Integration of Transcriptomics and Metabolomics for Pepper (Capsicum annuum L.) in Response to Heat Stress

Round 1

Reviewer 1 Report

In this study, authors investigated the transcriptomic and metabolomic responses to heat stress in two pepper cultivars, both heat-tolerant and sensitive. Their results showed some genes and pathways play important roles in the heat resistance, including the accumulation of osmotic adjusting materials and signal transduction. 

Overall, the manuscript is clearly written. The experiments and bioinformatic analysis  were carefully carried out. However, I still have a few concerns that must be clarified or addressed.

Major points:

1) In Line 341, the low-quality reads were filtered out from the datasets, but authors did not mention what kind of reads are of low quality.

2) Since there are the three replicates for each cultivar or condition, why did mapping each of the library to the reference genome separately rather than map the reads of all the three replicates together with one single run? HISAT2 can deal with replicate data using one command and there is no need to map them individually and then merge together.

3) In this study, gene expression levels were measured by FPKM values and DEGs were detected by DESeq2, however, DESeq2 requires raw read counts instead of FPKM. Using FPKM will violate the basic assumption of DESeq2's statistical model, and reduce the accuracy of DEG prediction.

4) In Line 394, authors should provide the details that how transcriptomic and metabolomic data were converted to the same model. And what kind of information did employ  for correlation analysis between genes and metabolites?

5) For GO and KEGG analysis, what significance criteria did use for enrichment analysis? And in Line 135-137, DEGs were said to assign into 10, 16 and 21 GO terms. Are these terms of significant enrichment?

6) In Figure 4, the color for the legend of Pvalue is hard to read. According to the plot, we can only know that all the p-values of the all the terms listed are less then 0.5, which is not a general view of statistical significance. I would suggest to use white color to represent p-value = 0.05, and red and blue colors range from 0-0.05 and 0.05-1, respectively.

7) In Line 182, what do the QC samples and inspected samples mean?

8) In Figure 7A, what does the mixed group mean?

9) In Line, 213, authors said DEGs and DAMs were categorized into the same group. I just wonder how did they been categorized? Please provide some details .

10) For qPCR validation, Figure 2 only shows the qPRC results, and there is no comparison between qPCR and RNA-seq results, thus there is no evidence supporting the consistency between these two measurements.

11) Some references that should be cited are not included, including the references for DESeq2 in Line 346, iTAK in Line 351 and all the databases used in Line 378. Please check the entire manuscript carefully to make sure all the views published in other studies, software and public databases used in this studies have been cited.

Minor points:

1) In Line 226, the red and blue numbers represent the log2 fold changes, but the fold change of what? And in the same sentence, what does 2144 mean?

2) All the fill names of the abbreviations used in the main text should be provided when they first appear. Although some of them have been provided in Materials and Methods, they have been use multiple times before the definition, which is inconvenient for readers. Here I only list some of the cases and please check the entire manuscript to make sure all the full names have been provided when they first appear.

In Figure 1, "RCK", "RT", SCK" and "ST";

In Line 135, "GO";

In Line 142, "KEGG";

In Line 193, "DAMs";

In Line 384, "VIP.

3) In the main, please define which cultivar is heat-tolerance or sensitive. Although they have been mentioned in the abstract, I would prefer to define them again in the main text at where they first appear, in order to help readers to follow the main story of this manuscript.

4) In Line 132-133, authors showed there are 1770 up-regulated genes and 1529 down-regulated genes, while in previous paragraph, they also mentioned two percentages of the same gene sets in Line 124. I would suggest two to merge these two sentence together since they talk about the same thing.

Author Response

Response to Comments of manuscript “Integration of Transcriptomics and Metabolomics for Pepper (Capsicum annuum L.) in Response to Heat Stress”

(ijms-599090)

In this study, authors investigated the transcriptomic and metabolomic responses to heat stress in two pepper cultivars, both heat-tolerant and sensitive. Their results showed some genes and pathways play important roles in the heat resistance, including the accumulation of osmotic adjusting materials and signal transduction.

Overall, the manuscript is clearly written. The experiments and bioinformatic analysis were carefully carried out. However, I still have a few concerns that must be clarified or addressed.

Major point 1：In Line 341, the low-quality reads were filtered out from the datasets, but authors did not mention what kind of reads are of low quality.

Response major point 1: Thank you for your comment. The sequencing raw reads were filtered when the reads with ambiguous nucleotides > 10%, and reads in which the low quality (Q≤5) base number > 50%. We have added the type of low-quality reads in line 362-363.

