Heterosis has contributed greatly to yield in maize, but the nature of its contribution is not completely clear. In this study, two strategies using whole-genome oligonucleotide microarrays were employed to identify differentially expressed genes (DEGs) associated with heterosis and yield. The analysis revealed 1 838 heterosis-associated genes (HAGs), 265 yieldassociated genes (YAGs), and 85 yield heterosis-associated genes (YHAGs). 37.1% of HAGs and 22.4% of YHAGs expressed additively. The remaining genes expressed non-additively, including those with high/low-parent dominance and over/under dominance, which were prevalent in this research. Pathway enrichment analysis and quantitative trait locus (QTL) co-mapping demonstrated that the metabolic pathways for energy and carbohydrates were the two main enriched pathways influencing heterosis and yield. Therefore, the DEGs participating in energy and carbohydrate metabolism were considered to contribute to heterosis and yield significantly. The investigation of potential groups of HAGs, YAGs, and YHAGs might provide valuable information for exploiting heterosis to improve yield in maize breeding. In addition, our results support the view that heterosis is contributed by multiple, complex molecular mechanisms.

Heterosis plays an important role in crop production and plant evolution. Although heterosis has been widely exploited by plant breeders, the underlying molecular mechanisms are not well understood. We analyzed gene expression of the highly heterotic maize hybrid Zhengdan 958 and its parents, Zheng 58 and Chang 7-2 during spikelet and floscule differentiation using the GeneChip® Maize Genome Array. Pairwise comparison among Zhengdan 958 and its parents at the two stages of immature ear development identfied 1 089 and 1 352 differentially expressed genes. Gene ontology (GO) functional analysis showed that these genes participate in many functional categories, and those encoding response to stress and transcription factor may play important roles in heterosis. Pathway analysis showed that the differentially expressed genes are involved in various metabolic processes, and those participating in lipid metabolism, signal transduction, transport, and catabolism may contribute to heterosis. A non-additive expression pattern was prevalent in genes that were differentially expressed between the hybrid and its parents during both spikelet and floscule differentiation. Because genes that are differentially expressed in a hybrid and its parents could underlie heterosis, nonadditive expression patterns might contribute to the manifestation of heterosis.