Comparative Transcriptome and Pigment Analyses Reveal Changes in Gene Expression Associated with Flavonol Metabolism in Yellow Camellia

Round 1

Reviewer 1 Report

In the current study under review, the authors made a nice attempt to perform a comparative transcriptome and pigment analyses to reveal the changes in gene expression associated with flavonol metabolism in yellow camellia. However, some major changes need to be made to increase the readability and enhance the presentation of the obtained results.  Furthermore, some queries need to be answered/clarified before the manuscript could be judged as suitable for publication. The following are some comments, suggestions, or queries regarding the manuscript.

1.     Several typos and grammatical errors are there hindering the reader’s ability to understand the intended meaning. Some examples are added to the attached manuscript.

2.     Abstract is very long. Several sentences could be removed or shortened. Some examples are added to the attached manuscript.

3.     Introduction

a.      This section needs major improvement to clearly show the drive force of the study and the importance of producing yellow cultivars of camellia in terms of market demand.

b.     Studies analyzed the transcriptomic changes and their relation to the biosynthesis of flavonoids is vase; however, the authors didn’t mention a single study of them. Studies applied the similar approach followed by the authors should be added to the introduction section.

4.     Materials and methods

a.      What were the growth conditions under which the studied plants were growing?

b.     Add citation for Fastp software!!!

c.      How was the quality of the de novo assembled transcriptome assessed? Please provide the data for QC of this assembly as a supplementary.

d.     Bowtie aligner was used to align reads as it required by the RSEM tool. Why didn’t the authors used bowtie2? I think bowtie2 is better in terms of aligning reads to long reference genomes. However, the authors didn’t mention the length of the assembled transcriptome. They have 78064 unigenes represented by 104037 which is very long genome to align against using bowtie.

e.      What tools were used for GO and KEGG pathway enrichment analyses of the identified DEGs?

f.      Were the up- and down-regulated DEGs used for GO and KEGG pathway enrichment analyses separately or were they combined together before analyses? Please add more explanation to the methods section.

g.     In supplementary table 1, add the gene names of each gene based on the annotation against data bases. The reader would like to know which gene was selected for qRT-PCR and what are their roles in flavonol biosynthesis.

h.     Excel 2017 accuracy in statistical analysis is questionable. The statistical analysis should be re-performed using more specialized statistical analysis software e.g., SAS or SPSS.

5.     Results

a.      The alignment rate ranged between 58.98- 61.83%. These alignment rates are very low. Were there problems in reads quality? Were there problems in the assembly quality?

b.     Why did the threshold of log2 (FC) ≥ 4 was used? Was it randomly chosen?

c.      Several parts of the results section was improved based on the comments raised on the previous submission of this manuscript.

6.     Discussion section is good.

7.     Conclusions are fair.

Comments for author File: Comments.pdf

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 2 Report

The manuscript has been greatly improved and now I have a few more minor suggestions:

1) In our previous work which previous work? cite it line 74

2) The SBS has occurred only once, so it is better to remove the abbreviation line 107.

3) replace De Bruijn with de Brujin.

4) replace compo-nent with the component line 145.

5) Primer Premier 5.0 was used, cite it? Line 151.

6) It is unclear which three separate experiments were performed? Line 154

7)  Table 1 is too messy so adjust it accordingly, and P should be in the small case at line 178.

8) Again, In our previous work which one? better to cite it 230.

9) ACC has occurred only once, so it is better to remove the abbreviation line 335.

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 3 Report

Line 62: Replace "and so on" with "among other factors"

Line 79: Please correct" for C. nitidssim’  s yellow"

Line 79: "To do this, we used CN and its hybrids..."

Line 145: "... biological process, cellular component and molecular ..."

Line 188: "... obtained 104,037 transcript sequences and 78,064 unigenes."

Please use a 1,000 separator (,) in all numbers

Line 230: "In our previous work [citation]”

Line 286: "The results indicated that Qu7G and Qu3G might be the main flavonols affecting the yellowness of petals in camellia" It seems that the levels of Ka3G are also correlated with the yellowness of petals.

