A Systems Biology Route Towards Understanding Temperature-Sensitive Subgenome Interplay in Triticum aestivum Endosperm

Due to the economic and agronomic importance of hexaploid bread wheat, characterizing gene expression and the corresponding regulatory patterns is of particular importance. The urge to decipher the underlying transcriptional mechanisms is increased by the predicted yield losses due to climate changes. In the present work we show how varying temperature regimes alter the interplay of the A,B, and D subgenome in three different endosperm cell types. Already in previous work, focused on the developing endosperm, we found clear signs of cell-type and time-dependent biases towards one of the subgenomes, without any of the three subgenomes dominating on a global scale (Pfeifer, Kugler et al. Science 2014). One backbone of our analyses is a set of 8605 homoleogous triplets, defined by genes having a 1:1:1 copy relationship between the A, B, and D subgenome. This set, comprising 25,815 genes, allows us to project gene-specific expression biases on a genome-wide representation. As inspecting expression patterns of single genes fails to depict the molecular dynamics, we apply a network-based approach for extracting groups of genes sharing similar expression patterns (modules) in order to capture and visualize the underlying transcriptional events. By using an information-theoretic measure we find that while for most genes all three subgenomes contribute equally, there exist several temperature sensitive network modules with cell-type-specific expression that are dominated by only one or two subgenomes. Our results will enable future studies to inspect how contributions from single subgenomes can be used to adapt cereals for environments with changing temperatures.