Major point 2：Since there are the three replicates for each cultivar or condition, why did mapping each of the library to the reference genome separately rather than map the reads of all the three replicates together with one single run? HISAT2 can deal with replicate data using one command and there is no need to map them individually and then merge together.

Response major point 2：Thank you for your comment. In our study, we mapped each library to the reference genome to get expression information of each sample. If we mapped replicate data to reference genome, HISAT2 will deal the data as one single sample. There is no point to include three replicates in this paper.

Major point 3: In this study, gene expression levels were measured by FPKM values and DEGs were detected by DESeq2, however, DESeq2 requires raw read counts instead of FPKM. Using FPKM will violate the basic assumption of DESeq2's statistical model, and reduce the accuracy of DEG prediction.

Response major point 3: Thank you for pointing this out. We have corrected the wrong description in line 365-367.

Major point 4: In Line 394, authors should provide the details that how transcriptomic and metabolomic data were converted to the same model. And what kind of information did employ for correlation analysis between genes and metabolites?

Response major point 4：Thank you for your valuable advice. According to your comments, we have added the detailed description about data processing. In line 417-420.

Major point 5: For GO and KEGG analysis, what significance criteria did use for enrichment analysis? And in Line 135-137, DEGs were said to assign into 10, 16 and 21 GO terms. Are these terms of significant enrichment?

Response major point 5：We used the pvalue and qvalue of hypergeometric distribution for KEGG and GO enrichment analysis, respectively. We have redrawn the GO figure and these GO terms were significantly enriched.

Major point 6: In Figure 4, the color for the legend of Pvalue is hard to read. According to the plot, we can only know that all the p-values of the all the terms listed are less than 0.5, which is not a general view of statistical significance. I would suggest to use white color to represent p-value = 0.05, and red and blue colors range from 0-0.05 and 0.05-1, respectively.

Response major point 6：Thank you for your valuable advice. We have redrawn the figure 4 in the revised manuscript (line162).

Major point 7: In Line 182, what do the QC samples and inspected samples mean?

Response major point 7: Sorry for the confusion here. “Inspected samples” was a bad choice of words and we have used “tested samples” to replace it (196-197). The QC samples were mixed by all test samples and inserted into each test sample to check the repeatability of the analytical process. We have added this description in the revised manuscript (line 379-380).

Major point 8: In Figure 7A, what does the mixed group mean?

Response major point 8: Sorry for the confusion here. The mixed group means QC samples used in this study and we have added this description in line 207.

Major point 9: In Line, 213, authors said DEGs and DAMs were categorized into the same group. I just wonder how did they been categorized? Please provide some details.

Response major point 9: Sorry for the confusion here. We used inappropriate word and we corrected it line (line 230).

Major point 10: For qPCR validation, Figure 2 only shows the qPRC results, and there is no comparison between qPCR and RNA-seq results, thus there is no evidence supporting the consistency between these two measurements.

Response major point 10: Thank you for your comment. We have added the RNA-seq results into Figure 2 (line121).

Major point 11: Some references that should be cited are not included, including the references for DESeq2 in Line 346, iTAK in Line 351 and all the databases used in Line 378. Please check the entire manuscript carefully to make sure all the views published in other studies, software and public databases used in this study have been cited.

Response major point 11: Thank you for your valuable advice. We have cited the references and databases (line 367, 372,401-402) in the revised manuscript.

Minor points 1: In Line 226, the red and blue numbers represent the log2 fold changes, but the fold change of what? And in the same sentence, what does 2144 mean?

Response to minor points 1: Thank you for your comment. The red and blue numbers represent the log2 fold changes of differentially expressed genes or differentially accumulated metabolites. We have replaced 2144 with 17CL30 (line243-245).

Minor points 2: All the fill names of the abbreviations used in the main text should be provided when they first appear. Although some of them have been provided in Materials and Methods, they have been used multiple times before the definition, which is inconvenient for readers. Here I only list some of the cases and please check the entire manuscript to make sure all the full names have been provided when they first appear.

In Figure 1, "RCK", "RT", SCK" and "ST";

In Line 135, "GO";

In Line 142, "KEGG";

In Line 193, "DAMs";

In Line 384, "VIP.

Response to minor points 2: Thanks for pointing this out. We have corrected all abbreviations used in the entire manuscript.