Figure 1: include abbreviations used in the text (CN, ZHQ, HXL and XSJ)

Table 1: Content in ug g-1

I suggest to plot the Table 1 as a bar plot figure, and include these plots below each flower in Figure 1.

Figure 3: You need to describe the MYB TFs and FLS and ethylene biosynthesis genes that you are evaluating.

Data shown in Figure 4 has to be described in the text.

I encourage the authors to show the transcript/metabolite data in a metabolic pathway format, including the skeletal structures of the intermediate metabolites and transcriptional data for PAL, CHS, F3H, DFR, ANS, LAR and ANR homologs known to be associated with the synthesis of flavonoids, as well as their transcriptional regulators.

Authors could do the same for the genes involved in ethylene, carotenoid and betalain biosynthesis.

In petal coloring, anthocyanin seems to be controlling the orange, red, purple, and blue colors, and betalain controls the deep red and yellow color: https://doi.org/10.1016/j.cj.2020.10.012

In this work, a phytoene desaturase with a role in carotenoid biosynthesis is shown to be involved in the yellowish flower color of Brassica napus.

Author Response

Please see the attachment

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The authors have significantly improved the manuscript. However, the question regarding the assembly QC was not answered. 

How was the quality of the de novo assembled transcriptome evaluated? 

As the authors mentioned, "Trinity" assembler was used to assemble the transcriptome. Did the authors made any measures to evaluate the quality of the resulted assembly? This part should be explained both in methods and results as it is crucial for judging the reliability of the whole downstream analysis, especially 'as I still believe' that the mapping (alignment) rates are so low in this study. 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

1. Figure 1: Please include the full name of the flavonols

2. The results shown in Figure 4 are still not described in the text.

3. The authors must show the data in a metabolic pathway format including intermediate metabolites and the levels of the genes and transcription factors involved in flavonoid, carotenoid and ethylene biosynthesis in the four different hydrids.

As an example, please see Fig. 7 in:

https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0230154

Author Response

Please see the attachment

Author Response File: Author Response.docx

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 performed a comprehensive transcriptomic analysis for flavonol metabolism in yellow camellia. The results are significant by showing a varied expression level for different cultivars. Such results will be useful for future transgenic and breeding program improvement of camelia.
The overall quality of the manuscript and the design of the experiments along with the results is good. So, I have only a few suggestions, I was just wondering the RNASeq will be made publically available? because such data is needed for the improvement of camelia, and will provide a chance for other researchers to explore this dataset.

Also, try to italic the word de novo (line-100)
The references are too messy, and I pasted them as an example. It needs to be consistent all small with the first letter as a capital, as a reference 2.

1. Sasaki, N.; Nakayama, T., Achievements and Perspectives in Biochemistry Concerning Anthocyanin Modification for Blue 329 Flower Coloration. Plant and Cell Physiology 2015, 56, (1), 28-40. 330 2. Li, X.; Fan, Z.; Guo, H.; Ye, N.; Lyu, T.; Yang, W.; Wang, J.; Wang, J. T.; Wu, B.; Li, J.; Yin, H., Comparative genomics analysis 331 reveals gene family expansion and changes of expression patterns associated with natural adaptations of flowering time and 332 secondary metabolism in yellow Camellia. Functional & integrative genomics 2018, 18, (6), 659-671.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

The manuscript by Liu et. al. has focused on Transcriptome analysis of petals of Camellia

nitidssima and its hybrid and identification of key DEG involved in petal color.

Following few points authors should address to further improve this manuscript.

Authors are suggested to provide top 50 up and downregulated genes based on their study with their annotation as supplemental data as this will be helpful for the researcher in this area.

• Authors are suggested to include a diagrammatic model including possible factors responsible for flower (petal) color. In the model, they may highlight what is already known and the key factors authors have identified in this study.
• Authors are also suggested to include a workflow diagram for the transcriptome study they have performed as a supplemental figure.
• The existing supplemental figures, authors have given in the manuscript are not clearly visible. Therefore need to address this issue.

Author Response

Comments and Suggestions for Authors

The manuscript by Liu et. al. has focused on Transcriptome analysis of petals of Camellia

nitidssima and its hybrid and identification of key DEG involved in petal color.

Following few points authors should address to further improve this manuscript.

