Team:KU Leuven/Project/Oscillator/wetlab

A well-known problem associated with the use of insect repellents is habituation. After a while insects will get used to the smell of the repellent and they won’t react anymore. Therefore we decided to implement an oscillator in our project for the production of (E)-β-farnesene (EBF) and methyl salicylate. In this part we will explain the genetics of our oscillator. For more information on the theoretical system and our in silico design, we refer to our design page and oscillator modelling page. We propose an implementation of the oscillator design and during the summer we focused on creating specifically a feed forward loop incorporating a quorum sensing molecule, since this is the workhorse of the oscillator.

As mentioned above we chose to use two quorum sensing molecules and four other transcription factors. A difficult step is the implementation of all the required interactions between those transcription factors. In our proposition we accomplish this by simply using four inducing transcription factors (on top of the two quorum sensing molecules). This section discusses how these can be used to create the OR gates and the AND gates.

How to implement an OR gate

The most elegant solution would be to design a promoter that would perfectly exhibit the characteristics of an OR gate. In the wetlab we started a slightly altered design, in which the goal of a asymmetric time delay is still achieved. For a further elaboration on that option we refer you to the design page. We do propose a potential implementation for a genuine OR gate. This would consist of a gene that has two promoters in front of it. There would then be production when either one of the transcription factors is present and the increase when both are present would be reduced. In the case of C this looks as follows:

Figure 2 | The proposed implementation of the OR gate.

How to implement the AND gate

We have also worked on a proposal to implement an AND gate. We need a promoter that is activated by one transcription factor and inhibited by another. An extra complication is that the transcription factor that inhibits has to activate another promoter and most transcription factors only exhibit either inhibitory behavior or inducing behavior. We propose two different solutions to tackle these issues, of which the second appeals to us the most. Our first solution consists of using a second gene that is activated by the transcription factor that was meant to inhibit the AND gate. This second gene creates antisense DNA of the eventual gene en thereby blocks the production of the protein. In the case of A this looks as follows:

Figure 3 | The first proposal for the AND gate.

Our second solution takes notice of the fact that both transcription factors that play a role in the AND gate are activating transcription factors. We do need to turn the function of one of those around. The method we propose for this is literally turning the operator site for that transcription factor around. When that transcription factor is present it will bind to its operator site and attract polymerases. That polymerase will not be able to transcribe the gene, since it is oriented in the wrong direction and the polymerase will simply obstruct the binding of any other polymerase. In the case of A this looks as follows:

Figure 4 | The second proposal for the AND gate.

Let's put names on it

On the first part of the modeling page of the oscillator we showed (at least in silico) that there is only a low sensibility of the oscillating behavior to the exact choice of promoters and transcription factors. This study is far from complete, but we did not wait until the engineers were completely certain and put names on it. We have used several parts from the iGEM registry and different transcription factors with a widespread use. This results in the following scheme that exhibits the behavior as discussed in the design page and validated in the oscillator modelling page.

Figure 5 | The total implementation of the oscillator.

This would bring the total implementation to 7 constructs which is, at this stage in the evolution of synthetic biology still a tough task. This will become easily feasible when the artificial synthesis of DNA becomes more available and when the concept of a ‘minimal cell’ is attained. That is why we only focused on creating a feed forward loop.

The genes

Since we want a colony-wide synchronization, TF A has to be a quorum-sensing molecule (as well as TF B). We chose for the autoinducer synthetase (BBa_C0076) that leads to the production of the quorum-sensing molecule 3OH,C14:1-HSL. Before this gene we placed a TetR repressible promoter (BBa_R0040), so the quorum-sensing molecule is only produced if there is no TetR present. Because this is a proof of concept, we want to be able to control the production of the quorum-sensing molecule. Therefore we placed the gene coding for TetR (BBa_P0340) behind a T7 promotor (BBa_I719005). When we put the construct in a Rosetta strain, we can control the activity of the T7 promotor. This strain has a chromosomal copy of the T7 RNA polymerase gene behind a lacUV5 promotor, which can be induced by IPTG. By adding IPTG, T7 RNA polymerase will be formed, which leads to the transcription of the TetR gene so no quorum-sensing molecules will be produced.

To be active, 3OH,C14:1-HSL has to form a complex with the CinR activator. This complex can bind the promotor BBa_R0078 and activate transcription. The gene for the CinR activator (BBa_C0077) is also placed under control of a TetR repressible promotor. So if no TetR is present, CinR and 3OH,C14:1-HSL are formed and they combine to activate the promotor BBa_C0078 which is placed before the gene coding for AraC (BBa_C0080; representing TF X) and before a gene for GFP (BBa_K082003; representing TF C). So the complex makes sure that TF X and TF C are formed. In the real model TF C would be the gene for EBF, but in our proof of concept model we use GFP, which allows us to easily detect if the system works.

Figure 6 | Genetic network of the oscillator.

Results

As you can see on figure 2, the proof of concept of the oscillator exists of a lot of different genes. We started with creating TF C, which in our case exists of the GFP gene coupled to a promoter and a terminator. We ligated the GFP-reporter gene (BBa_K082003) to a double terminator (BBa_B0015). These form the new intermediate brick BBa_K1060006. On figure 3 you can see the insert and backbone after digestion of this new brick. Afterwards we focused on TF A, which consists of the genes coding for the auto-inducer synthetase and a receptor that binds the quorum-sensing molecule. We coupled the cinI gene which codes for the autoinducer synthetase (BBa_C0076) to the double terminator (BBa_B0015). These form the new intermediate brick BBa_K1060007.

Figure 7 ǀ Digestion of our new brick BBa_K1060006