100-fold increase in inhibitory activity than the control poloboxtide. The peptide 5 was further tested for its ability to perturb PLK1-PBD function during cell signaling. The results indicated that peptide 5 including 3,4-dichlorophenylalanine effectively inhibits the cell cycle regulatory proteins inducing mitotic arrest at the G2/M phase cell cycle, suggesting that it has good cellular permeability. The related statements have been added to the revised manuscript.
Point 4: Additional commentary regarding the purpose and goals of this research design are needed. Additional detailed questions or comments are listed below:
Lines 40-41: Grammar
Response 4: Thank you for the suggestion. The grammatical error has carefully been corrected in the revised manuscript (please see “1. Introduction”).
Point 5: Line 52: “pharmacophore-based database search was used…”. What was searched to produce hit sequences? Were the new sequences generated de novo?
Response 5: Thank you for the suggestion. The related statement has been corrected in the revised manuscript (please see “1. Introduction”).
Point 6: Line 84: “The 83 results indicated a good mapping of ligands on the pharmacophore model.” The crystal structures used are ligand bound forms. Was the ligand included in the pharmacophore modeling?
Response 6: Thank you for the suggestion. As shown in Figure 1, the bound ligands have been included in the pharmacophore modeling (please see “Figure 1”).
Point 7: Line 94-95: “The quality of generated structure-based model was assessed using the GH score as a metric to search a database that consisted of 1985 inactive and 15 active molecules.” How were “active” vs “inactive” molecules selected? From what database?
Response 7: Thank you for the suggestion. The 15 active peptides are collected from the reported literatures (Bioorg. Med. Chem. 2018, 26, 3429-3437; Angew. Chem. 2012, 124, 7800-7803) (Table S3). Previous studies have found that a phosphopeptide motif (HSpTA) has a pivotal role in targeting PLK1-PBD (Nat. Struct Mo.l Biol. 2009, 16, 876). Since the phosphopeptide sequence excluding a HSpTA motif is unable to target PLK1-PBD (Nat. Struct Mo.l Biol. 2009, 16, 876; Angew. Chem. 2012, 124, 7800-7803), 1985 non-phosphorylated peptides were selected as inactive peptides from the antimicrobial peptide database (http://aps.unmc.edu/AP/). The detailed statements have been added to the revised manuscript (please see “Materials and methods: 3.1. Pharmacophore model generation and validation and Table S3”).
Point 8: Line 109-110: “…the pharmacophore model was firstly used as a 3D query to identify potential peptide inhibitors from database containing ~32,000 peptides.” What database? How were these 32,000 molecules selected? Why not screen a much larger library?
Response 8: Thank you for the suggestion. In previous studies, we have successfully constructed the peptide database (Adv. Funct. Mater. 2019, 1904969: DOI: 10.1002/adfm.201904969). Compared with other reported peptide databases (Nucleic Acids Res. 2015, 44, D1087-D1093), the peptide database constructed by us is one of the most large libraries. The reference about the peptide database has been added to the revised manuscript (please see “Materials and methods: 3.2. Virtual screening”).
Point 9: Line 120: Figure 2 shows 5 peptides selected, however the text indicates that 9 hits were chosen.
Response 9: Thank you for the suggestion. The "5 peptides" has been updated as "9 peptides" in the revised manuscript (please see “Figure 2”).
Point 10: Line 172: “interference” should be “interfere”
Response 10: Thank you for the suggestion. The word “interference” has been corrected as “interfere” in the revised manuscript (please see “Results and discussion: 2.4. Effect of peptide 5 on cycle Arrest in HeLa Cells”).
Point 11: Line 205: “1985 peptide sequences without HSpTA were selected as inactive peptides”. Selected from where? What database?
Response 11: Thank you for the suggestion. Previous studies have found that a phosphopeptide motif (HSpTA) has a pivotal role in targeting PLK1-PBD (Nat. Struct Mo.l Biol. 2009, 16, 876). Since the phosphopeptide sequence excluding a HSpTA motif is unable to target PLK1-PBD (Nat. Struct Mo.l Biol. 2009, 16, 876), 1985 peptide sequences without the HSpTA were selected as inactive peptides from the antimicrobial peptide database (http://aps.unmc.edu/AP/). The detailed statement has been added to the revised manuscript (please see “Materials and methods: 3.1. Pharmacophore model generation and validation and Table S3”).
Point 12: Line 215-216: “MOE was applied to build a two-dimensional (2D) database containing ~32,000 phosphorylated peptides with the HSpTA motif.” How was this database built? Why was phospho-Thr included as a primary motif of the database knowing that the resulting peptides would demonstrate poor cell permeability?
