Identification of a Novel QTL for Chlorate Resistance in Rice (Oryza sativa L.)

Reviewer 1

1.       In materials and methods (L86-98), the details of number is not described. The reader are therefore not followed the traits evaluation.

Page 2, lines 92-93: we included the following statement “height (six seedlings per treatment per rice line)”.

Page 2, line 94…”5mL of” 0.1% potassium chlorate…was added.

2.       In materials and methods (L95), “The chlorate sensitivity” may be replaced with “The chlorate resistance”.

Page 2, line 99: “chlorate sensitivity” was replaced by “chlorate resistance”.

3.       In materials and methods (L95-98), the formulas may be confused. In addition measured organs should be only survived seedlings.

Authors appreciate the concern raised by worthy reviewer. Therefore, we presented the formula as follows on page 2, lines 100-101: [(number of survived seedlings in KClO₃) – total number of tested seedlings)/Total number of tested seedlings] × 100.

The formula used to calculate the inhibition percentage of shoot is presented as follows on page 3, lines 104-105: [(average shoot length in water)–average shoot length in KClO₃)/average shoot length in water] × 100. The same formula was used for root inhibition.

Page 3, line 105-106, the following statement was included : “Six replications were used to evaluate the phenotypic response of the DH population to potassium chlorate.”

4.       In materials and methods (L106-11), the details of R package is not described. The reader are therefore not followed the drawing plots.

Authors appreciate the concern raised by thereviewer. We have included the following manuscript, on page 3, lines 118-127:

“Prior to performing QTL analysis, phenotype and genotype raw data were formatted using fread, pheno.raw, geno.raw and cbind functions in RStudio environment v. 1.2.5042 © 2009-2020 RStudio, Inc. Major steps in formatting raw data included checking of compatibility of the dataset (Structure: columns and rows names, and file extension: comma-separated value, .csv), the conversion of genotype sequence data to binary file compatible with IciMapping software. The binning of redundant markers was done, and the linkage map was constructed for a bi-parental population with IciMapping software v.4.1.0.0, using position mapping, and Kosambl mapping functions [1]. A threshold of logarithm of the odds (LOD) of 3.0 (alpha =0.05) explaining the probability for detecting statistically significant QTL associated with the chlorate resistance, by permutation test [2].”

We included a new subsection in materials and methods referred to as 2.4 (page 3, lines 128-148), describing the genomic DNA extraction and genotyping for nitrate reductase and nitrate transporter genes using InDel (Insertion/Deletion) markers.

On page 4, lines 150-157, a new subsection, 2.5 describing the in silico analysis, was included.

5.       In results (3.1. Differential phenotypic …), Fig. 1 and 2 are unnecessary. “Fig. 1 and 2” may be replaced with “Fig. S1”.

Authors have moved previously named Figure 1 and Figure 2 to the supplementary materials, now referred to as Figure S1 and Figure S2. The clustering panels in old Figures 1A and 2A were deleted. Old figure S1 has been moved to the main text, and is now referred to as Figure 1 (page 5, lines 191-202)

6.       In results (L156), “kindship” may be replaced with “kinship”.

Page 7, line 214, kindship was corrected and replaced by “kinship”

7.       In results (161-162), the sentences (This could imply …) had better place in discussion.

The following sentence was removed from the results section: “This could imply that degree of chlorate resistance is independent to the shoot inhibition or root inhibition percentage, and vis-versa”, and placed on page 11, lines 290-295 (Discussion section) and rephrased as follows: “This would imply that the degree of chlorate resistance of the studied DH lines is independent to the shoot inhibition in response to KClO3. Under the same conditions, a strong negative correlation was observed between root inhibition and chlorate resistance. Therefore, it may be suggested that high degree of chlorate resistance could be explained in part by the low reduction in root inhibition under KClO3 treatment”

8.       In results (Fig. 4), some DH lines are very similar genome (Fig. 4A). The genotypic bias from those DH lines should be confirmed your statistical analysis.

In graph legend of Fig. 4C, “Tolerant” may be replaced with “Resistant”.

We appreciate the pertinence of the concern raised by the reviewer. Therefore, we would like to specify that the japonica cv. Milyang352 (P2 in our study) was developed from a cross between C18/Ungwang. C18 is a cultivar originated from China, which is also the same origin of the cv. 93-11 used as P1. It could be then thought that C18 and 93-11 would share some genomic regions, with regard to regard to the similarity in the leaf color character (data not shown). This could possibly explain the suspected genomic similarity in some of the DH lines.

We have replaced “Tolerant” by “Resistant” in Figure 4C caption (page 8, line 230)

9.       In results (3.3. Novel QTL …), if it is difficult to explain the threshold of QTL analysis, please consider dropping the QTL analysis and apply the manhattan plot by qqman of R package.

Authors appreciate the suggestion made by worthy reviewer. However, the results of the QTL analysis is found to be clear enough to explain. A brief description of threshold LOD has been included on page 3, lines 125-127.

10.    In results (L189), “Figure S2” may be replaced with “Figure S1”. In the first place Table 2 is unnecessary, because three is Fig. S1.

