Lévy noise induced switch in the gene transcriptional regulatory system

Yong Xu, Jing Feng, Juanjuan Li, Huiqing ZhangChaos 2013 · DOI · arXiv

Abstract

The study of random fluctuations in a gene transcriptional regulatory system is extended to the case of non-Gaussian Lévy noise, which can describe unpredictable jump changes of the random environment. The stationary probability densities are given to explore the key roles of Lévy noise in a gene transcriptional regulatory system. The results demonstrate that the parameters of Lévy noise, including noise intensity, stability index, and skewness parameter, can induce switches between distinct gene-expression states. A further concern is the switching time (from the high concentration state to the low concentration one or from the low concentration state to the high concentration one), which is a random variable and often referred to as the mean first passage time. The effects of Lévy noise on expression and degradation time are studied by computing the mean first passage time in two directions and a number of different peculiarities of non-Gaussian Lévy noise compared with Gaussian noise are observed.

Introduction

In the complex world of molecular biology, stochastic fluctuations—commonly referred to as "noise"—are often viewed as a nuisance. However, researchers are increasingly recognizing that noise is a fundamental feature of genetic regulation. While traditional models typically assume that environmental fluctuations follow a Gaussian (Normal) distribution, real-world biological systems often encounter sudden, large-scale "jumps" in their environment, such as rapid changes in temperature, flashes of radiation, or nutrient availability.

Comparison between a Gaussian process (a) and a Lévy process (b), highlighting the 'jumpy' nature of the latter.

This study explores the impact of Lévy noise, a non-Gaussian stochastic process characterized by "heavy tails" and "Lévy flights." Unlike the smooth fluctuations of Gaussian noise, Lévy noise accounts for rare but significant perturbations that can trigger state-switching in gene regulatory systems more dynamically than classical models suggest.

Methods

The researchers focused on the $\lambda$-phage regulatory circuit in E. coli, a well-characterized biological system that is naturally bistable. This means the system can exist in two stable states: an "on" state (high protein concentration) or an "off" state (low protein concentration).

Schematic representation of the $\lambda$-phage regulatory circuit involving the cI and cro genes.

To analyze this system, the team used a nonlinear Langevin differential equation to model the concentration of the $\lambda$-repressor protein (CI), incorporating a Lévy stable motion term. This noise is defined by four key parameters: intensity ($D$), stability index ($\alpha$), skewness ($\beta$), and location ($\mu$). Due to the analytical complexity of the resulting Fractional Fokker-Planck Equation, the study utilized Monte Carlo simulations with $10^6$ trajectories to determine Stationary Probability Densities (SPDs) and calculate the Mean First Passage Time (MFPT)—the average time required for the system to switch between states.

Key Results

The findings demonstrate that Lévy noise parameters function as critical control variables for gene expression:

  • Noise Intensity ($D$): Increasing intensity generally pushes the system toward the high-concentration ("on") state, effectively acting as a positive feedback mechanism for protein production.
  • Stability Index ($\alpha$): This parameter controls the "jumpiness" of the noise. At low $\alpha$ (highly jumpy noise), the system tends to stay "on." As $\alpha$ increases toward the Gaussian limit of 2, the system more frequently shifts to the "off" state.
  • Skewness ($\beta$): Adjusting the asymmetry of the noise can induce a switch from "on" to "off," leading to a decline in protein concentration.
  • Switching Speed: Lévy noise significantly reduces the MFPT for degradation (switching from "on" to "off"). This suggests that jumpy environments can cause gene products to clear out much faster than expected under smooth Gaussian fluctuations. Additionally, the MFPT for the "off" to "on" transition shows a non-monotonic relationship with $\alpha$, indicating an optimal level of environmental jumpiness for maintaining a particular state.

Discussion and Conclusion

This research highlights that non-Gaussian Lévy noise can fundamentally alter the dynamics of gene transcriptional regulatory systems. By acting as a control parameter, noise can trigger or repress gene expression, providing a mechanism for cells to respond to unpredictable environmental shifts.

While these computational results are compelling, they are based on a simplified single-gene model. Future research should involve experimental verification in live cells and explore how Lévy noise interacts with more complex genetic networks involving multiple feedback loops and cross-talk between many genes.

Summary updated

Abstract

The study of random fluctuations in a gene transcriptional regulatory system is extended to the case of non-Gaussian Lévy noise, which can describe unpredictable jump changes of the random environment. The stationary probability densities are given to explore the key roles of Lévy noise in a gene transcriptional regulatory system. The results demonstrate that the parameters of Lévy noise, including noise intensity, stability index, and skewness parameter, can induce switches between distinct gene-expression states. A further concern is the switching time (from the high concentration state to the low concentration one or from the low concentration state to the high concentration one), which is a random variable and often referred to as the mean first passage time. The effects of Lévy noise on expression and degradation time are studied by computing the mean first passage time in two directions and a number of different peculiarities of non-Gaussian Lévy noise compared with Gaussian noise are observed.
BibTeX
@inproceedings{xu2013levy,
  author = {Yong Xu and Jing Feng and Juanjuan Li and Huiqing Zhang},
  title = {{Lévy noise induced switch in the gene transcriptional regulatory system}},
  booktitle = {Chaos},
  year = {2013},
  doi = {10.1063/1.4775758},
}
Figure