Genetic Dissection of Phosphorous Uptake and Utilization Efficiency Traits Using GWAS in Mungbean

Round 1

Reviewer 1 Report

This research about genetic dissection of Phosphorous uptake and utilization  in Mungbean is very interesting. The manuscript is well- written. I have only few comments that are indicated in the attached document.

Comments for author File: Comments.pdf

Author Response

We would like to thank you very much for your critical comments suggestions. The necessary corrections have been incorporated.

Comments : (Vigna radiata)...Italics

The necessary changes have been done.

Key words separated from Abstract

Line 60: PUpE and PUtE...

Abbreviation described according to suggestions.

Line 85: GBS....

Abbreviation described according to suggestions.

Line 160: Necessary correction have been incorporated

Author Response File: Author Response.pdf

Reviewer 2 Report

The work is extremely interesting and is part of the coordinates of sustainable agriculture, efficient use of resources. The work would be even richer, more complete, if we take into account the suggestions made to the authors. However, it is not mandatory. These may be the subject of future research.

1. Usually, the scientific papers have 5-7 keywords. Here are 10.
2. It would be interesting to mention in the paper if it was observed in the 120 mung-bean genotypes any connection between certain phenotypic characteristics and the presence of candidate genes associated with P-uptake and P-utilization efficiency traits. In this way, the paper would take a step towards the applicative side of agronomy. No matter how high the scientific level of research, we must also take into account the practical finality of the results.
3. In the future research it is important that the experiments are continued in soil. There can be important differences in the efficiency of phosphorus absorption between hydroponic culture and soil.

Author Response

The work is extremely interesting and is part of the coordinates of sustainable agriculture, efficient use of resources. The work would be even richer, more complete if we take into account the suggestions made to the author. However, it is not mandatory. These may be the subject of future research.

We would like to thank you very much for your constructive and encouraging comments as well as suggestions.

1. Usually, the scientific papers have 5-7 keywords. Here are 10.

 

We thought the maximum limit of keywords was 10. Therefore, we kept it as 10 in this scientific paper.

 

1. It would be interesting to mention in the paper if it was observed in the 120 mungbean genotypes any connection between certain phenotypic characteristics and the presence of candidate genes associated with P-uptake and P-utilization efficiency traits. In this way, the paper would take a step towards the applicative side of agronomy. No matter how high the scientific level of research, we must also take into account the practical finality of the results.

 

We would like to thank you for the suggestion. We also agree with you on the note ‘No matter how high the scientific level of research, we must also take into account the practical finality of the results’. In this study, we evaluated four traits viz. total dry weight, P concentration, total P uptake, and P utilization efficiency. These traits are directly associated with root and shoot. Therefore, we have carried out digital gene expression analysis of all the putative candidate genes using different developmental stages of the shoot, root as well as shoot apical meristem. The results indicated that six genes exhibited higher expression in root and shoot tissues. Therefore, we concluded that these genes may be the potential candidate genes as they may change the root and shoot architecture in the mungbean. Please follow the line 329-334.  The Vradi07g24790 and Vradi03g02620 were found to be associated with the TDW traits. The Vradi06g14930, Vradi08g20910, Vradi08g11310, Vradi09g09030, Vradi05g06040 and Vradi06g03940 were found to be associated with PC trait. The Vradi07g06240 Vradi08g00070, Vradi05g20860 and Vradi06g12490 were associated with TPU trait.  The Vradi06g13450, Vradi08g10870, Vradi01g04370 and Vradi05g21880 were linked with PUtE trait.

 

1. In the future research it is important that the experiments are continued in soil. There can be important differences in the efficiency of phosphorus absorption between hydroponic culture and soil.

 

We would like to thank you for your valuable suggestion. We are planning to work in the direction of studying the effect of phosphorous on the different genotypes at field conditions.

