LOVpep and ePDZb

For initial testing and characterization of the system, we fused LOVpep and ePDZb Muller2014 to corresponding segments of the split firefly luciferase (1). We tested different positions of the split protein on the PDZ domain, while the split protein was kept at the N-terminus of the LOVpep domain due to the importance of the C-terminal peptide epitope. We also tested different ratios of both parts of the dimerizing system to determine the optimal conditions for further experiments (2B).

As luciferase activity was highest with split luciferase on the N-terminus of the ePDZ domain and a 1:3 ratio of the LOVpep:ePDZ constructs, all subsequent experiments were performed with this conditions. An important feature for real life applications is the ability of the system to be stimulated multiple times. Therefore, repeated association and dissociation was tested in real time, by adding luciferin to the medium and measuring luciferase activity upon induction by light ( 2 C). The system exhibited a delayed, but successful induction the first time, but the second induction was much weaker. These results indicate that the LOVpep/ePDZ system in this setup could not be induced more than once, so we decided to test an additional system.

 CRY2PHR-CIBN

As it has previously been shown on the example of split Cre recombinase Kennedy2010 the CRY2/CIB1 interaction upon excitation with blue light seems to be a suitable tool for the reconstitution of split proteins, allowing for spatial, temporal and dose-dependent optical control of protein dimerization. The CRY2/CIB1 system has also been used before at iGEM: Duke 2012, Marburg 2013, HNU_China 2014 and ANU_Canberra 2015

We adapted this system for the reconstitution of split luciferase to create a blue-light sensor, which enables easy characterization for further experiments. The N- and C-terminal split fragments of the firefly luciferase were fused to the C-terminus of the CRY2PHR and the CIBN proteins, since this topology has previously been shown to work with the Cre recombinase.

As luciferase activity was highest with a 1:3 ratio of CRY2PHR:CIBN constructs ( 4 A), all subsequent experiments were performed with this ratio. Next we tested if this system could be induced repeatedly in real time. The CRY2PHR/CIBN system showed a maximum activity after 2 minutes of induction and dropped to background 10 minutes after the stimulus was removed. The system could be induced repeatedly and reach high levels of activation at each stimulation ( 4 B).

 Light inducible proteases

To implement light as one of the input signals for protease-based signalling pathways or logic functions, we fused the CRY2PHR and the CIBN domains to the N- and C-terminal split domains of 3 different proteases (TEVp, TEVpE and PPVp) ( 5 A). Activity and orthogonality were tested by cleavage of a cyclic luciferase reporter . Cleavage of this reporter results in luciferase reconstitution and thus an increase in luminescence. Both the TEV and the PPV split proteases, coupled to the CRY2PHR/CIBN system, showed high catalytic activity and exquisite mutual orthogonality upon blue light illumination ( 5 B and C).

The reporter used for the western blot analysis was the luciferase reporter with the appropriate cleavage substrate inserted at a permissible site and the AU1 tag at the N-terminus. Uncleaved luciferase appears as a single band on a western blot, partially cleaved luciferase would result in two bands (uncleaved at 65 kDa and cleaved at 55 kDa), while complete cleavage would result in only the smaller band. Our results indicate that substrates carrying specific protease cleavage sites were cleaved after blue light illumination of cells harboring plasmids for expression of CRY2PHR:nTEV and CIBN:cTEV proteins ( 6 ).

Both methods confirmed successful, fast and dose-dependent responses. This is the first time TEVpE and PPVp were prepared as split proteins and shown to function in an inducible system.