Response 12: Thank you for the suggestion. In previous studies, we have successfully constructed the peptide database (Adv. Funct. Mater. 2019, 1904969: DOI: 10.1002/adfm.201904969). The reference 22 has been added to the revised manuscript (please see “Materials and methods: 3.2. Virtual screening”). Lots of studies have confirmed that the PBD of PLK1 has a crucial role in proper subcellular localization and mitotic functions of PLK1 by interacting with threonine or serine-phosphorylated (phospho-Thr or -Ser) peptides with the invariable serine residue at the 1 position (S-p-S/T motif) (Nat. Struct. Mol. Biol.. 2009, 16, 876; Science 2003, 299, 1228–1231; J. Biol. Chem. 2002, 277, 32282–32293). An intact phosphoepitope binding module is required for the interaction between the PBD and its binding targets (Nat. Struct. Mol. Biol.. 2009, 16, 876). At present, a large number of studies have proved that a phosphopeptide sequence including the HSpTA motif has a pivotal role in targeting PLK1-PBD (Nat. Struct. Mol. Biol. 2009, 16, 876; Angew. Chem. 2012, 124, 7800-7803; Bioorg. Med. Chem. 2018, 26, 3429-3437; Angew. Chem. 2012, 51, 10078-10081; Angew. Chem. 2011, 50, 4003-4006; Adv. Funct. Mater. 2019, 1904969). The related statement has been added to the revised manuscript (please see “Materials and methods: 3.1. Pharmacophore model generation and validation and Table S3”).
Point 13: Line 240: “sores” should be “scores”
Response 13: Thank you for the suggestion. The word “sores” has been corrected as “scores” in the revised manuscript (please see “Materials and methods: 3.3. Molecular docking”).
Point 14: Line 257-259: “We further performed the sensitivity by synthesizing two new fluorescent probes, FITC-GPMQTSpTPKNG-OH for PLK2-PBD and FITC-GPLATSpTPKNG-OH for PLK3-PBD, respectively”. These probe structures are not “new” sequences. They were reported in several publications as fluorescent probes for their corresponding PBD isoforms.
Response 14: Thank you for the suggestion. The word “new” has been removed in the revised manuscript (please see “Materials and methods: 3.4. In vitro PLK1-PBD inhibition assay”).
Reviewer 3 Report
In my opinion this work is clear and well presented. I have only one question for the authors. Why they used MMFF94 force field and not Amber, in my opinion a more specific forcefield for proteins and peptides.
Author Response
Response to Reviewer 3 Comments
Point 1: English language and style
( ) Extensive editing of English language and style required
( ) Moderate English changes required
(x) English language and style are fine/minor spell check required
( ) I don't feel qualified to judge about the English language and style
Response 1: Thank you for the suggestion. We have carefully corrected some grammatical errors in the revised manuscript.
Point 2: In my opinion this work is clear and well presented. I have only one question for the authors. Why they used MMFF94 force field and not Amber, in my opinion a more specific forcefield for proteins and peptides.
Response 2: Thank you for the suggestion. According to the manual of molecular operating environment (MOE), Amber is an all-atom forcefield parameterized for proteins and nucleic acids. This forcefield is not suitable for most small organic molecules and peptides (Manual of molecular operating environment (MOE), Version 2007.09). In previous studies, we have successfully used MMFF94 force field to perform energy minimization of for proteins and peptides (Molecules, 2019, 24, 3181; Adv. Funct. Mater. 2019, 1904969: DOI: 10.1002/adfm.201904969).
Author Response File: Author Response.docx
Reviewer 4 Report
The manuscript entitled “Structure-Based Virtual Screening and Biological Evaluation of Peptide Inhibitors for Polo-Box Domain” done by Fang Yan et al., utilizes the computational tools to design the PLK1-PBD inhibitors which inhibits the cell cycle regulatory proteins inducing mitotic arrest at the G2/M phase cell cycle in many diseases.
Exploring the beneficial characteristics of p-Cdc25C and cell cycle regulatory proteins by affecting the PLK1-PBD inhibitors might have high value by reducing the mitotic arrest at the G2/M phase and testing these compounds, activities in in vitro biological assays are interesting and more valuable.
The author done the virtual screening and pharmacophore model generations and testing the structure-activity relationship of the designed compound are more valid in cycle arrest in HeLa cells and in vitro PLK1-PBD inhibition assays.
PLK1-PBD inhibition assays for the properties validated in vitro biological assays. The work is oriented towards the characterization and validations of the PLK1-PBD peptide inhibitors through the biological assays and validated the IC50 values for the most active compounds is more added and value for the synthesized compounds.
The methodology followed was really impressive and interesting. The authors have well utilized the chemical synthesis procedure designing in silco screening and in vitro validations to characterize the compound properties to validate the molecular properties of PLK1-PBD peptide inhibitor compounds more valuable to inflammation or auto-immune diseases. The methods were explained in a detailed and systematic manner and the corresponding results were discussed in an interactive way. The paper has been written well with clear conclusions. Beyond the potentiality of the manuscript, I have a few minor concerns that the authors may address.
