Hybrid vigor in agriculture is defined as the increase in yield of an offspring over those of its inbred parents. Hybrid vigor is a critical component of the high productivity of many crops, including maize. While hybrid vigor has been efficiently employed in many crops, we lack a basic understanding of how it works. A better understanding could provide novel avenues for the improvement of hybrid vigor, and for extending its benefits to non-hybrid crops. This project will develop new approaches for discovering sets of interacting genes from the inbred parents. Knowing the identity of those genes will make it possible to use them as predictive tools and as targets of manipulation for crop improvement. The project will provide a model for using predictive gene network-based approaches to understand and improve complex traits in crops. Interest in using biological networks for crop research is growing, but access to the necessary technical capabilities remains limited. Therefore, the project will also provide plant scientists with free technical and analytical services with an emphasis on crop genomics-enabled research. These services will range from consultation on experimental design to data interpretation, ensuring that results are useful and capacity is not wasted. Training investigators who wish to become plant biological network experts will be the project's highest priority. Two to three undergraduate students will provide the services. These students will be trained in chemistry, biochemistry, computer science, and engineering so that they can support users in sample preparation, data generation, and network analysis.
This project will create unsupervised gene regulatory networks (GRNs) and protein kinase networks (PKNs) for two maize inbreds (B73, Mo17) and their hybrid (SX19). GRN modules are comprised of a transcription factor (TF) and its target genes. PKN modules are comprised of a protein kinase and its substrates. The GRN s will be made using three different proxies for TF activity: TF mRNA abundance, TF protein abundance, and level of TF protein phosphorylation; network targets will be measured by the levels of all mRNAs. The PKN will be made using phosphorylation of the activation-loop as a proxy for kinase activity; network targets will be measured by protein phosphorylation levels of all proteins. All measures will be made on 26 different tissues. GRNs will be constructed using the GENIE3 random forest algorithm. The PKN will be made using our previously described correlative method. We will measure the preservation and divergence of modules between the inbreds and their hybrid. We will test whether inbred-specific or hybrid-specific modules contribute to heterosis. Selected near isogenic lines (NILs) derived from SX19 with introgressed segments containing module regulators will be test-crossed to the donor parent. The Fl will be exactly like SX19 except for the segment containing the regulator, which will be homozygous. Comparison of the transcriptome and proteome of SX19 to the test-crossed NIL will reveal whether the regulator is acting on its targets as predicted by the network. If the module contributes significantly to heterosis then hybrid vigor will be reduced in the test-crossed NIL.
