Six-gene model underlying heterosis for culm length and flowering date in sorghum (Sorghum bicolor (L.) Moench).

Heterosis has been widely used to increase crop yield. Four genetic heterosis models have been proposed previously: dominance, overdominance, epistasis, and pseudo-overdominance models. In a previous study, the remarkable heterosis for culm length (CL) of the F1 variety 'Tentaka' (MS79 × 74LH3213) of high biomass sorghum (Sorghum bicolor (L.) Moench) can be explained based on the dominance model involving five genes. However, whether other genes or epistatic interactions contribute to the heterosis remains unclear. Here, we clarified the two points and propose an updated model of the heterosis. We conducted quantitative trait locus (QTL) and epistasis analyses for CL and flowering date by employing a joint F2 population, consisting of each F2 population of the progeny of MS79 × 74LH3213 cultivated for over 4 years. In addition to the five known genes, a sixth QTL was indicated on chromosome 6. The corresponding candidate gene was Ma1/SbPRR37, which was linked with Ma6/SbGhd7, one of the five genes. The genes showed a repulsive phase in the pericentromeric region, and pseudo-overdominance was observed. Although five QTL epistatic interactions were detected, their genetic effects were weaker than those of most single QTLs. In conclusion, as all the dominant alleles of the six loci were major elements for the heterosis, it can be explained mainly based on the dominance model and partially by the epistasis model. The present study provides new insights into the genetic basis of the heterosis and could serve as a model for understanding heterosis in other crops.