Author Response File: Author Response.pdf

Reviewer 3 Report

This manuscript investigates phosphorus uptake efficiency and phosphorus utilization efficiency in mungbean, an economically and culturally important legume crop in many countries.  This work builds on previous work by this lab to identify >55K SNPs in mungbean.  The SNPs are used to classify and characterized the population structure of the 120 mungbean genotypes chosen for the study.  Further, the SNPs are coupled with phenotypic measurements in 120 genotypes to identify 161 SNPs important to PUpE or PUtE. This identifies 61 genes with SNPs, six of them (?) exhibit high digital expression in their Arabidopsis homologs.  While I believe this study is an excellent start, it does require a substantial amount of work before it would be ready for publication. 

The only supplemental data provided are graphs of expression values in different tissues at different stages for 12 different Arabidopsis genes.  There are no figures or tables to better understand the text of the manuscript.  I can see that there should be based on the supplementary material descriptions, but they are not referenced in the manuscript, not embedded in the text of the manuscript, nor are they even included in the downloaded supplementary information (probably because they're not referenced within the body of the manuscript).  This is critical information to include with the manuscript and without it it's impossible to fully understand the study. 

 

Major Changes:

A GWAS study will identify small(ish) regions of the genome associated with the trait of interest (here PUpE and PUtE).  Each of these regions includes significant SNPs and is bound by insignificant SNPs.  All genes between the two bounding insignificant SNPs are potential genes of interest, not just the genes containing an identified SNP.  While most GWAS studies are done using a SNP chip, since that is not available for Mungbean, GBS is an appropriate alternative.  However, it is not clear to me whether all genes within a region containing a significant SNP, but bound by insignificant SNPs were considered, or only genes containing a significant SNP within the coding region of a gene were considered candidate genes.  Please clarify.  If only genes containing a significant SNP within the coding region are considered, please expand the number of potential candidate genes. 

Are the QTL identified in this study co-linear with nutrient use or abiotic stress QTL in other species? That would add a level of validation to the mungbean findings.

Are any of the SNPs identified in this study also identified in previous studies referenced in lines 91-23 “SNPs have been associated with PUE regulating traits in different grain legume crops like soybean [32], cowpea 92 [33], and mungbean [20].”?  This would again add validity to this study.

A large portion of work for this study went towards understanding the relationship and population structure of the genotypes used in the study.  Do any of the SNPs identified correlate with the population structure?  For example, are there any SNPs for PUpE or PUtE that are unique to any of the five sub-populations? 

The study identified 116 SNPs associated with PUpE and PUtE within 61 genes.  16 of these 61 genes are known to increase PUpE and/or PUtE.  Were there other SNPs NOT within a gene coding region?

The paper highlights six genes in the abstract, but only discusses two of the six genes in the paper itself.  What about the other four genes?  Further, is there anything interesting about the 61 genes?  Any over-represented gene ontology (GO) terms associated with the genes?  Are they all the same type of gene? Except for ’they are highly expressed somewhere in the plant’ what else is interesting about these genes?  In the discussion we learn 16  (including two that are mentioned in the abstract) have functions related to PUE.  What about the other 45 genes?

 Phosphorus is taken up by the roots and then utilized throughout the plant.  While I don’t dispute gene activity in other parts of the plant might be important in Phosphorus utilization efficiency, I think it’s important to propose how the gene is involved in phosphorus utilization efficiency.  Enhancing root growth is fairly obvious, more root surface area would allow more phosphorus uptake, but pollen tube is less obvious to me. 

Paragraph starting on ln 305:  If these genes don’t lie within a PUpE or PUtE QTL OR the expression of the genes is not altered by altered Pi availability, it is pure conjecture that these genes are involved in regulating PUpE or PUtE. IF these six genes were identified using “Expression Angler”, then their expression is correlated with the expression of the 61 genes, identified by the study, but they might not be part of the 61 genes.  Please clarify. IF these six are identified by Expression Angler, are binding sites for these genes present in the promoter region of genes identified by GWAS study?  

The manuscript talks a lot about "genomics assisted breeding".  In order for the findings of this manuscript to actually be utilized in genomics assisted breeding, markers (likely developed from identified SNPs) need to be developed to screen for 'favorable' or against 'unfavorable' alleles of the candidate genes.  