Minor: Consider the more recent studies in pharmacophore based drug discovery in silico studies like doi.org/10.3390/cells8030260, doi.org/10.1016/j.drudis.2018.11.005 are more added value to this manuscript.
Author Response
Response to Reviewer 4 Comments
Point 1: English language and style
( ) Extensive editing of English language and style required
( ) Moderate English changes required
(x) English language and style are fine/minor spell check required
( ) I don't feel qualified to judge about the English language and styleResponse 1: Thank you for the suggestion. We have carefully corrected some grammatical errors in the revised manuscript.
Point 2: Minor:
Consider the more recent studies in pharmacophore based drug discovery in silico studies like doi.org/10.3390/cells8030260, doi.org/10.1016/j.drudis.2018.11.005 are more added value to this manuscript.
Response 2: Thank you for the suggestion. The recent studies in pharmacophore based drug discovery like the literatures (doi.org/10.3390/cells8030260, doi.org/10.1016/j.drudis.2018.11.005) have been added to the revised manuscript (please see “References 15 and 16”).
Author Response File: Author Response.docx
Round 2
Reviewer 2 Report
Thank you to the authors for incorporating the suggested comments into their manuscript. The additional details provide more information on the design and scope of the work.
This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.
Round 1
Reviewer 1 Report
The authors build pharmacophore model for PLK1-PBD peptide inhibitor from three protein structure.
The model was first verified with a 2000 compounds including 15 active inhibitors. The model get 18 hits with 13 of them are active.
Then 32,768 peptides are screened by model, with hits are ranked by RMSD. low RMSD hits are minimized and interaction energies are used to rerank.
The 5 with lowest interaction energy are then measured with inhibitor effect, IC50. The 5th peptide, with highest inhibitor effect are tested in Hela cell, and show stronger inhibit effect than control on expression of p-Cdc25C and cell cycle.
The result and conclusion are reasonable and details are provided.
Minors,
line 92 "The E value for pharmacophore model was 96 as it hasidentified 13 active hits from 18 screened molecules". "screened" is not proper here, as total 2000 compounds are screened and the 18 are hits. line 471 "Table 2. Results of RMSD values and docking scores of the 5 selected peptides." The main text for the table is not proper as IC50 is also included. Could authors list more number of peptide identified from virtual screening in main text or supplementary to give more candidates for reader? As the docking result and IC50 result doesn't match well, it's possible some lower hits from virtual screening have better inhibition than the top 5 measured. For example, the result show the 5th rank in virtual screening is better in inhibition than the top 4; if only top 4 are measured, the best, 5th, would have been lost. Do all the 340 peptides with <1 RMSD have docking scores?
Reviewer 2 Report
Recommendation: Reject.
The work “Structure-based pharmacophore modeling, virtual screening, molecular docking and biological evaluation for identification of potential polo-like kinase 1 polo-box domain inhibitors” of Yan et al. developed a computer-assisted strategy to screen for Polo-box domain of polo-like kinase 1 (PLK1-PBD) peptide inhibitors using a combination of structure-based pharmacophore modeling, virtual screening and molecular docking studies. Regretfully, there are no evidence that this computational protocol was well developed.
I want to indicate these points:
It is hard to understand how authors derived pharmacophore model by using only PDB structures. Authors should indicate how they selected the pharmacophoric features by using only a binding site where there are too many features to select. It is not enough to say that “the pharmacophore was done by using the pharmacophore generation protocol within MOE”, because MOE needs structural information of the ligands to select the most relevant features to create this model. Nine pharmacophoric features to describe a binding site looks not very suitable, there are too many features for a right screening.
It is hard to understand Figure 1 since there is no ligand to orient how ligands are oriented in this model.
Authors said that they used “a database including 1985 inactive and 15 active molecules collected from the reported literature”. They should explain how inactive compounds were considered. Without a definition, an inactive compound is ‘any compound in the world’. Authors should provide the structures of the tested compounds in the supplementary material.
In docking section the docking protocol is not well described. The only description is that docking was done with MOE, but authors did not describe the protocol (size of grid, parameters of the search, etc). Authors did not indicate how they select their best solutions and how they demonstrate that their solutions are not artifacts.
Five compounds were found during this very strange protocol. The manuscript has no evidences of the orientation of these compounds with respect to the pharmacophore model and docking. In general, the discussion of the modeling results is weak and has no value for readers interested in the design of novel polo-like kinase 1 polo-box domain inhibitors.
Materials and methods section must show the PDB codes of the structures used in this study. Such information is very relevant for readers in this section.
Authors must check English usage. There many grammar problems and typos in the manuscript.
