The Noise On The Signal Processing Of The Gene Regulatory Network Motif

Genetic regulatory network is one very important sort of complex network. It lies at the heart of the signal processing functionality in real lives. A major challenge in the post-genomic era is to quantitatively understand topological and dynamical properties and functions of these complex biological networks that control cellular functions. However, the complexity in both structures and functions of these networks inhibits us to investigate them as a whole. Recently, scientists find that complex genetic regulatory network is actually consisted of simple building blocks-Motif. Many motifs in real genetic network can play specific signal processing roles. Therefore, the research of the functions of large scale genetic network can start from investigating the function of motifs.In most of the existed research on signal processing roles of genetic regulatory network motifs, deterministic differential equation model is widely used. However, in such model, researchers ignore the important fact that the signal processing in motifs is constantly perturbed by noise. Such noise originates from two very different ways: extrinsic noise coming from the perturbation of key parameters in the system and intrinsic noise coming from the stochastic nature of chemical reactions.In this paper, we present a stochastic model for the mixed-feedback loop (MFL), a motif found in both integrated cellular networks of transcription regulation and protein-protein interaction, and computer-simulated genetic networks. Previous bifurcation analysis indicates that this motif can serve as a bistable switch or a clock. We investigate how extrinsic and intrinsic noise affects its signal processing roles systematically.We find that this motif can exploit noise to enrich its dynamical performance. When the MFL is in the bistable region, under fluctuation of extrinsic noise, the MFL system can switch from one steady state to the other and meanwhile one protein's production is amplified for more than three orders of magnitude. Further, from an engineering perspective, this noise-based switch and amplifier for gene expression is very easy to control. Without extrinsic noise, spontaneous transition between states occurs as the consequence of intrinsic noise. Such a switch is controlled by the parameters and system size.On the other hand, intrinsic noise can induce sustained stochastic oscillation when the corresponding deterministic system does not oscillate. Such stochastic oscillation shows the best performance at an optimal noise level, indicating the occurrence of intrinsic noise stochastic resonance which can contribute to the robustness of this oscillator. When considering the effects of extrinsic noise near bifurcation points, a similar phenomenon of extrinsic noise stochastic resonance is unveiled.