A genetic differential amplifier: design, simulation, construction, and testing

S Nagaraj,S Davies

doi:10.1109/bmn.2006.330883

Abstract

Summary form only given. Differential amplifiers are devices that produce an output that is proportional to the difference between the two inputs. In electronics, they are useful in comparison/thresholding, feedback amplifiers and oscillators. A genetic analog to this circuit is equally useful in the construction of artificial gene networks. A genetic differential amplifier was developed using components from bacteriophage lambda. The activator protein CI was used as the non-inverting (positive) input. The repressor protein Cro was used as the inverting (negative) input. A mutated P RM promoter formed the core of the amplifier. The gene for enhanced green fluorescent protein was placed after the P RM promoter so that the output could be monitored. The mutations in P RM allow the representation of negative differences and correct an undesired feature of wild-type P RM . Negative differences are represented by a reduction in output from a baseline level. Wild-type P RM has a low basal level of activity. The first mutation added a bias to correct this problem. Wild-type P RM is also repressed by high concentrations of CI. The second mutation removed the repression, making CI strictly an activator. The circuit was modeled using a statistical thermodynamic and stochastic approach. Different variants of the circuit were constructed. The mutated PRM was implemented on both low and high copy number plasmids. The inputs, CI and Cro, were placed on a separate plasmid under the control of P tet and P lac , respectively, allowing control by aTc and IPTG. Both the low copy number and the high copy number variants of the circuit were tested over a range of aTc and IPTG concentrations. Output fluorescence levels were measured using a microplate fluorometer. Measurements of the low copy number variant showed that the output increased in response to CI and decreased in response to Cro, as expected. The circuit did not respond to very low input protein concentrations, and the output was clipped after a maximum concentration was reached. The high copy number variant showed similar results. This variant was also tested with the two inputs applied together. The output of the circuit reflected the difference between the two inputs, although the response was not linear. At very high levels of Cro, the circuit essentially shut down, regardless of CI concentration. The experimental results are consistent with the model predictions

Full Text