Association among aroma volatiles and other traits in one near-isogenic line with firm flesh texture

Abstract

Quality in melons (Cucumis melo L.) is an important criterion for consumer’s acceptance. Like any other fruit, melons development and ripening process are linked to a series of biochemical and physiological changes. The goal of these researches has been focused on determination of the global quality and aroma profile of a non-climacteric near-isogenic line (NIL) SC10-2 of melon developed from the Spanish cultivar T111 (ssp. melo) of the ‘Piel de Sapo’ type (PS) and the Korean accession PI 161375 Songwhan Charmi (ssp. agrestis) (SC), with different texture than its non-climacteric parental PS, in order to map QTLs associated with fruit quality and, particularly, aroma volatiles. A methodology for a consistent study of outliers of non-climacteric melon volatile organic compounds at harvest was reported. For extreme outliers, we proposed the study of compound classes in percentage of total area counts by principal component analysis followed by a further verification with box-whisker plot. The presence of esters usually at higher proportions than expected, or an excess of minority compounds such as ketones, alcohols, terpenes or others are usually a sign of volatile results that can be considered outliers. A comparison was performed in two seasons between the melon near-isogenic line SC10-2 and its parental PS in order to identify the most discriminant flesh aroma volatiles at harvest as potential biomarkers of textural differences (especially flesh firmness) due to introgressions in melon chromosome X, examining seasonal effects on these the former traits. Several univariate and multivariate statistical techniques were used to determine the association between the aroma volatile levels and differences in melon texture. Three methodologies involving different treatments of null values were applied to analyse the effects of introgression on volatile organic compounds (VOCs). Substitution of the zeros with the minimum value of the same VOC variable per season and line is proposed for the analysis of VOC data. SC10-2 showed higher whole fruit hardness, flesh firmness, and fibrousness, but lower juiciness than PS. One hundred sixteen VOCs were identified to evaluate the introgression effect, in which 60 VOCs were detected in two seasons with very different climatic conditions during fruit set and ripening; twenty-eight VOCs were involved in the differences on aroma production in the comparison of the NIL SC10-2 with PS. The environmental effects were more significant than the introgression effect. The environment particularly affected VOCs rather than textural traits, and at least one potential VOC quantitative trait locus (QTL) located in LG X alone and another interacting with the environment are proposed to affect VOC fruit metabolism. Transcriptomic analysis from RNA-seq revealed 2954 differentially expressed genes (DEGs) having introgression or introgression x ripening time effects and 2068 DEGs postharvest ripening time effect. Only 909 DEGs were exclusive of the introgression, and only 23 DEGs were exclusive of postharvest ripening time. Clusters of ripening time (six) revealed a general upregulation vs. harvest with slight interaction genotype x ripening time in 68.8% DEGs of four clusters, while clusters of the genotype (seven) revealed that the introgression generally produced noticeable differences in DEG upregulation or downregulation vs. the control PS in four clusters and 30% and 7.3% DEGs, respectively. Major functions of the DEGs associated with introgression or ripening time were delimited. About thirty-four genes were affecting direct and/or indirectly in the delay of ripening of SC10-2 versus PS in general and particularly to respiration rate, ethylene production, textural traits and volatile production and probably differential non-climacteric response. Of the former genes, we studied with more detail at least three that mapped in the introgression in linkage group (LG) X, and thirty-one outside it. The former results indicated the control of fruit physiology, but also textural and volatile traits traits, by an eQTL located in LG X. Some genes that mapped within the introgression in LG X that showed differential response between SC10-2 and PS could be associated with these quality differences between lines.