Modeling the "microbial chassis effect" on the performance of a genetic switch

Abstract

Escherichia coli is one of the most established bacterial hosts for genetic devices, partially due to the available knowledge and tools for ease of manipulation. However, there is an incentive to increase the current number of available recipients. For instance, marine bacteria are being recognized for their potential as microbial "chassis" due to their rich genetic and metabolic diversity. With this in mind, tools for simulating the behavior of genetic devices would help synthetic biology expand its reach to untraditional hosts. Hence, here in this study, the effect of recipient on the performance of a genetic switch, the "chassis" effect, was estimated across different bacteria with an aim to indicate the possibilities of marine microorganisms as hosts. The device considered in this study was assembled from two sub-parts 1) L-arabinose-inducible PBAD promoter that expresses tetR and gfp genes encoding production of the TetR repressor and GFP fluorescent protein; and 2) anhydrotetracycline (aTc)-inducible PTet promoter that controls araC and mKate expression, which codes for the AraC repressor and mKate fluorescent protein. AraC protein is a repressor of the PBAD promoter, while TeR represses PTet. The dynamic behavior and stability of this device was simulated by a mathematical model based on a system of ordinary differential equations (ODEs) that predicted possible "chassis" effect and compared its strength across selected bacterial hosts. To further our understanding of the performance of a genetic switch, a dynamic modeling framework was established, and a behavior was simulated for a set of marine bacteria and E. coli. This was done by building a mathematical model that included system of already parametrized non-linear ODEs which were solved using the R programming language. The parametrization of ODEs by a non-linear model resulted in the Hill (n) and activation (K) coefficient estimates. The non-linear regression was performed on a GFP fluorescence data collected from the induction study with E. coli. This assay estimated GFP-, GFP/OD600 signals and GFP rates from the cells induced with L-arabinose. The simulated dynamic response was quantified by a response time, a limiting factor for designing efficient gene circuits. The simulation estimated the fastest response of Vibrio natriegens and the slowest of Pseudomonas oceani. This outcome has indicated high potentials of V. natriegens for future applications in the synthetic biology. The "chassis" effect predicted by the model was estimated as a direct consequence of the specific growth rate.