Abstract: Dissecting quantitative traits into Mendelian factors is a great challenge in genetics. Apple fruit storability is a complex trait controlled by multi-genes with unequal effects. We previously identified 62 quantitative trait loci (QTLs) associated with apple fruit storability and genomics-assisted prediction (GAP) models were trained using 56 QTL-based markers. Here, three candidate genes, MdNAC83, MdBPM2, and MdRGLG3, were screened from the regions of QTLs with large G’ value and large genetic effects. Both a 216-bp deletion and an SNP934 T/C at the promoter of MdNAC83 were associated with higher MdNAC83 expression but an SNP388 G/A at the coding region significantly reduced the activity to activate the expression of the target genes MdACO1, MdMANA3, and MdXTH28. MdBPM2 and MdRGLG3 participated in the ubiquitination of MdNAC83. SNP657 T/A of MdBPM2 and SNP167 C/G of MdRGLG3 caused a reduction in the activity to ubiquitinate MdNAC83. By the addition of functional markers to the GenoBaits SNP array, the prediction accuracy of the updated GAP models increased to 0.7723/0.6231 and 0.5639/0.5345 for flesh firmness/crispness at harvest and flesh firmness/crispness retainability, respectively. The variation network involving eight simple Mendelian variations in six genes helps to gain insight into the molecular quantitative genetics, to improve breeding strategy, and to provide targets for future genome editing.

Key words: fruit storability, genetic variation, Malus domestica Borkh, protein ubiquitination, quantitative trait