Authors have no objection to the suggestion made by the reviewer. We removed old Table 2 from the main text and moved to the supplementary materials as Table S2.

11.    In results (L193-196), the sentences (Therefore, …) had better place in materials and methods.

The text in lines 193-196 has been moved to materials and methods section, 2.5. on page 4, lines 150-157, and has been rephrased as follows: “2.5. Comparative coding sequences (CDS) alignment of candidate genes between indica ssp. and japonica ssp.

We were also interested in knowing the similarity between homologs genes in indica and japonica subspecies in order to predict the similarity of the function the protein they code for. Therefore, we performed the alignment of coding sequences (CDS) of each gene, downloaded from Nipponbare genome database (http://rice.plantbiology.msu.edu) for japonica, and indica rice genome database (http://plants.ensembl.org/Oryza_indica), using ClustalW Multiple Alignment in BioEdit Sequence Alignment Editor [3].”

12.    “5. Discussion” may be replaced with “5. Conclusion”.

Authors apologize for the duplication of discussion title. We have corrected and changed to Conclusion (line 333)

Reviewer 2

In this manuscript, the authors identified a novel QTL for chlorate resistance using DH lines. However, the present manuscript has some problems to need improvement.

Major Comments:

1.       I think it’s better to reverse the order of describing the phenotypes DH lines and parental lines in the result.

Authors appreciate the suggestion made by the reviewer, and found it to be useful. We have reorganized the results section by describing first the DH lines phenotypic response followed by parental lines (page 4, lines 159-189. Figures have been changed following the suggestions. We hope that our understanding of the reviewer suggestion met the essence of his mind.

2.       The statistical analysis method is not appropriate. I don’t understand the need for cluster analysis, and I think the number of clusters is not correct because the clusters overlap each other (Fig. 1A, 2A).

Moreover, it describes phenotype correlations, but I think heat map are not enough and correlation analysis is needed.

P.3 L.135, the reason for setting the phenotype threshold is not clear.

We appreciate the concern raised by the reviewer, and agree with the comment. We have removed the cluster data from the main text, and moved Figures 1 and 2 to supplementary materials as Figures S1 and S2 (without old panels A and B).

We also included the results of the correlation analysis investigating the relationship between shoot and root inhibition, and the chlorate resistance of DH lines. New figure was included (Figure 2 + caption) on page 6, lines 203-206.

The chlorate resistance was calculated as the percentage of survived seedlings under potassium chlorate treatment. The correlation between the chlorate resistance and other traits was estimated (Table S3). We determined a scoring degree 0-50% of dead seedlings for sensitive group, and above 60% of survival (seedlings with strong or light green leaves, with not necrotic symptoms) were counted as resistant lines. We have included in materials and methods section, page 3, lines 101-103 the following statement : “Rice DH lines with a percentage of survival above or equal to 60% rate were scored resistant to potassium chlorate, while those with a survival percentage below or equal to 50% were scored sensitive.”

3.       Not enough information on QTL analysis. Which method was used for QTL analysis?

Authors would like apologize for the inconvenience. We have included description regarding QTL analysis on page 3, lines 118-127.

4.       QTL analysis was performed on seven traits, but only one QTL for chlorate resistance was detected. What do you think about the result that QTLs were not detected in other traits?

Authors appreciate the interest showed by the reviewer to improve our manuscript. We would like to specify that QTL analysis was performed for seven traits but only the QTL for chlorate resistance was detected. We would like to indicate here that other studies considered LOD scores lower than lower 2.5 [1] and 3 [2] for the detection of putative QTL for chlorate resistance in rice. Using this threshold, more QTL were detect in our study. However, for reducing false positive effect, we selected higher threshold LOD as of 3.0 following previous guidelines [3].

5.       I think the supplementary figures should be the main figure. However, the supplementary figures have insufficient figure legend.

Authors have no objection to the suggestion made by the reviewer. We have moved the Figure S1 from the supplementary materials the main text as Figure 1, and the figure legend was corrected (page 5, lines 191-202).

6.       Have you confirmed the genotype of the chlorate resistance gene reported so far in parental lines?

The parental lines don’t seem to be a special combination, but why do you think a new locus was detected? I think this is an important point of this manuscript.

Authors have performed the genotyping of nitrate reductase (NR) and nitrate transporter (NRT) gene using InDel markers. Data included in as Figure 5 in the manuscript show that the NR and NRT InDel markers amplified polymorphic bands between parental lines. Description of the results was included in subsection 3.4, page 10, lines 259-279.

We appreciate the pertinence of the concern raised by the reviewer. Therefore, we would like to specify that the japonica cv. Milyang352 (P2 in our study) was developed from a cross between C18/Ungwang. C18 is a cultivar originated from China, which is also the same origin of the cv. 93-11 used as P1. It could be then thought that C18 and 93-11 would share some genomic regions, with regard to regard to the similarity in the leaf color character (data not shown). This could possibly explain the suspected genomic similarity in some of the DH lines.

Besides, factors involved in the detection of a putative QTL associated with a particular trait of interest include the genetic background of parental lines and the mapping population.