 

Minor Changes:

Abstract:  What is digital gene expression?  Additionally, depicted is the wrong word.  Maybe ‘identified’ six genes with high expression.  Missing a ] at the end of the digital gene expression sentence.  Don’t include the AtIDs in the Abstract. 

Various plant tissues – very vague. 

Ln. 34:  The way this sentence is phrased requires a year be included; “Globally, this crop was grown in 7.3 million hectares with a production of 5.3 million tonnes in 2020”  or “Globally, in 2020, ….”

Ln 57-61:  These sentences are all important statements, but they all stand alone and don’t flow together for a paragraph.

Ln 65:  Please re-define PUpE nad PUtE here since it’s the fist application of the acronyms.

Ln 61-63:  The sentence “The breeding of cultivars with improved…” should either have the acronyms re-defined as it’s their first application, or should be moved to after the sentence “PUE can be differentiated into PUpE and PUtE.”

Ln 70:  Genomics assisted breeding isn’t involved in dissecting complex quantitative traits.  It IS used in application of findings to improve crop varieties in a targeted and efficient manner.

Ln 72:  Again, association studies, genetic mapping and expression profiling aren’t components of genomics assisted breeding.  The results of these types of studies can be used to generate markers, which can then be used in genomics assisted breeding. 

Ln 86:  GBS has been around for at least a decade, it’s not particularly novel, though this study may be it’s first application in mungbean.

Ln 149-157: Digital gene expression and validation.  This is a good approach to narrow candidate genes and back up your findings, but this is not validating your candidate genes are involved in PUpE or PUtE.  Arabidopsis has not been domesticated and has not undergone genome duplication events experienced by legumes.  Further, because of domestication and genome duplication, legumes have developed specialized networks and genes associated with phosphorus uptake and utilization.  These may or may not be reflected in Arabidopsis expression profiles.  To really validate your candidate genes, you need to do qRT-PCR in mungbean and compare NP and LP.  The tissues and timing of sample collection may be informed by your findings in Arabidopsis, but relying solely on Arabidopsis studies, can not serve as validation, only support. Sanger sequencing validated the GBS results (the SNP identification), which is appropriate and excellent. 

Does silencing any of these genes in Arabidopsis alter the PUpE or PUtE?  Given the lack of genomic tools for mungbean, showing altered PUpE or PUtE due to gene silencing would be a good indication that theses genes are associated with these traits. 

Ln 156: Genes in a list should be presented in ascending chromosomal order: VRADI01G, VRADI05G, VRADI08G. Please fix. Throughout the manuscript.

Ln 200:  What does the ‘No’ mean?  If it stands for ‘number’, it can be deleted.

Ln 200-201:  Are the 116 SNPs common to MLM and CMLM part of the 125 and 123 SNPs described in the previous sentence?  Which would mean only 9 SNPs unqiue to MLM and 6 SNPs unique to CMLM? 

Ln 218-224:  Identifying the Arabidopsis homolog of the mungbean candidate genes is actually information that should be included in the Materials and methods.  Further, additional BLAST information is needed.  Was the BLAST a nucleotide, peptide, or translated BLAST?   P-value <1?  That is incredibly permissive.  Similarlity Index is not a BLAST parameter – please clarify.

Ln 234: PUP1 ad PSTOL need to be written out with abbreviations in parentheses

Ln 237: ANXUR needs to be written out with abbreviation in parentheses.

Ln 244-245:  How can the total phosphorus uptake of the genotypes be higher under LP if the mean value of the total phosphorus uptake is lower under LP compared to NP?

Ln 270:  Given how inter-connected TDW, PC, TPU, and PUtE are, it would be unusual if all four traits weren’t identified. 

Ln 295: Which gene is ‘this gene’ refering to?  VRADI06G12490, PHR, or PSI?

Ln 307: VRADI08G00070 is listed twice, but VRADI01G04370 isn’t listed at all.

