Genome-wide Association Study of Individual Sugar Content and Sugar Conversion in Fruit of Japanese Pear (Pyrus spp.)

Background: Sweetness is one of the most important traits determining fruit quality. Sweetness is controlled not only by the total sugar content but also by the contents of individual sugars. The major sugars in mature Rosaceae fruits are sucrose, fructose, glucose, and sorbitol, which have different levels of sweetness. Among these, sucrose and fructose have high sweetness, whereas glucose and sorbitol have low sweetness. The objective of this study was to identify the quantitative trait loci (QTLs) associated with fruit traits including individual sugar accumulation and conversion, to infer the candidate genes underlying the QTLs, and to assess the potential of genomic selection for breeding pear fruit traits. Results: : We evaluated 10 fruit traits and conducted genome-wide association studies (GWAS) for 106 cultivars and 17 breeding populations (1112 F1 individuals) using 3484 tag single-nucleotide polymorphisms (SNPs) genotyped by double-digest restriction-site associated DNA sequencing (ddRAD-Seq). By implementing a mixed linear model and a Bayesian multiple-QTL model in GWAS, 56 SNPs associated with fruit traits were identified. Four loci were presumed to be associated with sugar conversion because the SNPs were significant for more than one individual sugar and the individual sugar contents associated with each SNP genotype were negatively correlated. In particular, a SNP located close to acid invertase gene PPAIV3 on chromosome 7 and a newly identified SNP on chromosome 11 had quite large effects on sugar conversion. We used ‘Golden Delicious’ doubled haploid (GDDH) 13, an apple reference genome to infer the candidate genes for the identified SNPs. In the region flanking the SNP on chromosome 11, there is a tandem repeat of early responsive to dehydration ( ERD6 )-like sugar transporter genes which might play a role in the phenotypes observed. Conclusions: : SNPs associated with sugar accumulation and conversion were newly identified at several loci, and candidate genes underlying QTLs were inferred using advanced apple genome information. Several QTLs showed clear effects with more than 10% of phenotypic variance explained by those SNPs in the breeding populations. By combining the effects of multiple QTLs, breeders would be able to select seedlings that will later bear fruit with high sucrose and fructose content.