Title:Emergent properties of collective gene expression patterns in multicellular systems

Abstract:Multicellular organisms are composed of a wide range of stable tissues containing different cell types. It remains unclear how a large number of interacting cells can precisely coordinate their gene expression during tissue self-organization. Here we develop a spin glass model of tissue gene expression which accounts for intracellular and intercellular gene regulation. The model provides a theoretical framework for tissue self-organization, with stable tissue configurations corresponding to local minima of the gene expression landscape. Intracellular gene regulation is defined using a Hopfield network, which enforces the stability of certain cell types in the absence of signaling. We represent multicellular tissue by a graph of interacting Hopfield networks. The edges of this graph represent cell-cell signaling which regulates the gene expression of adjacent cells. By tuning the signaling strength, the model exhibits a cascade of transitions between different self-organized tissue states. We specifically consider random cell-cell signaling networks and show that several properties of the distribution of stable tissues are preserved for different realizations. Counterintuitively, we find that random signaling networks tend to stabilize highly ordered tissue states. These results may explain why collections of cells in distinct organisms robustly self-organize into relatively simple tissues even when the microscopic interactions which drive the system are highly disordered.