Team:TU Eindhoven/Modeling/Results






Modeling Results



The model can be configured by filling in many parameters, which leads to endless combinations. Below, a few of these possibilities are displayed and will be discussed. These data will lead to a concrete outcome that can be compared to our result in the lab and may give more information on how to improve our biosensor.

Linker Simulation



The relation between the intracellular linker length & the distance between the COMBs


The linker was simulated using a Monte Carlo method. Two membrane proteins were virtually placed on a 2D membrane. Then the linkers connecting the membrane proteins to their corresponding intracellular domain were simulated (For more information, check the approach page). The results for different distances between the membrane proteins are shown in figure 1. These histograms show that for small distances between two membrane proteins, the linker has a strong effect on the distance between the intracellular domains. For larger distances between the membrane proteins, the effect is less.
Furthermore, in the last case the effect of the linker on the intracellular distance is more even, almost symmetrical, whilst for a small distance between the membrane proteins, the linker will almost always increase the distance between the intracellular domains.




Figure 1: Results of the Monte Carlo linker simulation. The histograms show for different radii between the membrane proteins, the modeled distance between the intracellular domains. The used linker length was 28.4 nm


The relation between BRET or FRET efficiency and the intracellular linker length


These distances were converted into BRET or FRET efficiencies, to check the linker’s effect on these. Figure 2 shows the effect of the linker on the BRET and FRET efficiency. It shows that for small distances between two membrane proteins, which is for example the case when the clicked aptamers are bound to thrombin, the linker causes a lower BRET and FRET signal, while for larger distances it shows a slight increase in signal intensity.






Figure 2: This graph shows the BRET or FRET efficiency as function of the distance between the involved membrane proteins, for different linker lengths.




System Simulation



The relation between extracellular ligand binding and BRET or FRET signal


To estimate how well our sensor works, the increase in BRET or FRET signal on ligand binding was modeled. First the BRET or FRET signal resulting from the binding of aptamers is modeled, using Smoldyn. Then the BRET or FRET signal from the interaction between the unbound membrane proteins is added. This signal is divided by the amount of BRET or FRET found when only unbound membrane proteins are modeled.
Two concentrations of the construct were modeled. The results are depicted in figure 3. It shows that lesser membrane proteins will results in a higher relative signal increase. However, the optimal amount of membrane proteins is also dependent on the absolute signal, as a lower amount of membrane proteins will also lead to a decreased absolute signal, which is more difficult to detect .


Figure 3: The relative amount of BRET or FRET signal on binding, compared to the BRET and FRET signal when no ligand is bound to the aptamers. The upper plot shows the result for a modeled E. coli when 10,000 proteins, 5,000 of each intracellular domain, are located on the membrane. The lower plot shows the same data for a modeled E.coli with 50,000 proteins on the membrane, for each intracellular domain 25,000 proteins.



Conclusion

When analyzing the obtained results from the model, it can be concluded that the peptide linker plays a huge role in the efficiency of BRET or FRET in our biosensor. However, for enhancing our sensor, adjusting the linker length is not an effective way, as shown in figure 2. Optimizing the amount of proteins on the outer membrane of E.coli is, in contrast, is a viable option in order to improve our biosensor. As shown in figure 3, both the dynamic range and signal increase on ligand binding are extremely dependent on the amount of proteins present on the outer membrane. Furthermore, this amount is easy to control using tunable promoters. It can be concluded that, according to our model, our biosensors can be vastly improved by adjusting the amount of proteins expressed on the outer membrane.