Mining candidate genes for grape seed traits based on a genome-wide association study

Abstract

Seedlessness has always been a valuable quality characteristic of edible grape varieties. Although the production of seedless grapes has been ongoing for decades, the genetic complexity of seedless grapes is not yet fully understood. Therefore, determining the genetic mechanisms and key regulatory genes of seedless grapes is of great significance for seedless grape breeding and meeting market demands. The emergence of high-throughput analysis software offers greater possibilities for mining genes related to plant organ development. Specifically, to mine a greater number of candidate genes related to grape seed traits, this study used the seed trait parameters analyzed by Tomato Analyzer as the target trait and then used a genome-wide association study (GWAS) to mine candidate genes. In the two-year analysis using principal component analysis (PCA), we extracted five principal components with a cumulative contribution rate of 96.586 %. The cumulative contribution rate for component 1 reached 87.352 %. Correlation analysis revealed correlation coefficients ranging from 0.54 to 0.98 among the seven basic traits. The GWAS results indicated that 370 SNP loci were significantly correlated with seed traits. These SNP loci were distributed on 18 chromosomes, except for chromosome 4, with most SNP loci distributed on chromosome 18. Based on the physical location of single nucleotide polymorphism (SNP) markers significantly associated with seed-related traits in the grape reference genome, candidate genes are screened within the range of linkage disequilibrium (LD) attenuation distance, both upstream and downstream of the significant SNP loci. These candidate genes were mainly transcription factor-related genes (VvMADS4 and VvMADS5), ubiquitin ligase-related genes (E3 ubiquitin ligase BIG BROTHER), serine/threonine protein kinase-related genes, and carbohydrate metabolism-related genes (Sucrose Synthase 2) and simultaneously controlled multiple (at least two or more) seed traits. These results indicate that seed traits are jointly regulated by some genes involved in seed morphology regulation. In this work, we identified new gene loci related to grape seed traits. Identifying molecular markers closely related to these seed traits is of great significance for breeding seedless grape varieties.