Genetic and Metabolic Marker-Based Prediction of Heterotic Traits in Arabidopsis and Maize

Heterosis is used intensely in hybrid breeding and has boosted productivity of many crop species. Despite this widespread use and intense research, the molecular mechanisms causing heterosis are still largely unresolved. To gain deeper knowledge of molecular principles, hybrid vigor is analysed in the C24 x Col-0 cross of Arabidopsis accessions that exhibits strong mid-parent heterosis (MPH) of vegetative growth. Heterosis-associated gene expression and metabolite composition has been analysed in parents and hybrids and 10 biomass-heterosis QTL and 232 heterotic metabolite QTL were detected using 429 recombinant inbred lines, 140 introgression lines, and test crosses thereof. A prominent QTL (chr. 4) fine mapped to a region of 14 genes is subjected to loss and gain of function analyses of candidate genes. Complex correlation analysis revealed a close link between plant biomass and a specific metabolic composition and the approach was extended to prediction of biomass heterosis. Integration of SNP and metabolite information revealed highest predictive power. To explore the potential of hybrid performance prediction for hybrid breeding, maize inbred panels and corresponding test cross hybrids are subjected to field and glasshouse phenotyping, genotyping, and molecular profiling analyses. As a first step, 285 diverse Dent inbred lines crossed to two Flint testers were used to predict combining abilities for several complex heterotic traits including dry matter yield by building models fitting SNP and metabolite effects. Observed prediction accuracies ranging from 0.71 to 0.81 for SNPs and from 0.60 to 0.80 for metabolites warrant further investigation of genetic and bio marker-based hybrid performance prediction.