Authors are suggested to provide top 50 up and downregulated genes based on their study with their annotation as supplemental data as this will be helpful for the researcher in this area.

Response: Thank you very much for your advice of this manuscript. We have added Supplementary Excel contained top up- and down-regulated genes of CN vs ZHQ, CN vs HXL, CN vs XSJ.

• Authors are suggested to include a diagrammatic model including possible factors responsible for flower (petal) color. In the model, they may highlight what is already known and the key factors authors have identified in this study.

Response: Thank you very much for your suggestion. We have added a diagrammatic model in Supplementary Figure 1 and highlighted the the factors related to this article.

• Authors are also suggested to include a workflow diagram for the transcriptome study they have performed as a supplemental figure.

Response: Thank you very much for your suggestion. We have added a workflow diagram in Supplementary Figure 2.

• The existing supplemental figures, authors have given in the manuscript are not clearly visible. Therefore need to address this issue.

Response: Thank you very much for your advice. We have revised the Supplementary Figures to make them see more clearly.

Author Response File: Author Response.docx

Reviewer 3 Report

In the current study under review, the authors made a nice attempt to examine the changes in gene expression associated with flavonol metabolism in yellow camellia. However, some major changes need to be made to increase the readability and enhance the presentation of the obtained results.  Furthermore, some queries need to be answered/clarified before the manuscript could be judged as suitable for publication. The following are some comments, suggestions, or queries regarding the manuscript.

1. Several typos and grammatical errors are there hindering the reader’s ability to understand the intended meaning. I suggest reviewing the manuscript by a native.
2. Abstract is very general. More focus on the main obtained results is needed. The potential benefit gained from this work should be showed in the abstract. Proposal of the future experiments that could be performed based on the obtained results is recommended.
3. Introduction section needs minor improvement to clearly show the aims of the study. In -other words, lines 75-79 need to be rephrased to increase readability.
4. Materials and methods
   1. More information regarding the PCR enrichment is needed. At least, the authors should provide information about the kits used for this purpose.
   2. Authors abbreviated RNA sequencing as RNA-seq; however, sometimes they use RNAseq. It is better to use the same abbreviation all over the manuscript. Indeed, RNA-seq is commonly used as an abbreviation for RNA sequencing technique.
   3. More information about the filtration rules used to obtain clean reads is needed. On what basis were the reads removed? and with which tools?
   4. Section 2.3. reads filtering and de novo assembly is not adequate and needs more information so that the experiment could be repeated.
   5. Did the authors use any method to assess the quality of the de novo assembled transcriptome and the obtained unigenes? The authors stated that the alignment ration of the clean reads against assembled transcriptome ranged between 58.98- 61.83% which is very low. I think the quality of the assembled transcriptome needs more improvement. Otherwise, the authors should explain the reasons behind such low alignment ratios.
   6. How were DEGs used for analysis of GO function and KEGG pathway enrichment? What tools were used? Were all the DEGs used together? Were up-regulated and down-regulated genes used separately?
   7. Why were DEGs randomly selected for validation using qRT-PCR? I think it would be better to validate the expression of the genes related to flavonol biosynthesis and especially those related to the biosynthesis of quercetin-3-O-glucoside, quercetin-3-O-rutinoside, quercetin-7-O-glucoside pigments. I think this part is a must and have to be done.
   8. Did the authors performed any biological or technical replicates for the genes validated using qRT-PCR?
5. Results
   1. The authors stated that they focused on inspection of the common DEGs in CN vs ZHQ, CN vs HXL and CN vs XSJ; however, in the rest of the manuscript they examined all DEGs even without separating up-regulated and down-regulated genes. A clarification is needed for these points.
   2. The authors stated that “two flavonol synthase (FLS) (c114511.graph_c1 and c132672.graph_c1), the key synthase genes for flavonol biosynthesis, were down-regulated in CN vs ZHQ, CN vs 190 HXL and CN vs XSJ”. I think down-regulation of these genes in CN as compared to other hybrids doesn’t indicate a higher synthesis of flavonoids. Figure 3A shows that these genes are up-regulated in CN as compared to other hybrids; however, the authors claim that they were down-regulated.
6. Discussion of the obtained results and the questions raised by them is lacking. Deeper discussion of the potential roles of the studied genes as supported by the obtained results would help.
   1. Figure 4 doesn’t show any FLS genes.
   2. Again, there is no Fig 4A in the manuscript and Fig 4 doesn’t show any changes in the expression of GTs examined in the current study.
   3. The obtained results showing changes in gene expression can’t be correlated with the intensity of flavonoids in the petals of the studied plants as the authors didn’t quantify the flavonoids in these petals.
7. Conclusions
   1. Conclusions regarding the roles of up-regulated FLSs, GTs and MYBs is not true and can’t be claimed based on the results obtained in the current study as the authors didn’t examine the flavonoids content in the examined plants. Moreover, the authors didn’t validate the expression of any of these genes using qRT-PCR. The results are all hypothetical and not validated, therefore no conclusions could be drawn based on them.
8. Overall, a significant improvement in the implementation of the experiment is needed.