Minor points 3: In the main, please define which cultivar is heat-tolerance or sensitive. Although they have been mentioned in the abstract, I would prefer to define them again in the main text at where they first appear, in order to help readers to follow the main story of this manuscript.

Response to minor points 3: Thank you for your careful reading of our manuscript, and we have added in line 84.

Minor points 4: In Line 132-133, authors showed there are 1770 up-regulated genes and 1529 down-regulated genes, while in previous paragraph, they also mentioned two percentages of the same gene sets in Line 124. I would suggest two to merge these two sentences together since they talk about the same thing.

Response to minor points 4: Thank you for your valuable advice. We have merged these two sentences together in line 131-132.

Author Response File: Author Response.pdf

Reviewer 2 Report

This manuscript needs grammar/style editing before publication.

How closely related are these two cultivars?  It would be useful to see if any of the sequences of the genes of interest vary between the two cultivars, as this may further inform on the heat stress tolerance adaptation.

Figure 2 needs a complete legend in order to be comprehensible on its own.  The plot titles should be made clearer as well.  What is given currently cannot be readily understood without reference to the SI.

I would recommend citation of KEGG.

In figures 4-6, the scale circles are not sized in a way that covers the full range of results.  This makes the figures harder to interpret and I would recommend correcting this.

The metabolite heatmap in figure 7B contains so many metabolites and conditions as to make it next to impossible to glean useful information about any specific metabolite.  A more detailed discussion would help here.

The abbreviated names of the samples are extensively used before they are explained in the Materials and Methods.  This should be established earlier.

Line 370 - “The collision gas (CAD) was high” needs clarification.  Does this mean high-energy CID?

Considering the wealth of information obtained by these techniques, there is a lack of interpretation of this data which cannot be ignored.  The discussion mostly covers broad pathways without addressing which specific steps are significantly affected by heat stress, while there is no attempt made to explain why anything is expressed differently.  This severely curtails the potential impact of this paper and leaves a valuable dataset practically unused.  I would not call this a complete manuscript without an expanded analysis of the data obtained.  The technique is not novel, so the most valuable information that can be obtained here would be the actual alterations to protein and metabolite expression, which are too vague to provide much benefit as far as specific proteins and metabolites involved in heat stress response.

Author Response

Response to Comments of manuscript “Integration of Transcriptomics and Metabolomics for Pepper (Capsicum annuum L.) in Response to Heat Stress”

(ijms-599090)

Point 1: This manuscript needs grammar/style editing before publication.

Response point 1: We are sorry for the grammar and writing errors. We have carefully revised the manuscript to polish the English.

Point 2: How closely related are these two cultivars? It would be useful to see if any of the sequences of the genes of interest vary between the two cultivars, as this may further inform on the heat stress tolerance adaptation.

Response point 2: Thank you for your comment. These two cultivars belong to the same species. The trait of heat-tolerance in pepper is a very complex quantitative trait, which needs lots of genes and different types of heat-tolerant mechanism to coordinate. In this study, we integrated the transcriptomics and metabolomics data to understand the extremely complex regulatory mechanisms under heat stress in pepper. So many genes and metabolites were considered to be related to heat stress. And we found that glutathione metabolic pathway played a critical role in pepper response to heat stress. In future studies, we will have a more in-depth description of the function of the key genes in this pathway.

Point 3: Figure 2 needs a complete legend in order to be comprehensible on its own. The plot titles should be made clearer as well. What is given currently cannot be readily understood without reference to the SI.

Response point 3: Thanks for pointing this out. We have changed the figure and added a complete legend in line 122-125.

Point 4: I would recommend citation of KEGG.

Response point 4: Thank you for your valuable advice. We have cited KEGG in line 369.

Point 5: In figures 4-6, the scale circles are not sized in a way that covers the full range of results. This makes the figures harder to interpret and I would recommend correcting this.

Response point 5: Thank you for your comment. We have modified these pictures in line 162,165 and167.

Point 6: The metabolite heatmap in figure 7B contains so many metabolites and conditions as to make it next to impossible to glean useful information about any specific metabolite. A more detailed discussion would help here.

Response point 6: Thank you for your careful reading of our manuscript. We have redrawn this figure used all differentially accumulated metabolites in line 203.

Point 7: The abbreviated names of the samples are extensively used before they are explained in the Materials and Methods. This should be established earlier.

Response point 7: Thanks for pointing this out. We have added abbreviated names of the samples in line 105-107.

Point 8: Line 370 - “The collision gas (CAD) was high” needs clarification. Does this mean high-energy CID?