 

Author Response

This manuscript investigates phosphorus uptake efficiency and phosphorus utilization efficiency in mungbean, an economically and culturally important legume crop in many countries.  This work builds on previous work by this lab to identify >55K SNPs in mungbean.  The SNPs are used to classify and characterized the population structure of the 120 mungbean genotypes chosen for the study.  Further, the SNPs are coupled with phenotypic measurements in 120 genotypes to identify 161 SNPs important to PUpE or PUtE. This identifies 61 genes with SNPs; six of them (?) exhibit high digital expression in their Arabidopsis homologs.  While I believe this study is an excellent start, it does require a substantial amount of work before it would be ready for publication. 

We would like thank you for critical comments and suggestions.

The only supplemental data provided are graphs of expression values in different tissues at different stages for 12 different Arabidopsis genes.  There are no figures or tables to better understand the text of the manuscript.  I can see that there should be based on the supplementary material descriptions, but they are not referenced in the manuscript, not embedded in the text of the manuscript, nor are they even included in the downloaded supplementary information (probably because they're not referenced within the body of the manuscript).  This is critical information to include with the manuscript and without it it's impossible to fully understand the study. 

We would like to thank you very much for raising the concern. We have referenced the supplementary fig. in the revised manuscript. Supplementary tables, main text tables, and figures are already cited in the main text.

 

Major Changes:

A GWAS study will identify small(ish) regions of the genome associated with the trait of interest (here PUpE and PUtE).  Each of these regions includes significant SNPs and is bound by insignificant SNPs.  All genes between the two bounding insignificant SNPs are potential genes of interest, not just the genes containing an identified SNP.  While most GWAS studies are done using a SNP chip, since that is not available for Mungbean, GBS is an appropriate alternative.  However, it is not clear to me whether all genes within a region containing a significant SNP, but bound by insignificant SNPs were considered, or only genes containing a significant SNP within the coding region of a gene were considered candidate genes.  Please clarify.  If only genes containing a significant SNP within the coding region are considered, please expand the number of potential candidate genes. 

In this study, we have considered the SNPs present within 6 kb upstream, 4 kb downstream, coding region and intron region of the gene. Therefore, we consider significant as well as insignificant SNPs flanked the gene.

Are the QTL identified in this study co-linear with nutrient use or abiotic stress QTL in other species? That would add a level of validation to the mungbean findings.

We have carried out the Meta analysis of the QTL identified in this study. However, we have not found any co-linearity of these genomic regions with other species and mungbean also. The genetic inheritance of PUpE and PUtE traits has been explored very less in mungbean. In fact, to the best of our knowledge, this effort remains one of the earliest studies to decode the genetic complexity of PUpE and PUtE traits in mungbean. Therefore, the limited exploration of these traits in mungbean could be the possible reason of not having co-linearity of these genomic regions with other species and mungbean.

Are any of the SNPs identified in this study also identified in previous studies referenced in lines 91-23 “SNPs have been associated with PUE regulating traits in different grain legume crops like soybean [32], cowpea 92 [33], and mungbean [20].”?  This would again add validity to this study.

In this study, all the SNPs associated with candidate genes are novel. In our previous study, the reported SNPs are different from the SNPs identified in this study.

A large portion of work for this study went towards understanding the relationship and population structure of the genotypes used in the study.  Do any of the SNPs identified correlate with the population structure?  For example, are there any SNPs for PUpE or PUtE that are unique to any of the five sub-populations? 

We have not identified any unique SNP belonging to a particular sub-population.

The study identified 116 SNPs associated with PUpE and PUtE within 61 genes.  16 of these 61 genes are known to increase PUpE and/or PUtE.  Were there other SNPs NOT within a gene coding region?

The candidate SNPs were found to be present at regulatory, coding and intronic region of the gene.

The paper highlights six genes in the abstract, but only discusses two of the six genes in the paper itself.  What about the other four genes?  Further, is there anything interesting about the 61 genes?  Any over-represented gene ontology (GO) terms associated with the genes?  Are they all the same type of gene? Except for ’they are highly expressed somewhere in the plant’ what else is interesting about these genes?  In the discussion we learn 16 (including two that are mentioned in the abstract) have functions related to PUE.  What about the other 45 genes?

