We acquired the strains E.coli BL21-4A5-R-IG with recombinant plasmid vector pSB4A5-R-IG, according to the initial experimental methods.
     The recombinant colonies was transferred into LB culture medium with ampicillin , and shaked overnight at 37℃, then took the medium for fluorescence detection. We found the fluorescence signal was weak (Figure 1) and cannot be detected under the condition of its background expression as we expect. Thus it is speculated that the phenomenon may be caused by the low effective expression of the LuxI protein, in which cases there is not enough AHL.

     To test our hypothesis, we next verified the module-LuxI. The fluorescence intensity significantly increased after adding different concentrations of AHL (Figure 3). Therefore, the system was probably activated by adding AHL which indicated that there was a problem with the LuxI module. By analyzing the LuxI gene sequence, we found out there is an LVA tag in LuxI module, which was cloned from iGEM part (BBa_C0061). This tag will accelerate the degradation rate of its corresponding proteins, so that it can prevent the LuxI protein from playing a stable role in cell. It may, therefore, be concluded that the AHL signal was too weak to activate the positive feedback system.

     After fermenting E.coli BL21-4A5-R-IG-△LVA and adding AHL at the same time, we measured fluorescence values comparing with the control group. The fluorescence value of test group increased rapidly (Figure 4).
     Our system was activated that indicated the positive feedback was effective. However, after 20 hours, the AHL-free control group was still up to xx, in other words, the system background expression could also make the system to be activated and enable the positive feedback system to be constructed successfully.

     In order to avoid the impact of system detection limits and the influences of the prolonged cell growth period (See modeling for details), we added lac operon to the positive feedback system expecting the positive feedback system to be activated when IPTG was added into the culture medium in the stationary period of E.coli.
     With adding IPTG, the results showed that the positive feedback was normal; without IPTG, the positive feedback was closed (Figure 6). This indicated that the PLux/Lac switch functions well.
     Then we added IPTG to the medium in the cell stationary phase. After 12h induction, no positive feedback was found, which means the system didn’t work normally.
And then AHL was added in the stationary phase, in that case the positive feedback system still didn’t work . Through reading lots of literature, we speculated that model-LuxR may not work in the stable period, leading to the failure of activating positive feedback system ^[1]. (If you want to read more details, you can click here to see our contribution)
     So we would rather choose logarithmic growth period than stationary phase to detect the input signal.

     After adding a series of concentrations for AHL, we measured fluorescence of E.coli BL21-4A5-RG and E.coli BL21-4A5-R-IG-△LVA. The results showed that the measurement accuracy of E.coli BL21-4A5-R-IG-△LVA could reach the concentration of 10^-9M(Figure 7) while the one of E.coli BL21-4A5-RG, which has no positive feedback, was 10^-8M. As for the ordinary high performance liquid chromatography, when AHL concentration accuracy reached 10^-6M (Figure 8), its detection was not very obvious. Comparing the measurement accuracy of E.coli BL21-4A5-R-IG-△LVA with that of E.coli BL21-4A5-RG, the former is approximately an order of magnitude higher than the latter. When compared with High Performance Liquid Chromatography(HLPC), the measurement accuracy of E.coli BL21-4A5-R-IG-△LVA was nearly three orders higher than that of HLPC. To achieve the accuracy of 10^-9M, you can also use HPLC-TMS, but its operation will be slightly complex^[2].

     A high background expression can affect the detection limits of the system. Fortunately, AiiA hydrolytic enzymes can hydrolyze the AHL protein while LVA tag accelerates degradation of LuxI, thereby reducing the generation of AHL. Both of these ways are able to control AHL background expression. However, the effect of LVA tag will be strengthen with the enlargement of positive feedback system. The reason is that LVA tag exists on model LuxI. With the expression of LuxI, the total number of intracellular tags will increase accordingly, which may lead to a too strong control ability on the positive feedback effect, during the stationary period. According to the Michaelis-Menten equation, when the AHL is enough, the degradation rate of AHL is constant. So using AiiA to control background expression would not have a significant effect on the positive feedback. Thoughtfully, we chose to add AiiA for system control. When compared with E.coli BL21-4A5-R-IG-△LVA, the added AiiA will delay the threshold, but the increase rate is basically same as before(Figure 9), which was in line with our expectations. The decrease in the end point may be caused by the thing that the positive feedback effect had already stopped before cells reached the stationary period, which also confirmed our previous speculation that LuxR did not work stationary period.

     We constructed E.coli BL21-1C3-RT to detect whether the converter is normal or not. It can be found that the background level was high, which means the double plasmid system may need more AiiA hydrolase to control the background expression of AHL. We intended to switch to a stronger promoter or RBS to improve AiiA expression, but due to the limited time, the improvement hasn’t been done.
     If we could solve the problem of AHL background expression ,we would isolate E.coli BL21-4A5-AiiA/1C3-IT which contains both pSB4A5-R-IG-△LVA-AiiA and pSB1C3-LuxI-TetR plasmids, then add different concentrations of Tetracycline into LB medium to detect performance of that system.

     1. Change the positive feedback system to make sure that can work in a stationary phase
     2. Change the different intensities of the promoter or RBS to eliminate background expression

     [1] Zhu J. Winans SC. Auto inducer binding by the quorum sensing regulator TraR increases affinity for target promoters in vitro and decreases TraR tumover rates in whole cells. Proc.Natl.Acad. Sci. USA, 1999，96:4832-4837.
     [2] Chenchen Ma, Jie Ou, Bolin Li. Method for measuring AHLs by HPLC-TMS:CN 103063774 B[P]. 2014.

Number	Plasmid name
1	pSB4A5-RG
2	pSB4A5-R-IG(BBa_K2311000)
3	pSB1C3-RFP-TetR
4	pSB1C3- LuxI-TetR
5	pSB4A5-R-IG-△LVA(BBa_K2311002)
6	pSB4A5-R-IG-△LVA-lac5
7	pSB4A5-R-IG-△LVA-AiiA(BBa_K2311001)

Number	Strain name
1	E.coli BL21-4A5-RG
2	E.coli BL21-4A5-R-IG
3	E.coli BL21-1C3-RT
4	E.coli BL21-4A5-R-IG-△LVA
5	E.coli BL21-4A5-R-IG-△LVA-lac
6	E.coli BL21-4A5-R-IG-△LVA-AiiA

     If you want to read more details about how to construct the plasmids and strains, you can click here to see our protocol and notebook!