Phase separation as a tunable regulator of canonical gene regulatory motifs

Gene regulatory networks (GRNs) govern essential cellular processes, such as signal transduction, metabolism and cell fate control. The biophysical principles governing the dynamical behaviour of GRNs remain elusive. Recent studies highlight phase separation of transcription factors (TFs) and nucleic acids as a key organizing principle of intracellular biochemistry. In this work, we explore how phase separation of TFs influences the dynamics of two canonical GRNs, namely the toggle switch and the repressilator, using mean-field theory and stochastic simulations. Our mean-field analysis reveals that phase separation alters the stability of fixed points and reshapes the basin geometry of the toggle switch and modifies the oscillatory cycles of the repressilator. A key finding for both networks is that when multiple TFs undergo phase separation, the one with the lowest saturation concentration for phase separation dominates system dynamics. Interestingly, stochastic simulations show that the impact of phase separation on fluctuations (or noise) in the abundance of TFs for GRNs depends strongly on the topology of the circuits. This behaviour stands in sharp contrast with the noise reduction effect observed in the expression of isolated genes. Overall, our results show that biomolecular phase separation acts as a physical principle for tuning stability and noise in GRNs, providing new insights into cellular decision-making.