Response point 8: Sorry for the misuse of words. We have replaced the “was” with “was set at” (392-393).

Point 9: Considering the wealth of information obtained by these techniques, there is a lack of interpretation of this data which cannot be ignored. The discussion mostly covers broad pathways without addressing which specific steps are significantly affected by heat stress, while there is no attempt made to explain why anything is expressed differently. This severely curtails the potential impact of this paper and leaves a valuable dataset practically unused. I would not call this a complete manuscript without an expanded analysis of the data obtained. The technique is not novel, so the most valuable information that can be obtained here would be the actual alterations to protein and metabolite expression, which are too vague to provide much benefit as far as specific proteins and metabolites involved in heat stress response.

Response point 9: We thank the reviewer for pointing this out. In this study, transcriptomics and metabolomics analyses were used to investigate the heat stress tolerance mechanism of pepper. We found that the content of glutathione was increased much more dramatically in heat-tolerant cultivar under heat stress. As a result, we integrated the transcriptomics and metabolomics data related to glutathione metabolic pathway for further research contributed to rising content of glutathione. GSTs were identified as key genes in this pathway. Up-regulated expression of GSTs can promote the synthesis of glutathione which can reduce the damage of toxic substances caused by various stresses. We will further study the function of genes in the future.

Author Response File: Author Response.pdf

Reviewer 3 Report

The manuscript contains data about the effect of elevated temperature on two cultivars of pepper. A transcriptomic and metabolomic approach was applied in the research. In my opinion, results are interesting and I am sure that they will be also interesting to the wider scientific community. The results presented in the manuscript should be published, but first, the manuscript must be improved.

Step by step, following the manuscript.

Keywords must be changed because keywords should not be the same as words already used in a title. Figure 1E presents Total protein content. The problem is, that the Bradford method does not measure total protein content. The Bradford method allows measuring only soluble proteins. Another problem concerns the unit of protein content (mg/ml). In my opinion, the protein content should be expressed in mg g FW^-1, similarly as MDA, soluble sugars and proline content. Instead of ‘/’ use -1 in superscript. And finally, please change the reference no 30. If the Bradford method was used, the original paper should be cited (Bradford MM (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72:248-254).

General advice concerning all figures. Figures must be self-explanatory, so all abbreviations must be explained in the figures’ captions. Alternatively, abbreviations like RCK, RT, SCK, ST may be added to the Abbreviation list. Moreover, figures’ captions should be more detailed and they should contain a brief description of results presented in figures. The caption of Figure 1 is enough, but captions of all other figures must be improved.

Line 112-113. ‘334,699,604 and 344,747,058 clean reads’ are rather not presented in Supplementary Table 1. It must be corrected. Line 114-116 and Figure 2. I have many doubts about the randomly selected genes for qPCR analysis validating the RNA-seq data. Why these genes were selected randomly? In my opinion, the genes should be precisely selected. For qPCR genes of, for example, HSP, glutathione metabolism or highly up- or down-regulated should be chosen. Moreover, it is written that ‘The results showed that the expression profile of these genes WAS COINCIDENT with that of RNA-seq analysis (Figure 2)’, but no coincidence is visible in Figure 2. Figure 2 presents only relative gene expression obtained by qPCR. Additionally, all genes are up-regulated, but in Material and methods (line 401) is written, that up- and down-regulated genes were selected for qPCR. In my opinion, this part of the research must be performed again, and another set of genes should be chosen and tested by qPCR. Lines 120-122. Instead of DEGs (4 times), it should be written ‘genes’. For example ‘up-regulated genes’, not ‘up-regulated DEGs’. It should be written like in lines 132-133. Figure 3 is difficult to read. The font size must be increased. Line 142. KEGG is used for the first time, so the full name should be written here. Figure 8 is very difficult to read. The font size must be increased. Line 215-216 ‘Interestingly, we found that “glutathione metabolism” was markedly ACCUMULATED and the detailed network of this pathway was mapped’. A metabolism rather cannot be accumulated. Maybe ‘affected’ will be better? Figure 9. Why L-Serine is in green? What does 2144 mean in the figure caption? Should 17CL30 be? All abbreviations must be explained in the figure caption. The ending of the Discussion must be changed. It looks like it has no end. The last sentence is not appropriate. Moreover, Nikolaos is the first name, so instead of ‘Nikolaos’s [63]’ should be ‘Labrou’s [63]. Line 320-321. The explanation of why the 28-h period of HS was applied is not enough. I do not see that the selection of the 28-h period of HS is proven by Figure 1AB. It must be precisely explained why the 28-h period was applied. Line 382-338. The number of replicates of the transcriptome sequencing must be added in this subsection.

