Molecular Characterisation of a Supergene Conditioning Super-High Vitamin C in Kiwifruit Hybrids

Round 1

Reviewer 1 Report

The authors reported the identification of the qAsA26.1 region, which contributes to high vitamin C concentration in kiwifruit.  The contents are quite interesting with being rigid experimental approaches.  I have no additional comments in the current manuscript; thereby, the manuscript could be acceptable as it is.　

I have an interest how this region has been inherited during the evolutionary process of this species.

Author Response

We thank the reviewer for their positive comments regarding the manuscript.

We are also very interested in the question of the evolutionary  history of the QTL and discuss this in L223-236 of the revised ms. This question will be central to future studies but we cannot speculate other than to hypothesise that it confers fitness value and has been preserved during speciation of divergence of A. eriantha

Reviewer 2 Report

This manuscript describes the use of pooled genotyping from divergent progeny showing large differences in AsA levels that appears to be controlled by a single major effect QTL. This work identified a number of SNPs associated with the trait on Chrm 26 between positions 2453803 and 9436328. Markers developed to a locus at 7647158 mapped to 1.5 cM from the locus controlling the high AsA trait, and a marker that maps to 8453577 was ~ 10 cM from the AsA controlling locus. RNAseq on lines with high and low AsA levels identified a region on Chrm 26 that contained 82 of the 113 differentially expressed transcripts. Although the authors did not discuss this possibility, it seems unlikely that so many genes with different functionalities would be differentially expressed and clustered to the same chromosomal region. It would be interesting to have the authors describe possible reasons for this observation (a deleted region in the low AsA lines for example?).   The methods are sound with adequate replication in the RNAseq portion of the study, and adequate progeny numbers in the mapping portion of the study. However, the methods section needs some work to better describe the processes used. As a reviewer, I read the methods section first, and significantly misinterpreted both the crosses that were only made clear in the results section, as well as the number of individuals that were phenotyped/genotyped in the analysis of the linked markers. However, the methods section does do an excellent job of describing the bioinformatics approaches that were used. The only way I could think of to improve it would be to include a supplemental file indicating the specific command scripts and program sources that were employed. Doing so might assist others (like myself) to both replicate their results or use their methods for similar analyses. The only other major – but easily remedied concerns, is a tendency to use language that could be interpreted to suggest that the associated SNPs are causal. It would also be nice to see the specific physical mapping locations of the 82 DETs on Chrm 26 to see how they relate to the mapping of the homozygous SNPs associated with low AsA levels. Doing so would clarify if there is a deletion in the low AsA lineages, and if so, how big it might be.  These and other minor suggestions and comments are noted in the marked up version of the manuscript.

Comments for author File: Comments.pdf

Author Response

Response to Reviewer 2: ‘Molecular Characterization of a Supergene Conditioning Super-High Vitamin C in Kiwifruit Hybrids’

We thank this reviewer for their detailed and very constructive consideration of our submission and for raising some important areas for clarification and improvement.

Point 1 Differential expression

‘ it seems unlikely that so many genes with different functionalities

would be differentially expressed and clustered to the same chromosomal

region. It would be interesting to have the authors describe possible reasons

for this observation (a deleted region in the low AsA lines for example?).’

The differential expression analysis we performed was based on transcripts from a de novo assembly from hybrid transcriptomes. Therefore, where loci are highly divergent between species  such as in the QTL region described it will tend to highlight transcripts that are only present in the lines harboring the  eriantha  alleles. We raised this in our discussion (L166-168) but have made revisions to clarify interpretation . Our original discussion clearly reported these results as ‘differentially expressed transcripts’, not differentially expressed genes.

We have made edits to emphasise differential transcript, not gene expression differences at:

·       L 162 Fig 8B  legend

·       L 193 ‘of transcripts homologous to…’

The whole chromosome alignment we present shows clearly that the two genome are highly collinear apart from the terminal repetitive regions {L 127-128; Fig A1}. We interpret the differential transcript expression seen in this region as primarily reflecting allelic divergence (not deletion) and therefore similar to the read assignment analysis. 

Point2 Methodology

·       Pedigrees of populations

We sympathize with the reviewer on this matter! These are exotic hybrid pedigrees that the lead author was much challenged to comprehend. We chose to present the pedigree figure in the results body as its is germane to considering the transmission of the A. eriantha allele to the populations used in GWAS and interpreting the segregation of AsA in Figure 1. If the editors consider that it is better to be placed in the methods section we are happy to comply.

·       Population sample sizes

o   This key detail has been added in section 4.1.2

·       Bioinformatics Methods

We thank the reviewer for their positive comments. During revision of this paper we greatly reduced detail of the bioinformatics methods following criticisms justifying rejection by another journal. We have added back some details into the methods , adding key command flags where possible. We have also added  a note under supplementary materials stating that details can be supplied on request

Point3 Causality vs Association of Markers

·       In no way do we wish to suggest causality of any variants identified, since our data show strong linkage over many megabases.

·       We have checked the Ms for any possible cases and are unable to identify any cases where we imply causality

·       We are happy to address any specific wording that the referee can identify

Point 4 physical mapping locations of the 82 DETs on Chromosome 26

·       These were provided in Supplementary table 1

·       We have also added a column to this table giving  in the gene models SNPS are located in

·       No differentially expressed transcripts mapped to genes containing SNPs identified in the poolseq

Reviewer 3 Report

McCallum et al. observed that in a cross between the cultivated and a wild kiwifruit, the concentration of vitamin C showed bimodal segregation, possibly indicating a major locus for this trait. They proceeded to analyzing this by GWAS of pooled samples and found a single QTL for this trait. Due to limited recombination in their hybrid populations, the QTL spans a long range, 5 Mb. However, they could sho that there is inhibition of recombination in this region, which therefore behaves like a single large genetic ocus, a “supegene”.

 As content of vitamin C is important in fruits intended for human consumption, their finding will definitely help breeding of kiwifruit into an even more healthy crop. However the manuscript has also general interest, since this kind of apprach is suitable for many similar situations. The paper is well written and was a pleasure to read.

Author Response

We thank the reviewer for  devoting time for review and for their positive comments regarding our manuiscript