In results section, the details are provided about the other four genes (line 250-260) and are also provided in Table 3. The detail about all the SNPs and 61 genes, including 45 remaining gene provided in a separate supplementary table Table S2). Most of these genes belong to transporter class and also having zinc finger, bHLH, metallophos and pentatricopeptide repeat domain. They are also reported to be involved in heavy metal stress and abiotic stress.

 Phosphorus is taken up by the roots and then utilized throughout the plant.  While I don’t dispute gene activity in other parts of the plant might be important in Phosphorus utilization efficiency, I think it’s important to propose how the gene is involved in phosphorus utilization efficiency.  Enhancing root growth is fairly obvious, more root surface area would allow more phosphorus uptake, but pollen tube is less obvious to me.

In this study, we dissected phosphorus utilization and uptake efficiency traits with a molecular breeding perspective. Therefore, we have not focused on the physiological perspective of phosphorus utilization and uptake efficiency in mungbean. The details about the regulation of phosphorus utilization and uptake through interacting gene networks have been discussed extensively in earlier studies (Heuer et al. 2016; Dissanayaka et al. 2018).

Paragraph starting on ln 305:  If these genes don’t lie within a PUpE or PUtE QTL OR the expression of the genes is not altered by altered Pi availability, it is pure conjecture that these genes are involved in regulating PUpE or PUtE. IF these six genes were identified using “Expression Angler”, then their expression is correlated with the expression of the 61 genes, identified by the study, but they might not be part of the 61 genes.  Please clarify. IF these six are identified by Expression Angler, are binding sites for these genes present in the promoter region of genes identified by GWAS study?  

We have carried out the expression analysis of 16 genes which were found to have a crucial role in the imposition of phosphorous stress tolerance in the plants. These genes are considered as putative candidate genes. Among these genes, six genes showed significantly higher expression in the root, shoot, shoot apical meristem, and leaf tissues. Therefore, we concluded these genes as potential candidate genes. The VRADI05G20860 gene contains SNP in the upstream regulatory region. We carried out promoter analysis of this gene. However, we have not found any transcription factor binding site harboring the SNP.

The manuscript talks a lot about "genomics assisted breeding".  In order for the findings of this manuscript to actually be utilized in genomics assisted breeding, markers (likely developed from identified SNPs) need to be developed to screen for 'favourable' or against 'unfavourable' alleles of the candidate genes. 

We would like to thank you for your suggestion. We are trying to convert these SNPs into PCR-based marker system. This will lead to develop breeder friendly marker system which can be easily used in lab.

Minor Changes:

Abstract:  What is digital gene expression?  Additionally, depicted is the wrong word.  Maybe ‘identified’ six genes with high expression.  Missing a ] at the end of the digital gene expression sentence.  Don’t include the AtIDs in the Abstract. 

Necessary corrections have been made.

Various plant tissues – very vague. 

Necessary corrections have been made.

Ln. 34:  The way this sentence is phrased requires a year be included; “Globally, this crop was grown in 7.3 million hectares with a production of 5.3 million tonnes in 2020”  or “Globally, in 2020, ….”

Necessary corrections have been made.

Ln 57-61:  These sentences are all important statements, but they all stand alone and don’t flow together for a paragraph.

Necessary corrections have been made.

Ln 65:  Please re-define PUpE nad PUtE here since it’s the fist application of the acronyms.

Necessary corrections have been made.

Ln 61-63:  The sentence “The breeding of cultivars with improved…” should either have the acronyms re-defined as it’s their first application, or should be moved to after the sentence “PUE can be differentiated into PUpE and PUtE.”

Necessary corrections have been made.

Ln 70:  Genomics assisted breeding isn’t involved in dissecting complex quantitative traits.  It IS used in application of findings to improve crop varieties in a targeted and efficient manner.

Necessary corrections have been made.

Ln 72:  Again, association studies, genetic mapping and expression profiling aren’t components of genomics assisted breeding.  The results of these types of studies can be used to generate markers, which can then be used in genomics assisted breeding. 

