Multivariate genome-wide association study of leaf shape in a Populus deltoides and P. simonii F1 pedigree

Abstract

Leaf morphology exhibits tremendous diversity between and within species, and is likely related to adaptation to environmental factors. Most poplar species are of great economic and ecological values and their leaf morphology can be a good predictor for wood productivity and environment adaptation. It is important to understand the genetic mechanism behind variation in leaf shape. Although some initial efforts have been made to identify quantitative trait loci (QTLs) for poplar leaf traits, more effort needs to be expended to unravel the polygenic architecture of the complex traits of leaf shape. Here, we performed a genome-wide association analysis (GWAS) of poplar leaf shape traits in a randomized complete block design with clones from F1 hybrids of Populus deltoides and Populus simonii. A total of 35 SNPs were identified as significantly associated with the multiple traits of a moderate number of regular polar radii between the leaf centroid and its edge points, which could represent the leaf shape, based on a multivariate linear mixed model. In contrast, the univariate linear mixed model was applied as single leaf traits for GWAS, leading to genomic inflation; thus, no significant SNPs were detected for leaf length, measures of leaf width, leaf area, or the ratio of leaf length to leaf width under genomic control. Investigation of the candidate genes showed that most flanking regions of the significant leaf shape-associated SNPs harbored genes that were related to leaf growth and development and to the regulation of leaf morphology. The combined use of the traditional experimental design and the multivariate linear mixed model could greatly improve the power in GWAS because the multiple trait data from a large number of individuals with replicates of clones were incorporated into the statistical model. The results of this study will enhance the understanding of the genetic mechanism of leaf shape variation in Populus. In addition, a moderate number of regular leaf polar radii can largely represent the leaf shape and can be used for GWAS of such a complicated trait in Populus, instead of the higher-dimensional regular radius data that were previously considered to well represent leaf shape.

Genome-wide association study of leaf-related traits in tea plant in Guizhou based on genotyping-by-sequencing

“…Genome-wide association analysis has identified many SNP associated with leaf morphological characteristics [ 73 , 74 ]. However, marker-assisted selection (MAS) breeding is still relatively less applied.…”

Association Mapping and Expression Analysis of the Genes Involved in the Wood Formation of Poplar

“…However, it was found that growth variation in the offspring of artificial interspecific crosses was significantly genetically different, showing a great deal of variability [ 29 ]. Therefore, artificial populations of half-sib or full-sib families with a clear genetic background and rich variation have become widely studied in forest tree breeding [ 30 , 31 , 32 ]. In addition, although the genomic recombination rate in the artificial breeding population is low, it can amplify the frequency of rare mutation sites and detect some additional significant associations through adjusting the threshold of the p -value [ 32 ], also a more convenient and simpler method of preserving germplasm resources than natural populations [ 33 ].…”

“…Therefore, artificial populations of half-sib or full-sib families with a clear genetic background and rich variation have become widely studied in forest tree breeding [ 30 , 31 , 32 ]. In addition, although the genomic recombination rate in the artificial breeding population is low, it can amplify the frequency of rare mutation sites and detect some additional significant associations through adjusting the threshold of the p -value [ 32 ], also a more convenient and simpler method of preserving germplasm resources than natural populations [ 33 ]. Therefore, there is potential and feasibility to utilize a full-sib artificial hybrid population of P. deltoides with P. × euramericana for GWAS in this study.…”

“…The QQ plot and genomic inflation λ are two metrics commonly used to select an optimization association model [ 26 , 27 ]. Yang et al selected a model to associate with the leaves of the F1 pedigree of P. deltoides and P. simonii based on good genomic control (genomic inflation equal to 1) [ 32 ]. Furthermore, Zhang et al selected the most effective model for poplar GWAS based on the degree of deviation between the observed and expected p values [ 34 ].…”

Genome‐wide association study of multiyear dynamic growth traits in hybrid Liriodendron identifies robust genetic loci associated with growth trajectories

“…Moreover, because of differences in the genetic bases of these offspring, their growth rhythms also vary at different stages (Ji et al., 2005; Li et al., 2017; Zhao & Zhang, 2013; Zhu, 2020), resulting in abundant diversity in growth processes, which improves the efficiency of a GWAS. Thus, for tree species with a weak breeding basis, artificial populations of half‐sib or full‐sib families with clear genetic backgrounds and rich variation have become popular (Chen et al., 2021; Lu et al., 2019; Osorio‐Guarin et al., 2019; Tan et al., 2017; Yang et al., 2021). Additionally, although an artificial breeding population has a low rate of genome recombination, it can amplify the frequency of rare mutant loci and, compared to natural populations, is also a more convenient and simpler approach for the preservation of germplasm resources (Li, Boer, et al., 2021; Müller et al., 2019), especially for tree species that depend on cross‐breeding.…”