Earliness and ripening behavior are important attributes of fruits on and off the vine, and affect quality and preference of both growers and consumers. Fruit ripening is a complex physiological process that involves metabolic shifts affecting fruit color, firmness, and aroma production. Melon is a promising model crop for the study of fruit ripening, as the full spectrum of climacteric behavior is represented across the natural variation. Using Recombinant Inbred Lines (RILs) population derived from the parental lines “Dulce” (reticulatus, climacteric) and “Tam Dew” (inodorus, non-climacteric) that vary in earliness and ripening traits, we mapped QTLs for ethylene emission, fruit firmness and days to flowering and maturity. To further annotate the main QTL intervals and identify candidate genes, we used Oxford Nanopore long-read sequencing in combination with Illumina short-read resequencing, to assemble the parental genomes de-novo. In addition to 2.5 million genome-wide SNPs and short InDels detected between the parents, we also highlight here the structural variation between these lines and the reference melon genome. Through systematic multi-layered prioritization process, we identified 18 potential polymorphisms in candidate genes within multi-trait QTLs. The associations of selected SNPs with earliness and ripening traits were further validated across a panel of 177 diverse melon accessions and across a diallel population of 190 F1 hybrids derived from a core subset of 20 diverse parents. The combination of advanced genomic tools with diverse germplasm and targeted mapping populations is demonstrated as a way to leverage forward genetics strategies to dissect complex horticulturally important traits.

Melon is an important crop that exhibits broad variation for fruit morphology traits that are the substrate for genetic mapping efforts. In the post-genomic era, the link between genetic maps and physical genome assemblies is key for leveraging QTL mapping results for gene cloning and breeding purposes. Here, using a population of 164 melon recombinant inbred lines (RILs) that were subjected to genotyping-by-sequencing, we constructed and compared high-density sequence- and linkage-based recombination maps that were aligned to the reference melon genome. These analyses reveal the genome-wide variation in recombination frequency and highlight regions of disrupted collinearity between our population and the reference genome. The population was phenotyped over 3 years for fruit size and shape as well as rind netting. Four QTLs were detected for fruit size, and they act in an additive manner, while significant epistatic interaction was found between two neutral loci for this trait. Fruit shape displayed transgressive segregation that was explained by the action of four QTLs, contributed by alleles from both parents. The complexity of rind netting was demonstrated on a collection of 177 diverse accessions. Further dissection of netting in our RILs population, which is derived from a cross of smooth and densely netted parents, confirmed the intricacy of this trait and the involvement of major locus and several other interacting QTLs. A major netting QTL on chromosome 2 co-localized with results from two additional populations, paving the way for future study toward identification of a causative gene for this trait.

Deciduous trees require sufficient chilling during winter dormancy to grow. To decipher the dormancy-regulating mechanism, we carried out RNA sequencing (RNA-Seq) analysis and metabolic profiling of European pear (Pyrus communis L.) vegetative buds during the dormancy phases. Samples were collected from two cultivars that differed greatly in their chilling requirements: Spadona' (SPD), a low chilling requirement cultivar; and Harrow Sweet (HS), a high chilling requirement cultivar. Comparative transcriptome analysis revealed >8500 differentially expressed transcripts; most were related to metabolic pathways. Out of 174 metabolites, 44 displayed differential levels in both cultivars, 38 were significantly changed only in SPD, and 15 only in HS. Phospholipids were mostly accumulated at the beginning of dormancy, sugars between before dormancy and mid-dormancy, and fatty acids, including α-linolenic acid, at dormancy break. Differentially expressed genes underlying previously identified major quantitative trait loci (QTLs) in linkage group 8 included genes related to the α-linolenic acid pathway, 12-oxophytodienoate reductase 2-like, and the DORMANCY-ASSOCIATED MADS-BOX (DAM) genes, PcDAM1 and PcDAM2, putative orthologs of PpDAM1 and PpDAM2, confirming their role for the first time in European pear. Additional new putative dormancy-related uncharacterized genes and genes related to metabolic pathways are suggested. These results suggest the crucial role of α-linolenic acid and DAM genes in pear bud dormancy phase transitions. © 2018 The Author(s).

BACKGROUND: Genomic analysis technologies can promote efficient fruit tree breeding. Genotyping by sequencing (GBS) enables generating efficient data for high-quality genetic map construction and QTL analysis in a relatively accessible way. Furthermore, High-resolution genetic map construction and accurate QTL detection can significantly narrow down the putative candidate genes associated with important plant traits. RESULTS: We genotyped 162 offspring in the F1 'Spadona' x 'Harrow Sweet' pear population using GBS. An additional 21 pear accessions, including the F1 population's parents, from our germplasm collection were subjected to GBS to examine diverse genetic backgrounds that are associated to agriculturally relevant traits and to enhance the power of SNP calling. A standard SNP calling pipeline identified 206,971 SNPs with Asian pear ('Suli') as the reference genome and 148,622 SNPs with the European genome ('Bartlett'). These results enabled constructing a genetic map, after further stringent SNP filtering, consisting of 2036 markers on 17 linkage groups with a length of 1433 cM and an average marker interval of 0.7 cM. We aligned 1030 scaffolds covering a total size of 165.5 Mbp (29%) of the European pear genome to the 17 linkage groups. For high-resolution QTL analysis covering the whole genome, we used phenotyping for vegetative budbreak time in the F1 population. New QTLs associated to vegetative budbreak time were detected on linkage groups 5, 13 and 15. A major QTL on linkage group 8 and an additional QTL on linkage group 9 were confirmed. Due to the significant genotype-by-environment (GxE) effect, we were able to identify novel interaction QTLs on linkage groups 5, 8, 9 and 17. Phenotype-genotype association analysis in the pear accessions for main genotype effect was conducted to support the QTLs detected in the F1 population. Significant markers were detected on every linkage group to which main genotype effect QTLs were mapped. CONCLUSIONS: This is the first vegetative budbreak study of European pear that makes use of high-resolution genetic mapping. These results provide tools for marker-assisted selection and accurate QTL analysis in pear, and specifically at vegetative budbreak, considering the significant GxE and phenotype-plasticity effects.

Abstract Dormancy release is greatly affected by chilling unit (CU) accumulation. Lack of CU has a major impact on spring vegetative budbreak (VB). To understand the genetic mechanism governing the chilling requirement (CR), we conducted a QTL analysis of VB date in F1 population, derived from a cross between ‘Spadona’ (low CR) and ‘Harrow Sweet’ (high CR). Using a unique methodology of tree mobility, replicates of the same genotypes were exposed during the winter, over two consecutive years, to climates that differ greatly in their CU and to the same heat conditions to induce VB, in order to evaluate CR genetic impact and to distinguish it from the heat factor. Broad-sense heritability within locations ranged from 0.62 to 0.66. Due to a strong impact of GxE interaction, it was reduced to 0.46 for the overall mean. We examined the previously discovered apple QTLs detected in linkage groups (LG) 9 and 8, based on the synteny between the species. Our analysis confirms significant QTLs in LG8 (R2 = 12%–24%) and LG9 (R2 = 20%–38%) for all locations and years