Update of Genetic Linkage Map and QTL Analysis for Growth Traits in Eucommia ulmoides Oliver
Round 1
Reviewer 1 Report
English writing has to be improved throughout the whole document, especially focusing on verb conjugation, and numerous typos. Annexed the original manuscript with, highlighted in yellow, some of words and sentences that have to be corrected.
Hereafter, some specific comments.
Abstract
Line 15 : “..can elucidate the genetic mechanism of…” what ? Some words are probably missing
Introduction
Lines 33-34: the sentence is difficult to understand
Line 38: The arguments cited by the authors are more marketing arguments than scientific ones. What does it mean “tonifying kidney and impotence? Anti-fatigue? anti-aging? Please be more factual in your assertions or delete this sentence.
Materials and Methods
Mapping population
I understand that each F1 progeny individual is represented by only 1 plantlet, is it correct? This design without any replication (by cutting, grafting or tissue culture) is not optimal for the further estimation of genetic value and heritability.
SSR analysis
There is no indication on how the 2200 pairs of SSR primers were obtained, and no bibliographic reference on this point. I think readers of this article would be very interested to get this information.
The technique of AgNO3 staining used to visualize the alleles separated by electrophoresis on polyacrylamide gel is laborious and low throughput. For this reason, it has been usually replaced in many genotyping labs by other medium-throughput or even high-throughput techniques using fluorescently labeled primers. Though I understand the authors probably did not have other available options, I encourage them to envisage improving their SSR technologies in possible future studies.
QTL analysis
QTL detection should formally apply on genotypic value, i.e. when environmental variations have been estimated and subtracted from the phenotypic observations. Your planting design without replication does not allow the estimation of environmental variance, and consequently the correction of phenotypic values.
The search of candidate genes underlying QTLs is not well explained. Do the authors mean that they did BLASTX search with the sequence of the SSR nearest to the QTL ?
Results
SSR Analysis
How do the authors explain they did not find any SSR with 3 alleles (segregation ef x eg) or 4 alleles (segregation ab x cd)? Is it resulting from the way these SSRs were obtained? Please give an explanation.
Genetic linkage map
The authors did not mention in the introduction the haploid number of chromosome for E. ulmoides. I wonder if this number of 19 linkage groups match with the number of chromosomes of this species. Give further details and possibly comments in the discussion section.
Lines 160-161: maximum = 15, minimum = 154? Please correct.
Lines 171-173: This note is identical to the one of Table 2 and does not apply here for Table 1.
QTL analysis
In the M&M section, the authors mention the calculation by permutation test of the threshold value for LOD score signification, but this threshold value does not appear here. Please give this value.
Explanations on how were identified candidate genes were not convincing. It lacks a lot of crucial information to be sure that genes identified by BLASTX are good candidate genes that might explain the observed QTLs. Otherwise this BLASTX search does not make sense. The necessary data are related to the genetic distance between the flanking marker and the QTL, the approximate physical distance (in nucleotides) it represents, and the expected number of genes in this interval.
Discussion
Segregation distortion
Segregation distortion affecting some loci in a biparental population is a very common feature, and the amount of 4.87% mentioned here is not out of the standards. This comprehensive discussion about segregation distortion is therefore irrelevant.
Genetic linkage map
The increase number of linkage groups from 12 to 19 since the last study should be further discussed. The ideal situation is when the number of detected linkage groups is equal to the number of chromosomes. Is it the case here ?
Line 256: Once again, segregation distortion cannot generally be anticipated, and it is not always a flaw. It can even in some cases be informative about biological features of the studied species.
QTL analysis
The discussion on QTL analysis is difficult to understand. The authors claim the identification of 72 QTLs, but obviously, it appears that many of them are not different QTLs, but QTLs that contribute to the variation of one trait (i.e. tree height or ground diameter) during several years. It should therefore be interesting to get a synthetic presentation (a table) figuring how many distinct loci were identified, on how many years they have been detected, and possibly, if they were detected for different traits (ground diameter and tree height for example).
Author Response
Please see the attachment.
Author Response File: 
Author Response.docx
Reviewer 2 Report
The authors established earlier a linkage map in Eucommia ulmoides (2n = 34), composed of 706 markers in 25 linkage groups and with 3.1cM average density. In this paper, the authors describe an updated version of 869 markers mapped into 19 major groups with 2.2 cM average density. The current map integrates many SSR markers from available transcriptomic data. The authors detected 75 growth-related QTLs affecting tree height, ground diameter and crown diameter over 10 consecutive years. Few candidate genes are reported.
The subject matter falls within the scope of "Forests" journal. The paper shows an original contribution in the area of Eucommia genetics. However, I cannot recommend the current version for publication. There are several inconsistencies and major defaults in the manuscript, which are explained in detail below.
Linkage mapping
- There were quite a number of markers which could not be linked to any other marker. This is unexpected in a progeny of this size and given the total marker number, which is enough to get a very dense map. While 1949 markers used for the leakage analysis, only 869 were grouped to the 19 linkage groups. The remaining 1080 markers were either unlinked (784 markers ) or formed doublets, triplets etc. This may point to large number of missing values (technical problems encountered during analysis) or considerable genotypic errors. No explanation is provided.