Necessary corrections have been made.

Ln 86:  GBS has been around for at least a decade, it’s not particularly novel, though this study may be its first application in mungbean.

Necessary corrections have been made.

Ln 149-157: Digital gene expression and validation.  This is a good approach to narrow candidate genes and back up your findings, but this is not validating your candidate genes are involved in PUpE or PUtE.  Arabidopsis has not been domesticated and has not undergone genome duplication events experienced by legumes.  Further, because of domestication and genome duplication, legumes have developed specialized networks and genes associated with phosphorus uptake and utilization.  These may or may not be reflected in Arabidopsis expression profiles.  To really validate your candidate genes, you need to do qRT-PCR in mungbean and compare NP and LP.  The tissues and timing of sample collection may be informed by your findings in Arabidopsis, but relying solely on Arabidopsis studies, can not serve as validation, only support. Sanger sequencing validated the GBS results (the SNP identification), which is appropriate and excellent. 

Though the qRT-PCR is a way to check the expression of candidate genes; however, digital gene expression analysis has also been used in a number of studies to determine the differential gene expression pattern of candidate genes (Li et al. 2010; Singh et al. 2017; Singh et al. 2015).Thus, in our study, we have performed digital gene expression analysis in order to check the differential gene expression pattern (Ryan et al. 2016) using several tissues, including root, shoot etc. We choose Arabidopsis, as it is considered a model plant. Moreover, there is no availability of digital gene expression platform for legumes. In addition, we have also validated the candidate genes using the Sanger sequencing technique in a selected set of mungbean genotypes. These studies have unequivocally validated the role candidate genes identified in this study. Necessary details have been suitably added in the revised manuscript to highlight these points.

Li, L., Li, H., Li, J. ShuTu Xu, XiaoHong Yang, Sheng Lijan et al. (2010) A genome-wide survey of maize lipid-related genes: candidate genes mining, digital gene expression profiling and co-location with QTL for maize kernel oil. (2010) Sci. China Life Sci. 53 690–700 https://doi.org/10.1007/s11427-010-4007-3

Singh, V. K., & Jain, M. (2015). Genome-wide survey and comprehensive expression profiling of Aux/IAA gene family in chickpea and soybean. Frontiers in plant science, 6, 918. https://doi.org/10.3389/fpls.2015.00918

Singh, V.K., Rajkumar, M.S., Garg, R, Jain M Genome-wide identification and co-expression network analysis provide insights into the roles of auxin response factor gene family in chickpea. Sci Rep 7, 10895 (2017). https://doi.org/10.1038/s41598-017-11327-5

Does silencing any of these genes in Arabidopsis alter the PUpE or PUtE?  Given the lack of genomic tools for mungbean, showing altered PUpE or PUtE due to gene silencing would be a good indication that these genes are associated with these traits.

All the Arabidopsis orthologue of identified candidate genes have been well characterized. These genes are reported to be involved in PUpE or PUtE related pathway. The functions of these genes have been validated with overexpression or knock-out lines. The necessary references have been cited in the table of main text.   

Ln 156: Genes in a list should be presented in ascending chromosomal order: VRADI01G, VRADI05G and VRADI08G. Please fix. Throughout the manuscript.

Necessary corrections have been made.

Ln 200:  What does the ‘No’ mean?  If it stands for ‘number’, it can be deleted.

Necessary corrections have been made.

Ln 200-201:  Are the 116 SNPs common to MLM and CMLM part of the 125 and 123 SNPs described in the previous sentence?  Which would mean only 9 SNPs unqiue to MLM and 6 SNPs unique to CMLM? 

Ln 218-224:  Identifying the Arabidopsis homolog of the mungbean candidate genes is actually information that should be included in the Materials and methods.  Further, additional BLAST information is needed.  Was the BLAST a nucleotide, peptide, or translated BLAST?   P-value <1?  That is incredibly permissive.  Similarlity Index is not a BLAST parameter – please clarify.