Author Response

Comments and Suggestions for Authors

In the current study under review, the authors made a nice attempt to examine the changes in gene expression associated with flavonol metabolism in yellow camellia. However, some major changes need to be made to increase the readability and enhance the presentation of the obtained results.  Furthermore, some queries need to be answered/clarified before the manuscript could be judged as suitable for publication. The following are some comments, suggestions, or queries regarding the manuscript.

1. Several typos and grammatical errors are there hindering the reader’s ability to understand the intended meaning. I suggest reviewing the manuscript by a native.

Response: Thank you very much for your advice of this manuscript. We have asked native-speaker English editing to polish the manuscript.

1. Abstract is very general. More focus on the main obtained results is needed. The potential benefit gained from this work should be showed in the abstract. Proposal of the future experiments that could be performed based on the obtained results is recommended.

Response: Thank you very much for your suggestion. We have revised the abstract.

1. Introduction section needs minor improvement to clearly show the aims of the study. In -other words, lines 75-79 need to be rephrased to increase readability.

Response: Thank you very much for your advice. We have revised the section “Introduction”.

1. Materials and methods
1. More information regarding the PCR enrichment is needed. At least, the authors should provide information about the kits used for this purpose.

Response: Thank you very much for your advice. We have added related information in part 2.2—“A total amount of 1 µg RNA per sample was used as input material for the RNA sample preparations. Sequencing libraries were generated using NEBNext® UltraTM RNA Library Prep Kit for Illumina® (NEB, USA) following manufacturer’s recommenda-tions and index codes were added to attribute sequences to each sample.”

1. Authors abbreviated RNA sequencing as RNA-seq; however, sometimes they use RNAseq. It is better to use the same abbreviation all over the manuscript. Indeed, RNA-seq is commonly used as an abbreviation for RNA sequencing technique.

Response: Thank you very much for your advice. We have revised the writing in article.

1. More information about the filtration rules used to obtain clean reads is needed. On what basis were the reads removed? and with which tools?

Response: Thank you very much for your advice. We have revised and added related information in section “2.3 Reads filtering and de novo assembly”— “The quality score of raw bases used the Phred formula for the base identification. Of the raw reads, those low in quality, adaptor related, and with high content of unknown base reads (N) were first removed, to leave only the clean reads.”

1. Section 2.3. reads filtering and de novo assembly is not adequate and needs more information so that the experiment could be repeated.

Response: Thank you very much for your advice. We have revised and added related information in section “2.3 Reads filtering and de novo assembly”—“After filtering the reads, Trinity [23] was used to perform de novo assembly of clean reads; the k-mer library was constructed by sequencing reads according to the specified k-mer interrupt; the component was obtained by clustering and the De Bruijn map constructed; and, finally, the transcript sequences were obtained.”

1. Did the authors use any method to assess the quality of the de novo assembled transcriptome and the obtained unigenes? The authors stated that the alignment ration of the clean reads against assembled transcriptome ranged between 58.98- 61.83% which is very low. I think the quality of the assembled transcriptome needs more improvement. Otherwise, the authors should explain the reasons behind such low alignment ratios.

