Engineering prokaryotic gene circuits

Abstract

Engineering of synthetic gene circuits is a rapidly growing discipline, currently dominated by prokaryotic transcription networks, which can be easily rearranged or rewired to give different output behaviours. In this review, we examine both a rational and a combinatorial design of such networks and discuss progress on using in vitro evolution techniques to obtain functional systems. Moving beyond pure transcription networks, more and more networks are being implemented at the level of RNA, taking advantage of mechanisms of translational control and aptamer-small molecule complex formation. Unlike gene expression systems, metabolic components are generally not as interconnectable in any combination, and so engineering of metabolic circuits is a particularly challenging field. Nonetheless, metabolic engineering has immense potential to provide useful biosynthesis tools for biotechnology applications. Finally, although prokaryotes are mostly studied as single cell systems, cell-cell communication networks are now being developed that result in spatial pattern formation in multicellular prokaryote colonies. This represents a crossover with multicellular organisms, showing that prokaryotic systems have the potential to tackle questions traditionally associated with developmental biology. Overall, the current advances in synthetic gene synthesis, ultra-high-throughput DNA sequencing and computation are synergizing to drive synthetic gene network design at an unprecedented pace.

Fig. 1

Transcriptional gene circuits. Pointed arrows denote activation, blunt-end arrows repression or inhibition. (a) Atkinson's genetic clock-toggle switch. The checkered promoter upstream of lacI can be either the NRI-P responsive glnK promoter (dashed arrow), resulting in a genetic clock, or a constitutive promoter, resulting in a toggle switch (Atkinson et al., 2003). (b) A synthetic system for studying different modes of regulation. Box I provides a repression element, Box II provides activation and the dotted cI gene provides positive feedback. Using a varying number of these elements, different modes of regulation were studied and used to predict the behaviour of a circuit comprising of all three elements (Guido et al., 2006).

Fig. 2

Circuit ‘debugging’ by directed evolution. A nonfunctional circuit is restored using a library of mutant cI genes (black cylinder) and screening for functionality (Yokobayashi et al., 2002).

Fig. 3

The ‘metabolator’ (Fung et al., 2005) couples metabolism and transcription to generate oscillations. Block arrows denote fluxes, pointed arrows denote activation and blunt-end arrows shows repression. Metabolism leads to accumulation of acetyl-CoA, which, through AcP and NRI, leads to expression of destabilized Acs and repression of destabilized Pta and GFP. As Pta is being degraded, the flux towards AcP decreases until it is insufficient to provide suitable levels of NRI-P for transcription from the glnAp2 promoter. This results in downregulation of Acs and expression of Pta and GFP, closing the oscillatory cycle.

Fig. 4

(a) The BD circuit (Basu et al., 2005). The output of the circuit is a FP that is expressed only within a certain range of input (AHL) concentrations. (b) Using a mixture of cells containing a BD circuit with green fluorescent output and cells containing a BD circuit with a red fluorescent output and lower LuxR levels, a bullseye pattern can be formed around a population of ‘sender’ cells (white circle; S) secreting AHL.