Feedback between stochastic gene networks and population dynamics enables cellular decision-making

Area of Science:

• Systems Biology
• Cellular Dynamics
• Quantitative Biology

Background:

• Phenotypic selection arises from differing reproductive success in genetically identical cells, driven by cellular noise.
• Cellular noise, originating from protein synthesis fluctuations, is vital for cell decision-making, stress responses, and drug resistance.

Purpose of the Study:

• To develop a general stochastic agent-based model for growing populations that links gene expression and cell division dynamics.
• To devise a finite state projection approach for analyzing gene expression and division distributions.
• To infer selection from single-cell data and quantify selection in multi-stable gene expression networks.

Main Methods:

• Stochastic agent-based modeling of growing cell populations.
• Finite state projection for analyzing gene expression and division distributions.
• Inference of selection from single-cell data using mother machine and lineage tree experiments.

Main Results:

• Quantified selection in multi-stable gene expression networks.
• Demonstrated that the trade-off between phenotypic switching and selection enables robust decision-making.
• Provided quantitative insights into bet-hedging responses to DNA damage and antibiotic adaptation in *Escherichia coli*.

Conclusions:

• The developed theory and inference methods offer quantitative insights into cellular decision-making and adaptation.
• The trade-off between phenotypic switching and selection is essential for robust synthetic circuits and developmental processes.
• This approach elucidates bet-hedging strategies in response to environmental challenges like DNA damage and antibiotic stress.