Single Nucleotide Polymorphism Detection for Peach Gummosis Disease Resistance by Genome-Wide Association Study

Abstract

Peach gummosis is one of the most widespread and destructive diseases. It causes growth stunting, yield loss, branch, trunk, and tree death, and is becoming a restrictive factor in healthy and sustainable development of peach production. Although a locus has been identified based on bi-parental quantitative trait locus (QTL) mapping, selection of gummosis-resistant cultivars remains challenging due to the lack of resistant parents and of the complexity of an inducing factor. In this study, an integrated approach of genome-wide association study (GWAS) and comparative transcriptome was used to elucidate the genetic architecture associated with the disease using 195 accessions and 145,456 genome-wide single nucleotide polymorphisms (SNPs). The broad-sense and narrow-sense heritabilities were estimated using 2-year phenotypic data and genotypic data, which gave high values of 70 and 73%, respectively. Evaluation of population structure by neighbor-joining and principal components analysis (PCA) clustered all accessions into three major groups and six subgroups, mainly according to fruit shape, hairy vs. glabrous fruit skin, pedigree, geographic origin, and domestication history. Five SNPs were found to be significantly associated with gummosis disease resistance, of which SNPrs285957, located on chromosome6 across 28 Mb, was detected by both the BLINK and the FarmCPU model. Six candidate genes flanked by or harboring the significant SNPs, previously implicated in biotic stress tolerance, were significantly associated with this resistance. Two highly resistant accessions were identified with low disease severity, which could be potential sources of resistance genes for breeding. Our results provide a fresh insight into the genetic control of peach gummosis disease.

Keywords: QTLs; candidate genes; genome-wide association study; gummosis disease; peach.

FIGURE 1

The density distribution of peach gummosis in 195 peach accessions. Disease scale of 0–9 was used, where 0 and 1 represent highly resistant and 9, highly susceptible accessions. The red density distribution is the phenotypic data obtained from 2018; green is the density distribution of the phenotypic data obtained from 2019, and blue represents the density distribution of the best linear unbiased predictions (BLUPs) expressed as estimated breeding values.

FIGURE 2

Narrow-sense heritability of the resistance of peach gummosis disease calculated by whole genome SNP markers. “Genetic” and “Residual” means estimated genetic and residual variance in the mixed linear model. The optimum compression information indicates the optimal algorithm to calculate the group kinship matrix, the optimal clustering algorithm, the optimal number of groups in the compress mixed linear model. “-2 LL” is the abbreviation of -2 multiply likelihood value, which means the level of model fitting.

FIGURE 3

(A) The geographic location of the origins of each accession in China and worldwide. (B) Phylogenetic dendrogram constructed by the neighbor-joining method. The accession name is represented by the accession code, which is coincident with Supplementary Table 1. (C) Principal component analysis (PCA) of accessions, with the proportion of variance explained by each PC indicated in parenthesis. Dots of different color indicate different cluster groups. (D) Linkage disequilibrium measures (r²) against physical distance between pairs of SNP markers for the three major groups.

FIGURE 4

Distribution of SNPs on the eight chromosomes. Tracks 1, 2, and 3 represent results from three statistical analysis models used for genome-wide association study (GWAS) of peach gummosis disease. Track 4 represents filtered SNPs on the eight peach chromosomes. The red dotted-line indicates the significance threshold (-log₁₀ P = 6). The red asterisks represent the significant SNP.

FIGURE 5

The boxplots show the comparison of the phenotypic performance illustrated by the BLUP value for different genotypes of the five significant SNPs. For each graph, X axis indicates the different genotypes of SNP rs22118 (A), rs96598 (B), rs142398 (C), rs191998 (D), and rs285957 (E). Y axis indicates the BLUP value.

FIGURE 6

Significant associations and candidate genes on Chromosome 6, underlying peach gummosis disease. (A) Manhattan plots showing the significance of SNP rs285957 at the chromosome-wide level. The vertical blue lines indicate the position of the significant locus identified by three different models. The annotated candidate genes and the gene structure are represented below the plot. (B) Pairwise correlation of LD (r²) between significant SNPs along the highlighted genomic region. (C) Relative expression of Prupe.6G315800 obtained by comparative RNA-Seq profile during the inoculation of “Huyou018.”