The next step of the working bacterial two-hybrid system is the examination of the interaction between the DNA binding domain and the possible binding site upstream of the promoter region. An in vitro interaction was demonstrated via electrophoretic mobility shift assay (EMSA) with the single OR1 binding site (BBa_K2082231) and the double OR1 and OR2 binding site (BBa_K2082239) of the phage 434. The DNA fragment without addition of the protein SH2-cI runs way faster than after adding the SH2-cI protein (figure 1A). This observation indicates an interaction between 434 cI an the binding site OR1. Binding of the protein to the DNA increased the weigth of the DNA fragment in the gel, which results in the slower movement and the visible band at the higher position in the gel. To disprove the possibilities of general DNA binding of the cI protein, another fragment was tested with nearly the same fragment size than the OR1 fragment but with a different sequence. However, the agarose gel did not reveal any recognizable differences between addition of SH2-cI and no addition of SH2-cI. This result confirmed our conjecture, that there is only an interaction between cI and the cI binding site. The binding at the single binding site OR1 of cI is compared with the binding at the double binding site OR1 and OR2(figure 1B). In both cases, a shift by the SH2-cI protein with the DNA is noticeable, but it is not possible to see a stronger shift with the doubled binding site. Therefore, the construct with one 434 binding site is sufficiently for binding with the cI protein. Moreover, it was demonstrated that one binding site upstream of the promoter is enough for a strong binding and adequate for the bacterial two-hybrid system used.

A new construct was produced as in vivo binding control. A GFP gene under the control of a constitutive Anderson promoter (BBa_K608010) was cloned directly upstream of the reporter construct of the single reporter part BBa_K2082211 and of the reporter and SH2-cI combined part BBa_K2082231 without any terminator. The expectation was given, that the Anderson promoter would be able to increase the transcription rate of the RFP if the RNA polymerase can not be stopped. The polymerase is able to read over the second promoter and leads to an expression of the normally weakly expressed RFP gene. If the cI protein would bind at the binding site between the GFP gene and the optimized lacZ promoter (BBa_K2082210), it is possible that the polymerase has to stop the expression due to steric inhibition. Therefore, the relation of RFP to GFP expression could give a clue if cI is binding in vivo. To analyze this hypothesis the GFP and RFP expression are measured from a liquid culture of two E. coli cultures carrying different plasmids. The first culture was transformed with the GFP coupled part BBa_K2082211 and the second with the GFP coupled part BBa_K2082231. The main difference of these two parts is the ability of BBa_K2082231 of producing the SH2-cI fusion protein.
A first measurement was performed in the Infinite M200 plate reader (Tecan). The culture without producing the cI protein shows a bit higher RFP intensity (figure 3). Integrating the error bars in the analysis, it is not possible to say if their is a real difference between the two constructs designed for this experiment.

Therefore, a second measurement was conducted in the FACS (fluorescence-activated cell scanning) for direct measurement of single cell events and verification of the results given by the TECAN. At first, the GFP expression of the two designed constructs was compared with the native reporterBBa_K2082211. In total, 50,000 cells were measured in the FACS system, which revealed a much higher GFP intensity produced by the GFP gene carrying cells (figure 4A). This supports the supposition, that GFP was correctly cloned upstream of the reporter sequences. A comparison of the RFP intensity of the reporter without the GFP gene with the modified version exhibits an about 85% stronger RFP production in the cells with the GFP gene and the reporter (figure 4B). The RFP expression increases with the read over of the RNA polymerase, after docking at the constitutive Anderson promoter region upstream of the GFP gene.
A comparison of the cI-SH2 and GFP producing cells with the cells only producing GFP also demonstrated differences in the RFP intensity (fugure 4C). The cells with cI-SH2 proteins had an about 35% lower RFP signal after measuring the RFP intensity in 50,000 single cells. Normalized on the GFP production, the differences raises to 44% lower RFP intensity per measured GFP intensity.

Therefore, an in vivo validation of the DNA binding domain is also possible. One possible explanation for the reduced RFP production copared to the native reporter level with the binding of cI is the possibility that the binding protein could compete with the RNA polymerase for the DNA sequence with the binding site. Therefore, the polymerase represses the binding cI protein and continues the expression of the genes. Under normal conditions the cI protein binds at four binding sites and builds an octamere structure to completely repress the transcription activity. The single binding site OR1 could lead to a lower binding strength of cI and therefore, a higher chance that the RNA polymerase could repress the cI protein. All in all the binding of cI was shown under in vivo and in vitro conditions.