Response: Thank you very much. The alignment ration of the clean reads against assembled transcriptome was about 75% in transcriptome sequencing of a single Camellia variety or species (Liu et al., 2022). This article uses four different species and varieties (Camellia nitidssima and its hybrids C. ‘Zhenghuangqi’, C. ‘Huangxuanlv’, C. ‘Xinshiji’) for transcriptome sequencing. Their genetic background is more complex than that of a variety. Therefore, the proportion is lower than that of a single variety.

1. How were DEGs used for analysis of GO function and KEGG pathway enrichment? What tools were used? Were all the DEGs used together? Were up-regulated and down-regulated genes used separately?

Response: Thank you very much for your advice. We have revised and added descriptions in part “2.4. Functional annotation and differential expression analysis”—“All DEGs were annotated to the biological process, cellular component and molecular function by the GO database, with topGO software used to perform the enrichment analysis of DEGs. All DEGs were annotated to KEGG biological pathways by the KEGG database and its enrichment factor used to analyze the enrichment degree of a given pathway.”

1. Why were DEGs randomly selected for validation using qRT-PCR? I think it would be better to validate the expression of the genes related to flavonol biosynthesis and especially those related to the biosynthesis of quercetin-3-O-glucoside, quercetin-3-O-rutinoside, quercetin-7-O-glucoside pigments. I think this part is a must and have to be done.

Response: Thank you very much for your suggestion. Random selection of DEGs is to ensure the reliability of gene expression of RNA-seq, such as Yang et al. (2018). This is a good suggestion to validate the expression of the genes related to flavonol biosynthesis. This transcriptome was sequenced 2 years ago and related materials lost. And now, these 4 camellias (Camellia nitidssima and its hybrids C. ‘Zhenghuangqi’, C. ‘Huangxuanlv’, C. ‘Xinshiji’) have not yet bloomed, and the petals cannot be obtained. Therefore, we used the expression data in RNA-seq for analysis. Next, we will continue to verify the function of flavonol-related genes.

1. Did the authors performed any biological or technical replicates for the genes validated using qRT-PCR?

Response: Thank you very much. We did three biological replicates.

1. Results
1. The authors stated that they focused on inspection of the common DEGs in CN vs ZHQ, CN vs HXL and CN vs XSJ; however, in the rest of the manuscript they examined all DEGs even without separating up-regulated and down-regulated genes. A clarification is needed for these points.

Response: Thanks for your question. All DEGs were used together for GO and KEGG analysis, made by company. Based on the GO and KEGG analysis, we further analyzed the possible pathways. The results showed that many related genes of flavonol biosynthesis and ethylene pathway were highly expressed in CN, which all belonged to down-regulated DEGs.

1. The authors stated that “two flavonol synthase (FLS) (c114511.graph_c1 and c132672.graph_c1), the key synthase genes for flavonol biosynthesis, were down-regulated in CN vs ZHQ, CN vs 190 HXL and CN vs XSJ”. I think down-regulation of these genes in CN as compared to other hybrids doesn’t indicate a higher synthesis of flavonoids. Figure 3A shows that these genes are up-regulated in CN as compared to other hybrids; however, the authors claim that they were down-regulated.

Response: Thanks for your question. “two flavonol synthase (FLS) (c114511.graph_c1 and c132672.graph_c1), the key synthase genes for flavonol biosynthesis, were down-regulated in CN vs ZHQ, CN vs HXL and CN vs XSJ”, this indicates that the genes are highly expressed in CN, which is consistent with the results in Fig 3A.

Our previous study found a higher accumulation of flavonol in petals of CN than ZHQ, HXL, XSJ, which was consistent with the expression trend of flavonol related genes. However, the relationship between gene expression and flavonol content needs to be further verified.

1. Discussion of the obtained results and the questions raised by them is lacking. Deeper discussion of the potential roles of the studied genes as supported by the obtained results would help.
1. Figure 4 doesn’t show any FLS genes.

Response: Thanks for your question. I'm sorry to write Figure 4 as Figure 3. We have revised the writing. Fig. 3 has FLS genes.

1. Again, there is no Fig 4A in the manuscript and Fig 4 doesn’t show any changes in the expression of GTs examined in the current study.

Response: Thanks for your question. I'm sorry to write Figure 4 as Figure 3. We have revised the writing. Fig. 3 has GTs genes.

