Towards in vivo computing

Abstract

Artificial regulatory networks (AGRNs) are instrumental in elucidating basic principles that govern the dynamics and consequences of stochasticity in the expression of naturally occurring gene regulatory networks. In contrast to state of the art computer engineering circuits, these AGRNs are evolutionarily highly optimized and fault tolerant. We draw motivation from the fact that non-deterministic polynomial-time (NP) and NP-hard computational tasks can not be solved using conventional computing techniques. This study is a stepping stone towards solving problems such as traveling salesman problem (TSP) in a time bound fashion using interconnection-free bio-computing devices. In this in vivo study we quantitatively show that a reporter encoding the green fluorescent protein (GFP) can be switched from high to low expression states and vice versa, thus mimicking a RS flip-flop. This was accomplished by using the bistable, transgenic AGRN incorporating the N-acyl homoserine lactone (AHL) sensing lux operon from Vibrio fischeri along with a toggle switch in Escherichia coli, developed by Collins et al. (2004). The inducers and temperature act as inputs to the AGRN. GFP expression was quantified using flow cytometry. The plug and play property was demonstrated by showing that any output could be expressed based on a similar logic. The software model, previously proposed by Collins et al. (2004) was extended and analyzed by incorporating the function of the inducer isopropyl-beta-D-thiogalactopyranoside (IPTG) and temperature. We also demonstrate that such a system is robust and fault tolerant