A coarse-grained bacterial cell model for resource-aware analysis and design of synthetic gene circuits

Abstract

Synthetic genes compete among themselves and with the host cell's genes for expression machinery, exhibiting resource couplings that affect the dynamics of cellular processes. The modeling of such couplings can be facilitated by simplifying the kinetics of resource-substrate binding. Model-guided design allows to counter unwanted indirect interactions by using biomolecular controllers or tuning the biocircuit's parameters. However, resource-aware biocircuit design in bacteria is complicated by the interdependence of resource availability and cell growth rate, which significantly affects biocircuit performance. This phenomenon can be captured by coarse-grained models of the whole bacterial cell. The level of detail in these models must balance accurate representation of metabolic regulation against model simplicity and interpretability. We propose a coarse-grainedE. colicell model that combines the ease of simplified resource coupling analysis with the appreciation of bacterial growth regulation mechanisms. Reliably capturing known growth phenomena, it enables numerical prototyping of biocircuits and derivation of analytical relations which can guide the design process. By reproducing several distinct empirical laws observed in prior studies, our model provides a unifying framework for previously disjoint experimental observations. Finally, we propose a novel biomolecular controller that achieves near-perfect adaptation of cell-wide ribosome availability to changes in synthetic gene expression. Showcasing our model's usefulness, we use it to determine the controller's setpoint and operation range from its constituent genes' parameters.

A coarse-grained bacterial cell model for resource-aware analysis and design of synthetic gene circuits

“…( 1) and ( 2) for each heterologous gene to describe an arbitrary synthetic gene circuit. Resource-and host-aware simulations of a circuit's behavior can then be performed using our Matlab implementation of the model found at https://github.com/KSechkar/ rc_e_coli 63 , enabling rapid and cheap prototyping of resource-aware biomolecular controllers in silico 64 . Besides numerical analyses, our framework's simplicity allows to obtain a range of analytical relations between a synthetic circuit's design parameters and the host cell's growth rate, reproducing the empirical relations commonly used to capture burden in synthetic gene expression models 6,[17][18][19] .…”

“…Our Matlab model implementation, along with all other scripts and data used to obtain the results described here, can be found at https:// github.com/KSechkar/rc_e_coli 63 . The deterministic simulations were run using Matlab R2022a's ode15s solver on a Dell OptiPlex 7000 PC with a 2.10 GHz 12th Gen Intel(R) Core(TM) i7-12700 processor and 16 GB RAM, running on Windows 11.…”

“…The parameters' starting values and other simulation details, as well as the details of parameter sensitivity analysis based on the fitting's outcome 87,88 , can be found in Supplementary Note S2.2. The parameter fitting script is provided in this manuscript's GitHub repository 63 .…”

“…All code used to implement the model and obtain the figures is provided in the GitHub repository https://github.com/KSechkar/rc_e_ coli 63…”