Nonlinear Biochemical Signal Processing via Noise Propagation

TL;DR

This paper introduces a quantitative method to analyze how biochemical noise propagates through nonlinear cellular networks, enabling the design of novel signal processing modules that leverage stochastic fluctuations for enhanced functionality.

Contribution

It develops a new approach to understand and engineer biochemical signal processing by analyzing noise propagation and nonlinearities in cellular networks.

Findings

Noise can enhance sensitivities in certain regions and reduce them in others.

Designed three modules: a concentration detector, a bistable switch, and a linear amplifier.

Method allows for understanding and engineering of noise-induced phenotypes.

Abstract

Single-cell studies often show significant phenotypic variability due to the stochastic nature of intra-cellular biochemical reactions. When the numbers of molecules, e.g., transcription factors and regulatory enzymes, are in low abundance, fluctuations in biochemical activities become significant and such "noise" can propagate through regulatory cascades in terms of biochemical reaction networks. Here we develop an intuitive, yet fully quantitative method for analyzing how noise affects cellular phenotypes based on identifying a system's nonlinearities and noise propagations. We observe that such noise can simultaneously enhance sensitivities in one behavioral region while reducing sensitivities in another. Employing this novel phenomenon we designed three biochemical signal processing modules: (a) A gene regulatory network that acts as a concentration detector with both enhanced amplitude and sensitivity. (b) A non-cooperative positive feedback system, with a graded dose-response in the deterministic case, that serves as a bistable switch due to noise-induced bimodality. (c) A noise-induced linear amplifier for gene regulation that requires no feedback. The methods developed in the present work allow one to understand and engineer nonlinear biochemical signal processors based on fluctuation-induced phenotypes.