Team:TU-Eindhoven/Modeling Approach
Modelling is a cost efficient and fast way to design synthetic systems and determining the properties and behavior of a system (keep in mind that modelling is based on approximations and gives no 100% guarantees). With our project we will create a heterodimeric and tetrameric version of the T14-3-3 protein, however, the working mutations known from previous research are not enough to create the amount of orthogonal pairs needed.[1] We used modelling to find new mutations, giving us enough orthogonal pairs for our project.
The Rosetta software package is a molecular structure predictor that includes several predicting algorithms for computational modeling and analysis of protein structures. Several applications of Rosetta are de novo protein design, ligand docking, structure prediction of macromolecules and macromolecular complexes.
The software is now used in places including government laboratories, research centers, and corporations. It is still on its way to improve even further to reach its goal of being able to clarify macromolecular interactions, completely custom molecules and becoming faster and more accurate.
Rosetta mostly uses knowledge-guided Metropolis Monte Carlo* approaches combined with knowledge based energy functions. These energy functions assume that the molecular properties can be derived from the available information (e.g. a Protein DataBase).
*A Monte Carlo Algorithm is an algorithm in which a random action is taken to bring a new state to a system, if this action contributes positively the state is automatically accepted, if it contributes negatively the state has a chance to be accepted, this prevents the system getting stuck in a local minimum.
In order to design an Orthogonal pair a couple of steps had to be taken. The first of these steps was making the available protein databank file (PDB), a T14-3-3 CT52 complex with Fusicoccin (PDB ID:2O98), useable in Rosetta by relaxing it into the Rosetta-forcefield, making sure there are no sterical clashes between atoms and other hindrances that could result in incorrect output. Then the CT52 was
backrubbedA backrub simulation is a simulation in which the backbone of a protein is rotated in order to find a more stable structure.
in order to simulate its backbone flexibility, this was not necessary for T14-3-3 since it is a very robust protein.
The next step was determining which residues (identity of an object in the PDB i.e an amino acid or fusiccocin) were important in the binding between T14-3-3 and CT52, these were found using computational alanine scanning on the Robetta server[2], tested residues were based on earlier findings of Ottmann C. et al.[3]
Using the results of the computational Alanine scanning, the residues that had a significant relevance to the binding interface were used to search for point mutants. In order to find an orthogonal pair, a mutation in the CT52 must first destabilize the interface, then this mutation needs to be compensated by introducing restabilizing mutations in the T14-3-3. CT52 is first destabilized because it is more flexible and thus more prone to structural change due to a mutation. Finally, to ensure orthogonality (i.e. no cross-linking), the mutations introduced in T14-3-3 should not only restabilize the binding with the mutated CT52, but also significantly destabilize the binding with the wildtype CT52.
Click here to download the protocol.
- [1] Skwarczynska, M., Molzan, M., & Ottmann, C. (2013). Activation of NF-κB signalling by fusicoccin-induced dimerization. Proceedings of the National Academy of Sciences, 110(5), E377-E386.
- [2] http://robetta.bakerlab.org/
- [3] Ottmann, C., Marco, S., Jaspert, N., Marcon, C., Schauer, N., Weyand, M., ... & Rigaud, J. L. (2007). Structure of a 14-3-3 coordinated hexamer of the plant plasma membrane H+-ATPase by combining X-ray crystallography and electron cryomicroscopy. Molecular cell, 25(3), 427-440.
