Template:Team:TU Eindhoven/Modeling HTML

Modeling

Every good sensor must meet two requirements: specificity and sensitivity;

Approach

Modeling our sensor can give a good insight in the behavior of its input to signal transduction. Even though it is hard to predict the exact signal a certain event will produce, a relative approach, by comparing different states of the device, can give valuable information on how to improve and optimize the sensor. This model therefore compares the signal produced by the intracellular component on presence of the ligand, and on absence, in order to get a signal to noise ratio of the sensor (Hereafter “noise” will be used as the measured signal of the sensor in absence of a ligand). Because two different types of constructs have been made in the lab, a construct using split luciferases and one using BRET (Bioluminescence Resonance Energy Transfer), different approaches have been applied. The methods and results are discussed below.

Affinity based approach

The mechanism of action of this construct is based on the binding of the two components of a split luciferase. These type of constructs are predicted to have relatively large noise values, in comparison to non-affinity based intracellular components, because the two split luciferases will non-covalently bind to each other even in the absence of a ligand. The software package Smoldyn was used to perform a molecular simulation in order to predict the raise of signal on ligand binding. In Smoldyn, a sphere was created to represent an E. coli bacterium. The radius of this sphere was set to 0.53 µm, in order to give it the same membrane area as a cylinder of 2.5 µm in length and 0.5 µm in width, which approximately resembles the shape of an E. coli. The system makes use of a spherical boundary which is periodic; a particle that leaves the sphere, enters it on the other side. The ligand is freely floating in the system and the membrane protein domains are attached to the sphere resembling the E. coli. These domains interact with each other based on the properties, i.e. the association and dissociation constants, of the aptamers used in the lab. The following reactions were entered in Smoldyn:

The simulation was run for 2 µs with a time step of 1 ns. To simulate the noise of this sensor, the same simulation was run, except in absence of the ligand. The results of these simulations can be viewed in the graphs below:

Non-affinity based approach (NanoLuc & mNeonGreen)

In contrast to the other construct, these intracellular components don’t have interaction with each other; instead this constructs makes use of BRET , which only works on when the donor and acceptor are located within 10nm of each other BRON.

Simulating the Linker

Since the main limitation of the efficiency of BRET and FRET is the distance between the two participating domains, the length of linker in the intracellular compartment plays an important role in the success of our sensor.