Reviewer 3 Report
This paper described the development and validation of a pharmacophore model to isolate peptide sequences that are predicted to bind the Plk1 PBD with high affinity. The development of the pharmacophore model seems to be relatively straightforward based on published X-ray data of ligand bound PBD structures. However little information was provided for the 15 active and 1985 inactive sequences used to develop the model. The source and/or contents of the ~32,000 sequences screened was also not described. The resulting 5 peptide hits were synthesized and found to inhibit the PBD in various in vitro validation assays.
The design and ambition of the paper is appreciated, however the results fall short of its potential. The 5 peptide hits predicted by the pharmacophore model all contain the core "PLHSpT" 5-mer sequence that has been published in numerous articles as a core PBD-binding motif. This core structure has been the basis for dozens of papers that utilize this peptide as a starting point for further SAR. Moreover, the 5 hits also highly resemble the PBD-binding ligands published by Abell and co-workers in 2011 and 2012, so the observation that they are able to modestly inhibit Plk1 in vitro is not novel. The manuscript concludes with the statement, "The protocol can be also used for virtual screening of other chemical databases to identify potent PLK1-PBD inhibitors with unknown scaffolds." Utilizing this model to discover or predict a novel or non-phosphorylated PBD-binding ligand would have been far more impactful.
There are several statements that could use elaboration, citing, or re-phrasing. One major issue is the described basis of the work, which the authors indicate to be poor affinity or selectivity of PBD-binding inhibitors. There are many paper that have reported high affinity and highly selective ligands of the Plk1 PBD. However very few have advanced as therapeutics due to the requirement of the anionic pSer/pThr residue and its resulting poor cellular penetration. This needs to be address in the paper. More detailed comments are provided on a line-by-line basis below:
line 12: grammar, should be “The Polo-box domain”
lines 13-16: “Although therapeutic peptides are relative safe and well-tolerated compared with conventional small molecules, their weak affinity and poor selectivity are a prominent problem for development of PLK1-PBD peptide inhibitors.”
This statement does not accurately describe the state of peptide inhibitors of the Plk1PBD. Nearly all published inhibitors demonstrate appreciable affinity, with some demonstrating single-digit nanomolar binding affinities. Also, poor selectivity is rarely an issue with Plk1 PBD peptide inhibitors based on the currently available literature.
line 25: grammar, should be “The centrosome”
line 40: grammar, “previous researches” should be “previous researchers”
line 41: grammar, “make itself” should be “makes it”
line 45: grammar, “catalytic domain” should be “the catalytic domain”
line 46: “various adverse effects can be observed.”
This observation should be cited with relevant literature.
lines 51-53: “Moreover, peptides are easier to synthesize, and have lower immunogenicity, better cell-penetrating and targeting ability compared with proteins or antibodies.”
This statement is somewhat irrelevant to the scope of the paper, and seems like a generalized opinion rather than a citable statement of fact. The comparison between peptides and protein-based therapeutics is highly case-dependent as to which format will work best for a given target.
lines 56-58: “However, due to some problems such as poor affinity, drug resistance or low specificity, the clinical use of these PLK1-PBD inhibitors has been restricted. Therefore, there is a pressing need to develop novel high-affinity PLK1-PBD inhibitors.”
This statement also does not reflect the state of the field for peptide-based Plk1-PBD inhibitors. Poor affinity and poor selectivity are rarely issues for published PBD-binding inhibitors, as there are dozens of literature articles available with nanomolar-affinity inhibitors of the PBD that demonstrate >100X selectivity for the Plk1 PBD vs Plks 2 and 3. I also do not understand “drug resistance” to be an issue for Plk1 PBD inhibitors in the literature. Perhaps this point could be further elaborated. The major issue of PBD inhibitors, as address by nearly all articles published on their development, is the dependence of binding of the phospho-Ser or phosphor-Thr motif and the resulting poor cellular penetration due to this di-anionic motif. This factor was not addressed at all in the manuscript.
line 88: A summary or table in the SI for the “15 active molecules” would be helpful. The molecules seem to be cited further in the manuscript, but the exact sequences used to develop the pharmacophore model should be included.
line 141: “inhibit the expression of p-Cdc25C” would make more sense written as “inhibit the phosphorylation of Cdc25C”.
line 162: “statistical parameters” is duplicated.
line 189: The probe used in the competitive fluorescence polarization is not described in any way. What is the sequence used? Where is it fluorophore modified? What is the affinity of the probe being used?
line 200: “5 x 105” should be “5 x 105”
line 201: “Cells were then treated with various concentrations of peptide 5 for 72 h.”
How was peptide 5 dosed onto cells? Was it a DMSO stock? If so, what concentration of DMSO in the assay? What concentration of RNase and propidium iodide were used?
