System Approach to Understanding the Metabolic Diversity in Melon

Abstract

Fruit quality is determined by numerous genetic factors that affect taste, aroma, ‎color, texture, nutritional value and shelf life. To unravel the genetic components ‎involved in the metabolic pathways behind these traits, the major goal of the project was to identify novel genes that are involved in, or that regulate, these pathways using correlation analysis between genotype, metabolite and gene expression data. The original and specific research objectives were: (1) Collection of replicated fruit from a population of 96 RI lines derived from parents distinguished by great diversity in fruit development and quality phenotypes, (2) Phenotypic and metabolic profiling of mature fruit from all 96 RI lines and their parents, (3) 454 pyrosequencing of cDNA representing mRNA of mature fruit from each line to facilitate gene expression analysis based on relative EST abundance, (4) Development of a database modeled after an existing database developed for tomato introgression lines (ILs) to facilitate online data analysis by members of this project and by researchers around the world. The main functions of the database will be to store and present metabolite and gene expression data so that correlations can be drawn between variation in target traits or metabolites across the RI population members and variation in gene expression to identify candidate genes which may impact phenotypic and chemical traits of interest, (5) Selection of RI lines for segregation and/or hybridization (crosses) analysis to ascertain whether or not genes associated with traits through gene expression/metabolite correlation analysis are indeed contributors to said traits. The overall research strategy was to utilize an available recombinant inbred population of melon (Cucumis melo L.) derived from phenotypically diverse parents and for which over 800 molecular markers have been mapped for the association of metabolic trait and gene expression QTLs. Transcriptomic data were obtained by high throughput sequencing using the Illumina platform instead of the originally planned 454 platform. The change was due to the fast advancement and proven advantages of the Illumina platform, as explained in the first annual scientific report. Metabolic data were collected using both targeted (sugars, organic acids, carotenoids) and non-targeted metabolomics analysis methodologies. Genes whose expression patterns were associated with variation of particular metabolites or fruit quality traits represent candidates for the molecular mechanisms that underlie them. Candidate genes that may encode enzymes catalyzingbiosynthetic steps in the production of volatile compounds of interest, downstream catabolic processes of aromatic amino acids and regulatory genes were selected and are in the process of functional analyses. Several of these are genes represent unanticipated effectors of compound accumulation that could not be identified using traditional approaches. According to the original plan, the Cucurbit Genomics Network (http://www.icugi.org/), developed through an earlier BARD project (IS-3333-02), was expanded to serve as a public portal for the extensive metabolomics and transcriptomic data resulting from the current project. Importantly, this database was also expanded to include genomic and metabolomic resources of all the cucurbit crops, including genomes of cucumber and watermelon, EST collections, genetic maps, metabolite data and additional information. In addition, the database provides tools enabling researchers to identify genes, the expression patterns of which correlate with traits of interest. The project has significantly expanded the existing EST resource for melon and provides new molecular tools for marker-assisted selection. This information will be opened to the public by the end of 2013, upon the first publication describing the transcriptomic and metabolomics resources developed through the project. In addition, well-characterized RI lines are available to enable targeted breeding for genes of interest. Segregation of the RI lines for specific metabolites of interest has been shown, demonstrating the utility in these lines and our new molecular and metabolic data as a basis for selection targeting specific flavor, quality, nutritional and/or defensive compounds. To summarize, all the specific goals of the project have been achieved and in many cases exceeded. Large scale trascriptomic and metabolomic resources have been developed for melon and will soon become available to the community. The usefulness of these has been validated. A number of novel genes involved in fruit ripening have been selected and are currently being functionally analyzed. We thus fully addressed our obligations to the project. In our view, however, the potential value of the project outcomes as ultimately manifested may be far greater than originally anticipated. The resources developed and expanded under this project, and the tools created for using them will enable us, and others, to continue to employ resulting data and discoveries in future studies with benefits both in basic and applied agricultural - scientific research.