Optogenetic Regulation of Proteolysis Using an Engineered Light-Switchable Split-Tev Protease

Cells use proteases and proteolytic cleavage to control diverse processes which include protein activation or inactivation by removal of inhibitory or catalytic domains, mediation of protein degradation by revealing stabilizing or destabilizing residues, and initiation of protein translocation by removal of peptide signal sequences or exposure of latent ones. Due to the potential for regulating such a diverse array of cellular processes, proteases and proteolytic cleavage are excellent candidates for synthetic control.

Here, we have engineered a synthetic light-switchable protease by fusing split halves of the NIa tobacco etch virus protease (TEVp) to the light sensing proteins Phytochrome B (PhyB) and Phytochrome Interacting Factor 6 (PIF6). PhyB and PIF6 form high affinity heterodimers under red light but rapidly dissociate under far-red light, making them ideal candidates for reconstituting otherwise inactive split-TEV fragments¹. By coupling the catalytic activity of TEVp to the light-dependent activity of PhyB and PIF6 we are able to rapidly, photoreversibly, and quantitatively tune TEVp activity using time-varying ratios of red and far-red light. This result suggests that split TEV fragments rapidly become functional and non-functional upon dimerization and dissociation of PhyB and PIF6. We show that this light-switchable protease can be used to cleave and modify target proteins in-vivo, and that light-gated proteolytic cleavage can be engineered to initiate target protein degradation through the conserved N-end rule pathway in the model eukaryotic organisms Saccharomyces cerevisiae and Dictyostelium discoideum.

Using a custom multichromatic LED-tissue culture plate device, we have characterized the spectral, steady-state, and dynamic performance of the photoswitchable TEVp protein degradation system. We measured action spectra (activating and deactivating) and determined peak activating and deactivating wavelengths to be 648 nm and 721 nm, respectively, which correlate closely with in-vitro measured PhyB P_r/P_fr absorbance peaks spectra (660/730 nm). The steady-state dose response to 648 nm light intensity follows a Hill-like relationship with fold-change = 4.5, n = 1.39, k = 0.02 µmol/m²*s. Step-function response dynamics, τ_ON,1/2 = 150 min, are fast compared to gene expression dynamics (cell-cycle time scales, 12 hrs in D. discoideum) but slow compared to reported half-times of N-end rule substrates (< 10 min). We have constructed a mathematical model that indicates dynamic range can be increased and target protein half-life under red light decreased by optimizing expression level of system components relative to host proteolysis components and we are currently implementing these optimizations in-vivo. Finally, our plate based characterizations will enable us to couple light-switchable TEVp with other optogenetic tools^2,3,4 without crosstalk for multi-channel control of protein production and degradation. This tool has many applications in science and biotechnology including improving control of dynamics and noise in eukaryotic synthetic circuits and the fabrication and modification of protein materials in vitro.

[1] Wehr M.C. et al., Nat Methods. 2006 Dec;3(12):985-93

[2] Renicke C. et al., Chem Biol. 2013 Apr 18;20(4):619-26.

[3] Tabor J.J. and Levskaya A, J Mol Biol. 2011 Jan 14;405(2):315-24.

[4] Muller K. et al., Nucleic Acids Res. 2013 Jul;41(12):e124.