- There are more linkage groups than there are chromosomes. With this progeny size (N= 152) and the involvement of new SSR markers it should not be so problematic to obtain a linkage map with better coverage and density. Again this point to genotypic errors or possibly to outcross (fathers from outside?) in the progeny.
Phenotypic traits
- The authors mention (p10, line 258) that growth rates between 5a (2014) and 6a (2015) was very slow due to transplantation in 2014. This is true for ground diameter. Diameter increased surprisingly almost 2 folds in year 6 (Table 2). Is it likely that frost damage or insect infection in the shoot apical meristem took place during early year 6? If this is the case, then some of the detected QTLs (in year 6) may reflect biotic or abiotic stress resistance and rather growth. The authors should clarify this.
- In the distribution of the traits, the values of the parents are not indicated.
QTL analysis
- QTL analysis for height and diameter has been measured previously over four consecutive years using an earlier map version (Li et al, 2014). I was wondering if the current analysis includes the 4-year phenotypic data used in the previous study. The authors must clarify this.
- In my opinion the most interesting part of the work is the assessment of the growth traits over 10 consecutive years. This provides the possibility to the authors to make valuable statement about QTL stability over the years. Nothing is really mentioned. The authors should estimate the QTL*year (or genotype * year) interaction and make relative conclusions. The QTL*year interaction might appeal to a broader readership.
- The authors need to evaluate their results more critical. For example, the phenotypic assessment was undertaken in a single growing environment. QTL analysis under different environmental conditions is not considered in the present study. The authors should refer to this when they make conclusions.
Candidate genes
- The candidate genes approach used is unclear. No sufficient details are provided.
-The significance of the predicted candidate genes is not discussed in sufficient detail.
Although these genes need further experimental validation, the authors need also to speculate the significance of the predicted genes to growth of Eucommia. For example, how the ABIL1 subunit of the WAVE complex which, induce branched F-actin, can affect growth in Eucommia? In plants, actin filaments affect cell shape by modulating transport properties of organelles and cargos to control mechanical properties of the cell wall. It is becoming clear lately (in plants, invertebrates and mammals) that cells can balance actin filament networks (i.e. branched vs long linear actin bundles) to coordinate processes like long-distance nutrient transport, endocytosis and secretion. Perhaps, this cellular control can modulate various components of the tree growth process.
Text writing
- I strongly recommend a critical screening of the manuscript for grammatical errors and inconsistency. Text is often unclear and inconsistent to results.
Few examples:
- p2, line 59, ... "high heterozygous of trees" should be "high heterozygosity of trees"
- p2, line 60, "paeudo-testcross mapping straety" should be "pseudo-testcross mapping strategy"
- p2, line 67..... "Li [21] have constructed a genetic linkage map .... for E. ulmoides, containing 529 markers distributed over 12 linkage groups (with more than 15 markers)". Where is this info coming from? This is inconsistent with the Li et al. 2014 (ref#21).
- p2, line 68 .... "(with more than 15 markers)" I would suggest ..."(at least 15 markers in each linkage group)".
- The authors should define the acronyms SRAP, AFLP, ISSR and SSR (once in the text).
- p4, line 161-162. It needs correction the following sentence: "The maximum number of makers per linkage group was 15 (G17), the minimum number was 154 (G1)".
Does not reflect the results presented in table 1.
- p10, line 264. "Decennial trees .... been the primary issues". "Tree cultivation is a long-term task, thus the construction of nursery and seed .... experimental results". These sentences need to be revised or omitted. Unclear meaning.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 3 Report
This manuscript describes the construction of an update linkage map, adding addition SSR markers developed in this study. The obtained map was used to predict QTLs for height, ground diameter and crown diameter. QTL mapping was done for each trait for each year separately.
Is the ploidy, and number of chromosomes known for this species?
Line 150: “452 polymorphic SSR markers were obtained from 365 SSR primers” How were the SSR loci genotyped? It seems from the results that they were scored as dominant markers (band presence-absence). This should be clarified.
Table 1 legend seems to be from a different table. The original linkage map contained 12 LG. How does this relate to the 19 LG found in this study? In particular, in Table 1, the column “Number of added SSR” is less than the number of markers in each linkage group.
Section 3.3 Growth traits. It might be more comprehensible to replace the year designations (H1, H2 etc.) with the actual years (H2010, H2011 etc).
Table 3. The SSR markers derived from this study could be differentiated from the markers from the original map.
In the discussion (lines 282-285), it is stated that several loci were linked to QTLs in multiple years, and that some QTLs were associated with several traits (Lines 291-292). These should be presented in the results. They can be found in Table 3, but these could be highlighted or placed in a separate table or figure to show which loci were found in multiple years or linked to multiple traits.
The changes in growth traits are discussed in the discussion (lines 258-268). This could be expanded to include some discussion about the variation in identified QTLs. In particular, are the different linked markers identified in each year related to changes in growth parameters (which could be affected by external factors, as mentioned in section 4.3). However, in section 4.4 (line 288-289), it is stated that “growth traits of trees are easily influenced by surrounding environment”. This should be illustrated by data from the current study – e.g. highlighting the years where growth was affected by competition or other factors, and if this had an influence on the identified QTLs.
Growth rates could be calculated from the data – could these be more useful for use in QTL mapping, to identify loci for biomass accumulation?
Spell check and grammar correction are needed throughout the manuscript.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Round 2
Reviewer 2 Report
No further revision is needed.
