Optogenetic Frequency Analysis of Bacterial Two-Component Systems and Synthetic Gene Circuits

Synthetic biology is not yet a mature engineering discipline, due in part to a poor understanding of the dynamical input/output (I/O) properties of genetic components. Despite the fact that gene regulatory systems are inherently dynamic, most attempts to characterize them have been static. We have previously engineered green/red (CcaS/R) and red/far red (Cph8/OmpR) light-switchable two component systems (TCSs) with transcriptional outputs in E. coli^1,2. In recent work, we developed quantitatively predictive mathematical models of the I/O dynamics of these signaling pathways by measuring their gene expression outputs in response to step changes in light input³. However, this characterization relied upon intuition to design non-linear models and iterative characterization and validation experiments. In contrast, the electrical engineering community has developed standardized, automatable, and scalable system identification (SI) methods that employ frequency analysis to construct linear transfer function (TF) models. These dynamical TF models are used to systematically characterize the performance of engineered components, reliably predict the behavior of those components when they are composed with others, and thereby design new systems with high predictability.

Here, we have combined frequency analysis, optogenetics, light-switchable TCSs and synthetic gene circuits to develop a standardized I/O characterization framework for transcriptional regulatory parts. Specifically, we subject E. coli populations to sinusoidally oscillating light inputs spanning a wide range of frequencies, and monitor the corresponding gene expression outputs with high temporal resolution. We then fit the sinusoidal response data to a small set of TF models, selecting one with the best fidelity for each system. We have validated this approach by characterizing both light-switchable TCSs and using the resulting TF models to predict gene expression dynamics in response to non-sinusoidal, time-varying inputs, with only 5% error (RMSE) over 12-hour experiments.  This result demonstrates that frequency analysis can supplant our previous intuition-guided approach. Additionally, our TF models have yielded a better understanding of the TCSs’ filtering characteristics, noise, and signal transduction limits, motivating characterization of other optogenetic components. We are also extending optogenetic frequency analysis to a wide range of genetic circuit components, including the Voigt TetR inverter family⁴ and large sets of CRISPRi-based inverters, and using the data to create technical data sheets. Finally, we will use the resulting TF models to design larger genetic circuits with predictable dynamic properties. Our standard, scalable, and quantitatively rigorous approach will help synthetic biology transition into a mature engineering discipline.

[1] Tabor JJ, Levskaya A, & Voigt CA (2011). Multichromatic control of gene expression in Escherichia coli. Journal of Molecular Biology, 405(2), 315–24. doi:10.1016/j.jmb.2010.10.038

[2] Schmidl SR, Sheth RU, Wu A, & Tabor JJ (2014). Refactoring and Optimization of Light-Switchable Escherichia coli Two-Component Systems. ACS Synthetic Biology, 3(11), 820–831. doi:10.1021/sb500273n

[3] Olson EJ, Hartsough LA, Landry BP, Shroff R & Tabor JJ (2014). Characterizing bacterial gene circuit dynamics with optically programmed gene expression signals. Nature Methods, 1–11. doi:10.1038/nmeth.2884

[4] Stanton BC, Nielsen AK, Tamsir A, Clancy K, Peterson T, & Voigt CA (2014). Genomic mining of prokaryotic repressors for orthogonal logic gates. Nature Chemical Biology, 10(2), 99–105. doi:10.1038/nchembio.1411