Author Response

Response to Comments of manuscript “Integration of Transcriptomics and Metabolomics for Pepper (Capsicum annuum L.) in Response to Heat Stress”

(ijms-599090)

The manuscript contains data about the effect of elevated temperature on two cultivars of pepper. A transcriptomic and metabolomic approach was applied in the research. In my opinion, results are interesting and I am sure that they will be also interesting to the wider scientific community. The results presented in the manuscript should be published, but first, the manuscript must be improved. Step by step, following the manuscript.

Point 1: Keywords must be changed because keywords should not be the same as words already used in a title.

Response point 1: Thank you for your comment. We have changed keywords in line 31-32.

Point 2: Figure 1E presents Total protein content. The problem is, that the Bradford method does not measure total protein content. The Bradford method allows measuring only soluble proteins. Another problem concerns the unit of protein content (mg/ml). In my opinion, the protein content should be expressed in mg g FW-1, similarly as MDA, soluble sugars and proline content. Instead of ‘/’ use -1 in superscript. And finally, please change the reference no 30. If the Bradford method was used, the original paper should be cited (Bradford MM (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72:248-254).

Response point 2: Thank you for pointing this out. We have corrected the unit of four substances (line 103). We are very sorry for our wrong citations. In this research, we used BCA method to measure total protein content. We have cited the correct reference in line 415.

Point 3: General advice concerning all figures. Figures must be self-explanatory, so all abbreviations must be explained in the figures’ captions. Alternatively, abbreviations like RCK, RT, SCK, ST may be added to the Abbreviation list. Moreover, figures’ captions should be more detailed and they should contain a brief description of results presented in figures. The caption of Figure 1 is enough, but captions of all other figures must be improved.

Response point 3: Thank you for your valuable advice. We have revised the captions of all figures. And all abbreviations used in this manuscript were added in the Abbreviation list.

Points 4: Line 112-113. ‘334,699,604 and 344,747,058 clean reads’ are rather not presented in Supplementary Table 1. It must be corrected.

Response point 4: Thank you for your comment. We have added this information in Supplementary Table 1.

Point 5: Line 114-116 and Figure 2. I have many doubts about the randomly selected genes for qPCR analysis validating the RNA-seq data. Why these genes were selected randomly? In my opinion, the genes should be precisely selected. For qPCR genes of, for example, HSP, glutathione metabolism or highly up- or down-regulated should be chosen. Moreover, it is written that ‘The results showed that the expression profile of these genes WAS COINCIDENT with that of RNA-seq analysis (Figure 2)’, but no coincidence is visible in Figure 2. Figure 2 presents only relative gene expression obtained by qPCR. Additionally, all genes are up-regulated, but in Material and methods (line 401) is written, that up- and down-regulated genes were selected for qPCR. In my opinion, this part of the research must be performed again, and another set of genes should be chosen and tested by qPCR.

Response point 5: Thank you for your careful reading of our manuscript. We have performed qRT-PCR again. Eight genes were chosen for qRT-PCR, including HSP20, HSP90, GST, NAC and WRKY40, which are related to heat stress. We have also added the RNA-seq results into Figure 2 line 121.

Point 6: Lines 120-122. Instead of DEGs (4 times), it should be written ‘genes’. For example, ‘up-regulated genes’, not ‘up-regulated DEGs’. It should be written like in lines 132-133.

Response point 6: We have corrected this in line 127 and 129and thank you for pointing this out.

Point 7: Figure 3 is difficult to read. The font size must be increased.

Response point 7: Thank you for your advice. We have increased the font size of Figure 3 (line 136).

Point 8: Line 142. KEGG is used for the first time, so the full name should be written here.

Response point 8: We have added the full name of KEGG in line (155) and thank you for pointing this out.

Point 9: Figure 8 is very difficult to read. The font size must be increased.

Response point 9: We have increased the font size of Figure 8 (line 222) and thank you for your valuable advice.

Point 10: Line 215-216 ‘Interestingly, we found that “glutathione metabolism” was markedly ACCUMULATED and the detailed network of this pathway was mapped’. A metabolism rather cannot be accumulated. Maybe ‘affected’ will be better?

