From our dedicated work in both wet lab and dry lab we’ve successfully proved that our STEP system can work in human cells and can be tuned to achieve an ideal output.

At the beginning, we improved a previous part of ECFP (BBa_E0020) to be BiFC components (BBa_K2886004 & BBa_K2886005), verified their reliability and employed them to test cell membrane localisation and binding ability of STEP (Figure 2). With cyan fluorescence observed in both circumstances, we can confirm that both our developed parts and the first part of STEP system works appropriately.

For another part of STEP, we proved the effectiveness of our transcription factor tTA and promotor pTight (Figure 3), and constructed a stable cell line with pTight-mCherry in both HEK293T and HeLa.

After both two sections of STEP were verified, we carried out experiment to confirm the whole system works and to explore the best condition for it to operate. We successfully figured out a relatively ideal transfection condition of TC/PC = 12 and total quantity of 2.44 μg per ml (Figure 4).
The results from our overall tests are confirmed by our model, which also suggests that there could be a certain set of TC/PC ratio and quantity that can lead to a best dynamic range which may be significantly higher than other conditions. Thus we believe that STEP can be further optimised through more subdivided gradients.

Apart from VEGF, we carried out experiments with D-Dimer-STEP, and achieved a larger dynamic range (Figure 5a). The VEGF-STEP was also tested in HeLa and extremely low background noise was observed during experiments (Figure 5b).

In our dry lab, to improve the binding affinity, we have redesigned 3 receptor binding domains: VEGF-scFv, KDR Ig domain 3 and D-Dimer-scFv (Figure 6). Also, we have redesigned the DNA sequence of the tTA promoter, to enhance downstream expression (Figure 7). Based on theoretical calculation of Gibbs free energy, we can now design or redesign receptors in silicon to get maximum binding affinity with the ligand. With input structure of receptors and ligands, we can find out the interface of the ligand-receptor complex, calculate binding energy, and mutate some residues to improve it. Even if there is only sequence information, homology modeling can be used to predict the structure. We believe, with the input sequence or structure of any ligand, we can redesign, or do de novo design of the sequence of receptor, thus extending additional functions of our system. For detailed results from dry lab, please visit our results page.