Abstract
Tomato fruit texture is one of the most critical quality traits for both the consumer and the production chain. Texture is a complex trait for which several QTL and genes were found. However, interactions between the molecular, histological, physical and biochemical components of fruit texture have been rarely investigated. In this work an integrative approach based on multiple co-inertia analysis (MCOA) was applied to point out links among the different levels: from protein to fruit, then to identify main physiological mechanisms involved in fruit texture. Three contrasted parental lines (‘Cervil’, ‘Levovil’, ‘VilB’) and three derived QTL-NILs harbouring texture QTL on chromosome 4 and 9 were analyzed. Measurements were performed at cell expansion stage, at red ripe stage and after 7-days postharvest storage at 20°C. To increase texture variability, water deficit was induced by decreasing water supply by 40% from flowering of the third truss. Three blocks of data (texture, physico-chemical traits and proteome datasets) were analysed. Results showed a common multi-scale structure obtained from the three datasets with a main contribution of texture and biochemical blocks. At all levels, MCOA outlined strong genotype discrimination, indicating that the genetic factor was the main factor of variability, in contrast with water deficit. The first common component separated the genetic background and correlated with firmness and sugar traits. The second component represented QTL effect. The percentage of the variance of the protein block taken into account to build the common structure was low. Proteins which mainly contributed to the common components at the three developmental stages were implicated in carbohydrate metabolism. Multiple co-inertia analysis provides an interesting tool to characterize complex trait such as fleshy fruit texture by integrating several levels of studies.