Response point 10: Thank you for your valuable advice and we have replaced “accumulated” with “affected” in line 233.

Point 11: Figure 9. Why L-Serine is in green? What does 2144 mean in the figure caption? Should 17CL30 be?

Response point 11: Thank you for pointing this out. We have corrected the Figure 9 and the figure caption in line 240 and 244.

Point 12: All abbreviations must be explained in the figure caption.

Response point 12: Thank you for your comment. We have corrected the all abbreviations in revised manuscript.

Point 13: The ending of the Discussion must be changed. It looks like it has no end. The last sentence is not appropriate. Moreover, Nikolaos is the first name, so instead of ‘Nikolaos’s [63]’ should be ‘Labrou’s [63].

Response point 13: We have rewritten the end of discussion and replaced “Nikolaos” with “Labrou” in line 329.

Point 14: Line 320-321. The explanation of why the 28-h period of HS was applied is not enough. I do not see that the selection of the 28-h period of HS is proven by Figure 1AB. It must be precisely explained why the 28-h period was applied.

Response point 14: Thank you for pointing this out. Such treatment duration was chosen because the phenotypic differences between two cultivars were greatest after 28-h heat treatment. At this time, leaves of 05S180 (heat-sensitive cultivar) all wilted, whereas few leaves began to curl in 17CL30 (heat-tolerant cultivar). We have added this in line 341-342 and line 95-96.

Point 15: Line 382-338. The number of replicates of the transcriptome sequencing must be added in this subsection.

Response point 15: We have added the number of replicates of transcriptome sequencing in line 344.

Round 2

Reviewer 1 Report

 I am happy to see that all my comments have been addressed. I have only one more concern on qPCR validation. In the revised manuscript, none of genes being validated in previous version have been included now. Thus, I wonder why authors perform validation with different eight genes, is that because the transcripome and qPCR results are not consistent?

Author Response

Point 1：I am happy to see that all my comments have been addressed. I have only one more concern on qPCR validation. In the revised manuscript, none of genes being validated in previous version have been included now. Thus, I wonder why authors perform validation with different eight genes, is that because the transcripome and qPCR results are not consistent?

Response point 1: Sorry for the confusion here. Validation with different genes was not because their inconsistent transcriptome and qPCR results. It was done according to the suggestion of another reviewer who recommended to use some specific genes such as HSP, GST, and others. To follow the reviewer’s suggestion, we validated with different eight genes.

Reviewer 2 Report

With the increased focus on glutathione metabolism, this manuscript is overall more cohesive, though likely less impactful than it could be if more of the data the authors obtained was addressed.  I would strongly encourage the authors to publish interpretations of the other data obtained, as there is likely significant information to be found therein.

This manuscript is still in considerable need of edits for English quality, though the revised version is quite scientifically solid, much clearer on technical points, and I would recommend publication after general grammar editing.

Author Response

Point 1: With the increased focus on glutathione metabolism, this manuscript is overall more cohesive, though likely less impactful than it could be if more of the data the authors obtained was addressed. I would strongly encourage the authors to publish interpretations of the other data obtained, as there is likely significant information to be found therein.

Response point 1: Thank you for your valuable advice. We will seriously consider publication of the interpretation of other data obtained.

Point 2: This manuscript is still in considerable need of edits for English quality, though the revised version is quite scientifically solid, much clearer on technical points, and I would recommend publication after general grammar editing.

Response point 2: Sorry for my poor English. We have sought professional editing (MDPI English editing) to further improve the English writing.

Reviewer 3 Report

The manuscript was considerably improved. All of my suggestions and requirements are introduced to the text, thus I have no objections and I can recommend acceptance of the manuscript. However, I have suggestions regarding figure 2. Firstly, the relative gene expression is not a continuous value, so it cannot be shown as a line graph. It must be presented only by bars or maybe by points, asterisks, etc. Secondly, the resolution of the picture must be increased.

Author Response

Point 1: The manuscript was considerably improved. All of my suggestions and requirements are introduced to the text, thus I have no objections and I can recommend acceptance of the manuscript. However, I have suggestions regarding figure 2. Firstly, the relative gene expression is not a continuous value, so it cannot be shown as a line graph. It must be presented only by bars or maybe by points, asterisks, etc. Secondly, the resolution of the picture must be increased.

Response point 1: Thank you for pointing this out. We have revised Figure 2 accordingly (line 122).

Round 3

Reviewer 1 Report

The authors have addressed my previous concerns properly. I don't have any additional comments.