The indica allele of nitrate reductase present in cv. 93-11, which allows to improve the nitrogen use efficiency in rice could not be detected in the panel of japonica rice cultivars.

The clear distinctive phenotypic responses observed between parental lines towards potassium chlorate treatment, indicates their genetic differences. In addition, previously detected QTL associated with chlorate resistance in rice have shown that qCHR-2, qCHR-8, and qCHR-10 explained about 26.5, 12.6, and 12.6% of the observed phenotypic variation (PVE), respectively. In the present study, the detected QTL (qCHR-3) using DH population counted for 14.9% of the phenotypic variance.

7.       Not enough phenotypes considerations. Regarding the description of the relationship between shoot growth pattern and chlorate sensitivity, the result and discussion do not seem to match.

Authors are thankful to reviewer for the interest and suggestions to improve the content of the manuscript. We have modified the discussion section and matched with the recorded results (page 11, lines 280-301)

8.       In the introduction, I think that there is insufficient description about the genetic knowledge of chlorate resistance. It would be better to introduce the papers cited in the discussion.

Authors have included changes in the introduction as suggested (page 1, lines 38-40; 43-45 p.2).

9.       Often times the version of RStudio is shown, but what should be shown is the statistical method and version of the package.

We agree with the reviewer, and necessary changes have been included in the manuscript regarding R packages and functions (page 3, lines 110-111, 119-120; Figures S1, S2 captions)

Minor Comments:

1.       Figure 1B and 2B do not have vertical and horizontal labels.

We added the labels in the y- and x-axis of previously named Figures 1B and 2B, now moved to supplementary materials as Figure S1A, and Figure S2A

2.       The vertical axis label in Figure 3C is incorrect.

Authors apologize for the labelling mistake. The label of y-axis has been corrected to “Root length (cm)” in Figure 3B, page 7, line 207.

3.       There is no citation in the text of Figure 2B.

Authors apologize for the inconvenience. Figure 2 was changed to supplementary Figure S2, panel B became panel A.

4.       The table 2 is out of alignment.

We are thankful for the scrutiny in the revision. Old Table 2 has been moved to supplementary materials as Table S2

5.       P.10 L.255, the heading is wrong.

Authors would like to apologize for the duplication of title. On page 12, line 333, “Discussion” was changed to section 5.“Conclusion.”

1.       In materials and methods (L100-105), the formulas are still confused. My understanding is that the shoot inhibition (%), SI = [(SLC - SL_KCLO3) / SLC] * 100.

We are thankful to the reviewer for the comments, which improved significantly the content and the quality of the manuscript. We have improved the presentation of the formulas as follows:

Page 3, lines 100-101, “the [(total number of tested seedlings – number of dead seedlings in KClO₃)/total number of tested seedlings] × 100.”

Page 3, lines 104-105, “[(SLC – SL_KClO₃)/SLC] × 100 (SLC: shoot length under control; SL_KClO₃: shoot length under KClO₃).”

2.       In materials and methods (L118-123), I think the sentences (Prior to performing … The binning of redundant markers was done, and) are unnecessary.

Authors agree with the suggestion made by the reviewer; therefore, changes were included in lines 118-120 (page 3) as follows: “An initial formatting of phenotype and genotype raw data was done using fread, pheno.raw, geno.raw and cbind functions in RStudio environment v. 1.2.5042 © 2009-2020 RStudio, Inc.”.

We also deleted the following from the manuscript: “Major steps in formatting raw data included checking of compatibility of the dataset (Structure: columns and rows names, and file extension: comma-separated value, .csv), the conversion of genotype sequence data to binary file compatible with IciMapping software. The binning of redundant markers was done, and”.

We kept few detail to allow other readers to follow up the flow of the analysis, as suggested by anonymous reviewer (page 3, lines 120-123)

3.       In materials and methods (L128-149), I think the analysis (Genomic DNA Extraction and Genotyping for nitrate reductase and nitrate transporter genes) is unnecessary in this paper

Authors appreciate the by worthy reviewer regarding the genotyping for nitrate reductase and nitrate transporter genes. However, we would like to indicate the section “.4. Genomic DNA extraction and Genotyping” was not present in the initial version of the manuscript, because we did not find the opportunity of including this data in the present manuscript. But, one of the anonymous reviewers formulated a specific recommendation regarding the confirmation of the presence of different alleles of nitrate reductase (NR) and nitrate transporter (NRT) in the parental lines. Upon this recommendation, we, therefore, performed the genotyping for NR and NRT using InDel markers.

The title of subsection 2.4. (page 3, line 124) has been modified as follows: ”Genomic DNA Extraction and Genotyping”.

Page 3, line 137, we included the following: “The Genotyping of parental lines for nitrate reductase (NR) and nitrate transporter (NRT) genes…”

4.       In graph legend of Fig. 4C, “Tolerant” may be replaced with “Resistant”. The figure legend is still not correct.

Page 8, line 226, “Tolerant” was already replaced with “Resistant” (Caption of Figure 4).

Authors apologize for the inconvenience. The legend of Figure 4 was corrected, “Tolerant” was replaced with “Resistant” (page 8, line 220)