Control of genes by self-organizing multicellular interaction networks

Abstract

Multicellular self-organization drives development in biological organisms, yet a comprehensive theory is lacking as basic properties of cells can complicate common approaches. Framing such properties by dynamic graphs led to new theoretical propositions for multicellular self-organization in Escherichia coli. Here, corresponding ideas are developed from biologically-general first principles. The resulting perspective could aid both experimental and computational approaches to multicellular biology as well as efforts to control and engineer it.

Figures (3)

• Figure 1: Dynamic graphs for multicellular self-organization. Cells divide (top left). Cells can adapt their identity and behavior by regulating genes using gene networks that interpret molecular signals, including inputs from their external environment (top right). Proximal cells can form chemical and mechanical interactions and can modify these interactions by adaptation (top right). Self-organization is determined by the specific interactions among a system’s fundamental units. Cells are the fundamental units of multicellular self-organization. Dynamic graph models, which represent the units and interactions of dynamically changing systems, are a logical framework for multicellular self-organization (and more generally, for any form of self-organization). In this framework, cells are represented as nodes and interactions as edges (bottom). As cells divide, adapt, and interact, new nodes and edges are added to a propagating dynamic graph. Clonality (where all cells descend from a single progenitor cell) simplifies propagation: all changes to nodes and edges result from division, adaptation, and interaction events on preceding graph stages, rather than from unpredictable external effects.
• Figure 2: Multicellular interaction networks. Interactions (edges) between cells (nodes) dynamically change and are distributed over populations of self-organizing cells. The set of such interactions in the dynamic-graph framework represents a multicellular interaction network. The individual edges of such networks can be directed, weighted, combinatorial, and/or diverse (or not). Many mechanical and some chemical interactions between cells can be encoded by such pairwise edges. However, some multicellular interactions, like those mediated by diffusible molecules (e.g., morphogens, metabolites, and so on), simultaneously connect many cells. These interactions would be better encoded as hyperedges which simultaneously connect many nodes. Hyperedges were developed to address the pair-wise limitation of classical graph theory for complex systems.
• Figure 3: Developmental daisy chains. As groups of cells propagate by clonal multicellular self-organization (represented as the transition from the graph in the upper left to the one in bottom right), each cell experiences specific combinations of interactions (inset, left) deriving from the communities’ overall multicellular interaction network (Fig. 2). These interactions can serve as inputs for each cell’s gene networks (inset, middle) via multicellular control, as articulated. The outputs of these gene networks drive adaptation, including cell-fate and -differentiation decisions. Moreover, these outputs can modify the cell’s local multicellular interaction network (inset, right). In clonal populations, such modifications are predictable because they were generated by progenitors and their neighbors. Modified interactions feedback iteratively as dynamic and distributed inputs for subsequent rounds of multicellular control. Such predictive and iterative feedback of gene networks and multicellular interactions networks in clonal self-organization has been described as multicellular daisy chaining. The resulting multicellular- or developmental daisy chains provide an underlying mechanism for clonal multicellular self-organization by linking the dynamics of cell division, adaptation, and interaction. Moreover, they model how clonal self-organizing populations can robustly and adaptively control the fate, differentiation, and morphogenesis of each of their cells.