• To have a positive control of the testing circuit, we characterized another circuit (Fig. 12) that simulates the behaviour of testing circuit when the T-REX device is idle for the absence of TRANS-repressor or in case that TRANS-repressor mRNA is unable to silence LacI translation.

Figura 12 - BBa_K201001 with BBa_K201002

Dh5alpha cells were co-transformed with the [http://partsregistry.org/Part:BBa_K201001:Experience BBa_K201001] on a high copy number plasmid (pSB1A2) and BBa_K201002 on a low copy number plasmid (pSB3K3). To characterize this device and its sensitivity to the inducer, we studied both the static and the dynamic response to IPTG induction.

Static response Dh5alpha were inoculated in 5 ml of M9 medium with 0, 10, 20, 40, 60, 80, 100 uM IPTG, respectively. After O/N growth at 37° (about 12 h) samples were collected and slides prepared for microscope analysis. Acquired images were analyzed with the VIFluoR software. To obtain a significant representation of bacterial fluorescence, it was necessary to acquire several images, each one reporting a sufficient number of bacterial cells (n=60). VIFluoR operates image segmentation and then recognises the bacterial cells, yielding the mean fluorescence per bacterium as the output. The experimental data (Fig. 13) were used to identify, by the mathematical model, the operator binding affinity for the repressor LacI (K= 1.7 nM).

Fig.13 - Experimental data (blue lines) of the static induction after 0, 10, 20, 40, 60, 80, 100 uM IPTG induction. Data were fitted by the model (green line) to identify the operator-repressor binding affinity (K= 1.7 nM)

After parameter identification, we computed by the model the static control curve for the LacI repressed GFP generator (LacI inverter) (Fig.14).

Fig.14 - Model prediction of promoter repression by Lac I.

Dynamic response Dh5alpha cells were inoculated in the morning (9 a.m.) in 5 ml of M9 medium with no IPTG. After daily growth (about 8 h) the culture was diluted to an OD=0.1. To perform the induction analysis, the culture was splitted in two. A half was induced with 100 uM IPTG and the other was grown in control medium. 200 ul of each sample were used to fill plate wells (6 wells each). Cells were grown into a fluorimeter (Tecan M200) O/N (about 12h) at 37°. OD and fluorescence were sampled every 15 min (Fig. 15 and 16, respectively).

Fig.15 - Growth curve for the uninduced (black line) and induced (100 uM IPTG, light blue line)system.

Fig.16 - Absolute fluorescence curve for the uninduced (black line) and induced (light blue, 100 uM IPTG)system.

Fig. 17 - Model fitting of the experimental data. Experimental data (black lines) were fitted by the model considering a constant (blue line) or a varying (green line) amount of RNA polymerase

Experimental data of the fluorescence/OD ratio (Fig. 17; blue symbols) were compared to model predictions obtained either considering a constant (purple line) or progressively increasing (green line) amount of RNA polymerase. A good fitting can only be obtained if RNA polymerase available for transcription increases up to 30fold with respect to the initial value. This is consistent with the required activation of the major sigma subunit of RNA polymerase for transcription of most of the genes expressed in the exponential growth phase (Jishage M, Ishihama A. Proc Natl Acad Sci USA 1998; 95: 4953–8).

Model Prediction of Testing Circuit

• To test T-REX device, we developed the genetic circuit (Fig. 18)

Fig.18 - Testing Circuit

After identification of model parametres, the GFP levels as a function of TRANS/CIS affinity was predicted by simulating the whole test circuit (Fig. 19)

Fig.19 - Model prediction of testing circuit GFP level