Necessary corrections have been made in material method and result sections.

Ln 234: PUP1 ad PSTOL need to be written out with abbreviations in parentheses

Necessary corrections have been made.

Ln 237: ANXUR needs to be written out with abbreviation in parentheses.

Necessary corrections have been made.

Ln 244-245:  How can the total phosphorus uptake of the genotypes be higher under LP if the mean value of the total phosphorus uptake is lower under LP compared to NP?

The mean values of TPU were significantly higher under NP condition than the LP condition. Whereas mean values of PUtE were higher under LP condition than NP condition. The low values of PUtE under NP condition were attributed to high TPU of genotypes. The formula, we used for calculation of PUtE is

The results presented here are in good agreement with the previous reports of P efficiency traits in rice (Wissuwa et al., 2015) and wheat (Yuan et al., 2017).

Ln 270:  Given how inter-connected TDW, PC, TPU, and PUtE are, it would be unusual if all four traits weren’t identified.

The total P uptake (TPU) (mg P/ dry weight) was calculated by multiplying the TDW and PC of the sample. The P utilization efficiency (PUtE) (g dry weight/mg P) was calculated using the following formula under both NP and LP conditions. The PUtE (g dry weight/mg P) = Total dry weight/ total P uptake by plant.

Ln 295: Which gene is ‘this gene’ refering to?  VRADI06G12490, PHR, or PSI?

It is PHR gene. The VRADI06G12490 is an SPX domain-containing gene family that is involved in the negative regulation of the PHR (Phosphate starvation response regulator) gene by regulating the expression of the PSI (Phosphate starvation-induced) gene

Ln 307: VRADI08G00070 is listed twice, but VRADI01G04370 isn’t listed at all.

Necessary corrections have been incorporated.

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report

I recognize the authors have put a lot of work towards revising this manuscript.  However, those modifications have really only addressed the surface and simple changes.  While the authors did clarify points of contention in the response to the reviewer letter, those clarifications were not incorporated into the body of the manuscript itself.  Thus, scientific concerns remain, which I feel can not be overlooked.  

I have addressed a number of them in the attached document. 

Comments for author File: Comments.pdf

Author Response

Comment: I recognize the authors have put a lot of work towards revising this manuscript; however, I am afraid I must still reject it. While the authors did clarify points of contention in the response to reviewer’s letter, those clarifications were not incorporated into the body of the manuscript itself. Further, there are still significant scientific concerns which I feel cannot be overlooked.

 

The majority of SNPs identified by GBS are likely outside gene coding regions. However, in this study, the only significant SNPs identified are within gene coding regions. This is statistically impossible. Further, the authors responded in the cover letter that they only considered SNPs within 6 and 4kb of a gene. Unless they have worked with a statistician or are basing this off a previously published study that did the same thing, I am extremely skeptical.

Response: In this study, we identified a total of 116 associated SNPs. Among these, 62 are in the intergenic region, 23 are in the coding region, 24 are in the upstream region and 7 are found in the downstream region. In our previous study (Reddy et al., 2020) we had annotated all the SNPs identified through GBS approach. We observed that 27% of the SNPs are residing in the intergenic region, 31% in the regulatory region, 23% in the coding region and 19% in the scaffold. The result suggests that most of the SNPs reside outside the coding sequence which is in line with the reviewer’s remark. In eukaryotes, generally 2-3 kb region as upstream and 1-2 kb region as downstream has been considered as ideal. However, more than 2 kb region can also be considered as upstream regulatory region as there may be the effect of enhancer or to check the repressive mark (Eden et al., 2001). Necessary details are included in the revised manuscript.

 

Comment: Table 1 provides excellent information. However, it should either be flipped (so gene IDs don’t have to be wrapped) or it should be provided as a supplementary file. Also, the column “Function” should be “Arabidopsis Function”.

Response: The necessary correction has been done in the revised manuscript.

 

Comment: Supplementary Table 1 only has 60 genotypes listed, not 120 as indicated in the Supplementary Table title, but also not the 144 genotypes referenced in the manuscript.

