In synthetic biology, modeling is a powerful tool that uses theoretical models and computational approaches to predict, improve and further understand experiments. This year, modeling turned out to be one of the crucial cornerstones of our project due to the cancelled lab time.
Over the course of the project, it became increasingly clear that we wouldn’t be able to test out our project in the laboratory, which meant for us to adapt to the current situation and focus on different aspects of our project such as the modeling part. In this context, we built a total of three distinct models to gain further insights into the underlying biochemistry of our wastewater treatment approach:
We used the Rosetta Commons Software [1] developed by Baker Lab as well as the computational power of the Lichtenberg high-performance-computer at the TU Darmstadt to predict enzyme structures , test their stability and predict enzyme-ligand-interactions. We wrote a python program based on the work of B. Qin et al. [2] to simulate the growth mechanics and -conditions of our biofilm in cooperation with the modeling team from iGEM Hannover as well as an ODE-based MATLAB-model [3] to conceptually represent the functionality of our kill switch.
Since we weren’t able to include any self-generated data into our models this year, we hope that future iGEM teams can be inspired to use and fill them with life by expanding on them and implementing their own parameters. Our biofilm model can be used to describe biofilms built by other bacteria than B. subtilis simply by utilizing the values that describe those best. This also holds true for the kill switch model, which can be adapted to different biofilm-related systems if given the right data.
to make it short, through our modeling work we proved important cornerstones of our project and layed the ground work for further optimizations.
The described models can be further examined on the following pages:
Click the monitors to get to different parts of our modeling team.
With our Protein-Ligand-docking experiments we were able to define possible starting points for the optimizing process of our bacterial laccases for diclofenac. We analysed the binding interaction of the laccases CotA und T. Versicolor with diclofenac. Consequently, we could identify residues prone to mutation experiments to improve the transformation ability of CotA and thereby support the enzyme engineering process
EreB modeling
We were able to create a structure candidate of EreB using a homology modelling approach with RosettaCM. This helped us to understand EreB substrate binding and its reaction mechanism in more depth. After we validated structural stability using an MD simulation, we were able to create a model of the tasA-EreB fusion protein using our modelled structure as one of the templates for comparative modelling. This functioned as a prove of concept for our idea of immobilizing transforming enzymes in the B. subtilis extracellular matrix. MD simulation was then used to prove structural stability of both protein domains. This way we validated functionality of fused enzymes for immobilization in our biofilm.
Biofilm modeling:
With the help of our software tool we can easily predict the growth of our biofilm as well as the density with different lab parameters. As our software tool is also capable of calculating the adhesion force, we are also able to predict the stability
of the biofilm in different conditions.
kill switch modeling:
With the help of our model, we are able to improve our kill switch in various ways. One way would be to try various protein levels of ComX and ComP in silico. Also, we could use lab data to refine the model and improve the stability of our kill switch by removing interference of other substances inside the cells.
