Combining ecophysiological models and genetic analysis: a promising way to dissect complex adaptive traits in grapevine

摘要

Designing genotypes with acceptable performance under warmer or drier environments is essential for sustainable crop production in view of climate change. However, this objective is not trivial for grapevine since traits targeted for genetic improvement are complex and result from many interactions and trade-off between various physiological and molecular processes that are controlled by many environmental conditions. Integrative tools can help to understand and unravel these Genotype × Environment interactions. Indeed, models integrating physiological processes and their genetic control have been shown to provide a relevant framework for analyzing genetic diversity of complex traits and enhancing progress in plant breeding for various environments. Here we provide an overview of the work conducted by the French LACCAVE research consortium on this topic. Modeling abiotic stress tolerance and fruit quality in grapevine is a challenging issue, but it will provide the first step to design and test in silico plants better adapted to future issues of viticulture. Introduction Exploiting genetic diversity or designing new scion varieties and rootstocks with better performance under water stress or high temperature is one of the possible paths to sustain high-quality viticulture in view of future climate change (Duchêne, 2016). However, this breeding challenge is not trivial for perennial fruit crops, including grapevine, since the main traits targeted for genetic improvement (e.g. plant growth and tolerance to abiotic stresses, yield, fruit quality) are quantitative and complex, as they result from many interactions and trade-off between various physiological and molecular processes that (i) act at different temporal, spatial and structural scales and (ii) depend on environmental conditions and management strategies.Ecophysiological process-based models (PBMs) can predict quantitative traits of one genotype in any environment, whereas quantitative trait locus (QTL) models predict the contribution of alleles under a limited number of environments (Tardieu, 2003). Approaches combining both ecophysiological modeling and QTL analyses have been developed recently (Hammer et al., 2006; Reymond et al., 2003; Yin et al., 1999), essentially in annual crops, to overcome the strong G×E interactions in the control of complex traits in plants and to improve QTL detection power. The method dissects the genotypic variation of a given complex trait into simpler ecophysiological model parameters linked to key underlying processes involved in this trait. Then, co-localization (or absence thereof) between QTLs for the trait and QTLs for model parameters can give new insights into the contribution of processes involved in the trait. Hence, it may help in the choice of candidate genes, or may give clues about the genomic regions to be combined in an ideotype. This approach is particularly well suited for studying plant adaptive responses to diverse environmental conditions (Prudent et al., 2011). It appears as a valuable tool to help make informed decisions with regard to genotypic adaptation options and ideotype design in the context of climate change (Ramirez-Villegas et al., 2015).Such an approach combining ecophysiological modeling and genetic analyses is still in infancy in the international grape community (Duchêne et al., 2012; Marguerit et al., 2012). We report here some of the pioneering work from the French LACCAVE research consortium on this topic. Models developed for plant drought response and berry sugar accumulation are outlined. These models consist of simple response curves for one trait or are able to simulate more complex physiological processes. Genetic parameters were defined and their variations among genotypes or segregating populations analyzed. The potential use of such models to simulate grapevine ideotype behavior under future climatic conditions is discussed. Gene-by-gene breeding approach remains elusive for complex traits in grapes Over the past century, conventional plant breeding has been used successfully to improve several crops. With the recent progress in molecular technologies for genome sequencing and functional genomics, genes have become tangible rather than virtual entities (Hammer et al., 2006). It is widely anticipated that a gene-by-gene approach will improve plant breeding efficiency. Indeed, there have been successes in developing plants that are more resistant to pests or tolerant to herbicides. Those cases involved single-gene transformations where plant phenotypic response scaled directly from the level of molecular action. However, this has not yet been extended to key complex traits where relationships among components and their genetic control involve quantitative multi-gene interactions (Tardieu, 2003).In grapes, up to now, few physiological functions have been clearly related to known gene sequences, and the tremendous progress in gene discovery has only weakly aided genetic sele