Highly designable phenotypes and mutational buffers emerge from a systematic mapping between network topology and dynamic output

Abstract

Deciphering the design principles for regulatory networks is fundamental to an understanding of biological systems. We have explored the mapping from the space of network topologies to the space of dynamical phenotypes for small networks. Using exhaustive enumeration of a simple model of three-and four-node networks, we demonstrate that certain dynamical phenotypes can be generated by an atypically broad spectrum of network topologies. Such dynamical outputs are highly designable, much like certain protein structures can be designed by an unusually broad spectrum of sequences. The network topologies that encode a highly designable dynamical phenotype possess two classes of connections: a fully conserved core of dedicated connections that encodes the stable dynamical phenotype and a partially conserved set of variable connections that controls the transient dynamical flow. By comparing the topologies and dynamics of the three-and four-node network ensembles, we observe a large number of instances of the phenomenon of ''mutational buffering,'' whereby addition of a fourth node suppresses phenotypic variation amongst a set of three-node networks.designability ͉ dynamical phenotype ͉ enumeration ͉ mutational buffering ͉ regulatory network D iscerning the structure and function of cellular networks is essential to the development of a true understanding of biological systems. Experimental and theoretical studies are steadily advancing our knowledge of the wiring and input-output characteristics of a variety of natural and designed biological networks. These efforts have focused on characterizing the components and interactions for specific biological networks and their dynamical behaviors, by using standard genetic and biochemical approaches in conjunction with mathematical analysis of discovered circuits (1-7). Valuable insights into certain design features of biological networks have emerged through these efforts (8-13) and are used to guide the design of synthetic systems (14-16). The choice of which synthetic circuits to build is often a matter of careful hand-picking guided by experimental restrictions.Reverse-engineering and modeling of specific experimental systems on a case-by-case basis is necessary and meritorious. However, the space of networks and associated dynamics is potentially very large, and parallel approaches that consider broad ensembles of networks may advance our understanding of general design principles in ways that the serial strategies may have difficulty revealing. Therefore, we have chosen to explore general design principles by using a global strategy. We expect that exploration of an entire ensemble of networks and associated dynamics will reveal statistical signatures connecting network architectures to categories of dynamical phenotypes. In particular, we analyze the relationship between the space of topology and the space of dynamics by employing an analogy to the protein ''designability principle,'' which states that compact protein structures that can be encoded by a wide array of ...

Cancer as robust intrinsic state of endogenous molecular-cellular network shaped by evolution

“…Even after mutations in its genome, it may still stay in the lysogenic state for many generations [38,50]. Thus, the existence of functional landscape with multiple stable states explains the robustness widely observed in biological systems [16,21,[51][52][53][54][55]. The existence of such multiple stable states may also corroborate well with the observations that cancer takes multiple steps [56]: If regarding moving from one stable state to another as one step.…”

Process-based network decomposition reveals backbone motif structure

“…Together, they reveal a number of valuable insights related to regulating the phases of the cell cycle. The first is that many of the motifs involve nodes [5,8,9,10 in budding yeast (16)(17)(18)(19)(20) and 2, 3, 4, 6 in fission (21)(22)(23)(24)(25)] known to be master regulators. This is not surprising, but it is a confirmation that the type of analysis presented here correlates with what is known by biologists.…”

Evolution of Transcription Networks — Lessons from Yeasts

“…At the network level, some regulatory tasks can be accomplished by different network architectures/topologies [47–49]. As a simple example, condition-specific expression can be achieved either by an activator or a repressor [31].…”