In the early stages of our project, before we decided to proceed with our final “kill Switch” design, we had considered using Cre recombinases in order to disable our genetic circuit. We designed our plasmid and flanked it with two Lox-P sites, so that once Cre recombinases were introduced into the cell, it would excise the gene that was flanked between these two Lox-P sites (see figure 1).

Figure 1: Cre Lox system

In order to examine whether using recombinases was a viable option for our system we understood that modelling the mechanism was crucial. Unfortunately, such a model is extremely complex, and for lack of time and experience, was not feasible for us to Create. Luckily, we discovered that Team Edinburgh UG, whose whole project was concerned with recombinases, had already created such a model. Team Edinburgh UG generously offered to help us model our system, and through intense collaboration and specific modifications to their original model, we managed to attain informative results that shaped our decision regarding which “kill switch” mechanism to use.

We modelled two different processes:

1. Cells transiently transfected with our ToloGen plasmid, assuming a copy number of 20
2. Cells with stable integration of out ToloGen construct, this means a copy number of 1

This allowed us to analyze how the Cre system would function in our lab as well as in future in-vivo applications.

Lastly, we analyzed two Cre delivery systems:

1. Where Cre enzyme is taken into the cell directly through endocytosis
2. Where a Cre enzyme producing plasmid is taken into the cell

Assumptions:

1. Homogeneous mixture assumed throughout (e.g. if Cre molecule number is set to 100 at time = 0, it is assumed that the 100 Cre molecules are dispersed evenly inside the compartment regardless of its delivery method)
2. Values for half-life are taken from ‘general’ proteins
3. Compartmentalization is not taken into account (i.e. ignored nuclear export and import rate & spatio-temporal separation of transcription and translation, if any)

Simulation Settings:

• mRNA half-life (if using): 2 minutes
• Protein half-life: 90 minutes
• Plasmid copy number: 1 (i.e. stable integration into the genome) or 20
• Cre delivery method: as individual protein; or as plasmid expressing Cre recombinase
• For plasmid expressing Cre:
• Transcription rate: 1 mRNA molecule per 100 seconds
• Translation rate: adjusted so that the burst factor is 2
• For delivery of individual Cre protein:
  • Simulation performed with delivering 100, 1000, and 10000 Cre molecules

Guide:

• PC2 = Recombination Product with two Cre protein molecules attach to it (Cre binds to target site with very high affinity, so with a low protein degradation rate, Cre stably associate with the target site in the product construct).
• Simulations are repeated for 2000 times and are shown as an average value with +/- 1 standard deviation.

Results:

Delivery of 100 Cre Proteins: Genome integration (Figure 2) and Transient expression (figure 3):

Figure 2

Figure 3

Delivery of 1000 Cre Proteins: Genome integration (figure 4) and Transient expression (Figure 5):

Figure 4

Figure 5

Delivery of 1 Cre expressing plasmid: Genome integration (Figure 6) and Transient expression (Figure 7)

Figure 6

Figure 7

Delivery of 5 Cre expressing plasmid: Genome integration (Figure 8) and Transient expression (Figure 9)

Figure 8

Figure 9

Delivery of 10 Cre expressing plasmid: Genome integration (Figure 10) and Transient expression (Figure 11)

Figure 10

Figure 11

Analysis:

Based on the modelling results and clinical consideration we decided to forgo using Cre recombinases for our “Kill Switch”. As can be seen in Figure 6, the Cre recombinase at times is incapable of achieving the 100% recombination we require. It is our understanding that any recombination based system may ever fully eliminate every copy of our plasmid, and as such we decided to focus on a “Suicide Mechanism” wherein every cell containing our plasmid will undergo apoptosis. This is both more effective for our purposes and clinically proven.

Lastly, we would like to thank team Edinburgh UG again for the amazing work they did, our project would not have been the same without their help.