(277b) Bistable Gene Regulatory Networks: Intrinsic Versus Externally Induced Bistability

In this paper we bring light to the topic of bistability in gene regulatory networks by making the important distinction between intrinsically bistable systems which arise solely from the balance of gene expression between two cross-repressing operons and those systems which are actually monostable but become bistable in the presence of an external inducer molecule. This property of a bistable gene network can affect how it is suited for a specific function in nature and how synthetic bistable systems can be custom designed for a particular application in biotechnology. Finally we clarify the meaning of bistability in the context of stochasticity as this phenomenon blurs the distinction between monostable and intrinsically bistable phenotypes and influences the long term stability of externally induced bistable systems.

For our study we have constructed a simple gene network in Escherichia coli comprised of two genetic operons carried on artificial plasmids. One operon consists of a promoter that controls the expression of the tetracycline-inducible repressor protein (TetR) and the red fluorescent, reporter protein (Mcherry). The promoter is regulated by the lactose-inducible repressor protein (LacI). The other operon consists of a TetR-repressible promoter that controls expression of the LacI protein and the green fluorescent, reporter protein (GFPmut2). Acting together, the two operons produce a bimodal phenotype in which each population is dominated by a different operon. To control the distribution of phenotypes within the population we use externally added, inducer molecules, IPTG and aTc, capable of binding LacI and TetR, respectively. In the induced state the repressor molecules undergo an allosteric transition resulting in reduced affinity for their target operator regions, thus promoter repression is relieved.

A very similar LacI / TetR based device, named ?the genetic toggle switch? was first constructed in 2000 by Gardner et al.¹ Beyond their work we highlight the need for a more rigorous definition of genetic bistability: 1.) when external inducer molecules are used in the system and 2.) when the system is subject to stochastic fluctuations of relatively high probability. Through in vivo experiments and in silico stochastic simulations we demonstrate that this particular network configuration is actually capable of at least 4 different classes of behavior each exhibiting a different form of bistability. These classes include: intrinsic bistability, IPTG induced bistability, aTc induced bistability, and a fourth class in which cells are induced by either IPTG or aTc. This more extensive classification will add to the range of application for bistable gene networks.

By adjusting repressor half-life through ssrA degradation tags, we constructed 4 different system variants to explore the different classes of bistable behavior. In addition to results from flow cytometry showing the distribution of phenotypes for each system variant, we also present in silico model data to demonstrate the benefit of highly mechanistic stochastic kinetic models for modeling gene regulatory networks and generating useful predictions for molecular and synthetic biologists.

1. Gardner TS, Cantor CR, Collins JJ. 2000. Construction of a genetic toggle switch in Escherichia coli. Nature 403(6767):339-42.