Response: We would like to request the reviewer to recheck the table as we found 120 genotypes to be listed.

 

Comment: I still don’t understand why so much time and effort was devoted to understanding the population structure if no SNPs were identified as unique to different sub-populations.

Response: Though, it is expected to have unique SNPs to different sub-population. However, in our study we did not find any.

 

Comment: I previously asked if the QTLs identified by the study were co-linear with other nutrient/stress QTLs in other legume species. The authors responded in the letter that the limited exploration of these traits in mungbean is why there’s no co-linearity with other species. However, the lack of data in mungbean is precisely why this type of analysis is so important. You can make a lot of scientific inferences about the role of genomic regions if those regions have been attributed to a QTL for nutrients in species A, the gene calls in species A and species B align, then the same region is likely to play a role for nturients in species B. While not a perfect answer, since the referenced genome for mungbean is a Vigna radiata and this study is on Vigna Wilczek, the candidate genes in the current study could be BLASTed to the V. radiata genome/gene calls on LegumeInfo.org and the results could be queried to see if genes in soybean and Phaseolus vulgaris are within nutrient/stress QTL regions. Simply identifying homologs of Arabidopsis genes known to be involved in phosphate processes is not enough.

Response: We would like to thank you very much for the constructive comments. In our GBS analysis we used Vigna radiata as reference genome only. The name Vigna radiata L. Wilczek is referred to the authorship purpose. We have also carried the BLAST search of the candidate genes against Phaseolus vulgaris genome and identified the orthologous gene which is provided in separate Supplementary Table 3.

 

Comment: Phosphorus is taken up by the roots and then utilized throughout the plant. While I don’t dispute gene activity in other parts of the plant might be important in Phosphorus utilization efficiency, I think it’s important to propose how the gene is involved in phosphorus utilization efficiency. Enhancing root growth is fairly obvious, more root surface area would allow more phosphorus uptake, but pollen tube is less obvious to me.

In this study, we dissected phosphorus utilization and uptake efficiency traits with a molecular breeding perspective. Therefore, we have not focused on physiological perspective of phosphorus utilization and uptake efficiency in mungbean. The details about regulation of phosphorus utilization and uptake through interacting gene network have been discussed extensively in earlier studies (Heuer et al. 2016; Dissanayaka et al. 2018).

That’s fine – you don’t have to consider the physiological aspect of all 16 genes, though it’s only 16. However, when a gene is so specialized (pollen tube), with no obvious connecgtion to the biological question at hand, I think the readers are owed an explanation, or the gene should be disgarded or expression proved through qRT-PCR. It’s also fine to say ‘we don’t know what role this gene plays, but it is recruited and participates in the processes, but it should at least be addressed.

Response: We are agreeing with the reviewer’s opinion. The expression of the genes in pollen tube has no obvious connection to the biological question relating to PUpE and PUtE. However, we have found the expression in the pollen tube comparatively at a lesser degree. We also found a smaller degree of expression of these genes in different kind of tissues in different developmental stages, viz. Ovary and hypocotyls. The phosphorous is an important component of biological macromolecules. After up-taking, it is sequestered in different component of plant tissues throughout the plant body. Therefore, it may be possible these genes play important role in sequestration in those tissues also. We believe that qRT-PCR can also not shed light on the exact mode of action of these genes. Functional characterization will be need which is beyond the scope of present study. Necessary details are included in the revised manuscript.

 

Comment: No additional information for the Expression Angler analyses was added to the text of the manuscript. Without more information/clarification, this isn’t publishable.

Response: Necessary details are included in the revised manuscript.

 

Comment: While the authors are correct that there is no digital gene expression platform for legumes, they could query their candidate genes in SoyBase and see what tissues and what conditions induce or repress expression in SoyBean – a closer relative than Arabidopsis.

Response: We would like to thank you for the constructive comments. We have performed BLAST search of the candidate genes against Phaseolus vulgaris genome. Then we performed expression analysis. The result is presented in a separate table 3.

 

Author Response File: Author Response.pdf