1. The obtained results showing changes in gene expression can’t be correlated with the intensity of flavonoids in the petals of the studied plants as the authors didn’t quantify the flavonoids in these petals.

Response: Thank you very much for your question. Our previous study found a higher accumulation of flavonol in petals of CN than ZHQ, HXL, XSJ (Li et al., 2019), which was consistent with the expression trend of flavonol related genes in this article. And the relationship between gene expression and flavonol content needs to be further verified.

1. Conclusions
1. Conclusions regarding the roles of up-regulated FLSs, GTs and MYBs is not true and can’t be claimed based on the results obtained in the current study as the authors didn’t examine the flavonoids content in the examined plants. Moreover, the authors didn’t validate the expression of any of these genes using qRT-PCR. The results are all hypothetical and not validated, therefore no conclusions could be drawn based on them.

Response: Thank you very much for your question. In our previous work, we analyzed the pigment components of CN and its hybrids ZHQ, HXL, XSJ. The results uncovered a significantly higher content of Qu3G and Qu3R (flavonol) contents in CN than its hybrids ZHQ, HXL, XSJ, while the content of Qu7G (flavonol) in CN was higher vis-à-vis HXL and XSJ (Li et al., 2019). In this article, we used RNA-seq for possible regulatory pathway analysis.

This is a good suggestion to validate the expression of the genes related to flavonol biosynthesis. Next, we will continue to verify the function of flavonol-related genes.

1. Overall, a significant improvement in the implementation of the experiment is needed.

Response: Thank you very much for your advice of this manuscript. We have revised the manuscript.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

It's good to see that the revised version of the manuscript by Liu et. al. has been significantly improved. 

But still, I could not see names of top upregulated and downregulated genes in the manuscript as asked to include based on sequencing and annotation performed in this study.

These data will be helpful for researchers working in this area. Therefore, authors are strongly recommended to include the same as supplemental data in the manuscript.

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 3 Report

1. More information about the filtration rules used to obtain clean reads is needed. On what basis were the reads removed? and with which tools?

The authors mentioned that reads with low quality and high content of N were removed. This information should be presented in numbers. For example, reads with average Q30 score below ….. and ….% of N bases were removed. This information is very critical in judging the quality of the obtained reads and consequently the quality of the assembled transcriptome and the whole results obtained in the study.

Moreover, the tools used for reads cleaning were not mentioned. Please add this information to the methods section.

2. The authors should explain the methods used to evaluate the quality of the de novo assembled transcriptome and the results of these QC measures should be presented in the supplementary material.

3. The authors mentioned that “all DEGs were used for GO and pathway enrichment analysis”. Why up-regulated and down-regulated ones were not separated. I think the data should be re-analyzed after separating them to get more insights about up- and down-regulated GO and pathways in different cultivars and their relations with the studied trait.

4. The authors should mention the number of biological and technical replicates used in qRT-PCR analysis in the methods section.

5. “Hence, in the following analysis, we focused on inspecting these common DEGs in CN vs. ZHQ, CN vs. HXL and CN vs. XSJ, to investigating the reasons for why the petals of CN’s hybrids (ZHQ, HXL, and XSJ) become a lighter yellow.” Please, delete this sentence, as it has no relation to the work done in the manuscript.

6. The authors claimed that “two flavonol synthase (FLS) (c114511.graph_c1 and c132672.graph_c1), the key synthase genes for flavonol biosynthesis, were down-regulated in CN vs ZHQ, CN vs HXL and CN vs XSJ”, this indicates that the genes are highly expressed in CN, which is consistent with the results in Fig 3A”. How could downregulation be explained as higher expression? Downregulation means lower expression against reference.

7. Still no work has been done to verify the relationship between changes in gene expression observed in the current study and the flavonol contents in the petals. I think it is better to wait until the cultivar blooms and complete the experiment.

8. As the authors agree “This is a good suggestion to validate the expression of the genes related to flavonol biosynthesis. Next, we will continue to verify the function of flavonol-related genes.” I think validation of genes related to flavonol biosynthesis using qRT-PCR is a must before publication.

Author Response

Please see the attachment