Section 1 : Introduction

Our Fusion Protein, functioning Tb³⁺ adsorbtion, is built on four separated proteins (dLBT, LanM, si-tag and oprf). In order to fuse these separated proteins in a reasonable way, we performed a few in-silico work as follows. With the help of computer, we acquired more quantitative information about the structure and learned how to use the research result from different databases to assist our project.

Section 2 : Protein fusion

The first question of this project is that we have to build a original protein, which have never been studied. Fortunately, the parts of the fusion protein has been studied before. With the help of different papers ^[1-3], we found the function domains of different part proteins. Besides, to bind the dLBT and LanM in a more efficient way, we aligned the linkers between the 4 function domains in LanM, and 2 function domains in dLBT. Using the alignment result, we want to find out a functional perfect linker binding the two together, with the help of linker database ([Linker Database (vu.nl))^[4]. Finally, we link the four protein parts: oprf(4RLC), si-tag(Q3YWU2), LanM(6MI5), dLBT(2OJR) together with GS linker and (EAAAK) linker.

Section 3 : Structure Prediction

Before molecular docking, we need to get 3D model of the Fusion Protein. We found 3D models of separated Fusion Protein parts on the RCSB PDB.

Features include relatively fast and accurate deep learning based methods, RoseTTAFold and TrRosetta is a efficient way to perform homology modeling. We use the Robetta server to predict the 3D model of the Fusion Protein.

After prediction, we submitted the model to verify the protein’s structure through SAVES(ERRAT2), and the Overall Quality Factor exceeded 92, which indicate that the quality of the model is dependable.

We also predicted the structure on the I-TASSER server. The results are as follows.

Section 4 : Molecular docking

EF hands

The literatures shows that the loops with the highest affinity for the lanthanides have the following primary sequence of coordinating ligands: Asp-Xxx-Asn-Xxx-Asp-XxxXxx-Glu-Xxx-Xxx-Glu (Xxx: any amino acid). Tyrosine residues at positions 8 and 2 further increase the Tb³⁺ adsorbtion effect.

LanM have 4 EF hand domains, which possess several unique features relative to other EF hands binding Ln³⁺ rather than Ca²⁺. First, LanM retains all of the metalbinding Asp and Asn residues present in typical EF hands but also features an Asp residue in the ninth position in each of its EF hands, whereas Asp is encountered at this position in only roughly one-third of EF hand sequences. An Asp residue at position 9 has been shown in a model EF hand to contribute ∼2 orders of magnitude selectivity for Ln³⁺ over Ca²⁺. Second, Asn is rarely if ever observed at the first position in functional EF hands, as it is in EF hand 4 (EF4) in LanM. Third, all of LanM’s EF hands also possess a Pro residue at the second position. Finally, LanM features unusually short sequences between each EF hand loop (12 or 13 residues) instead of the 24 or 25 residues present in canonical EF hands.

In this part, we docked the LanM and Tb³⁺ by AutoDockTools, using vina. We learned that vina is an open-source program for doing molecular docking. Comparing to AutoDock 4, vina tends to be faster by orders of magnitude. Besides, vina is ease of using. All that is required is the structures of the molecules being docked and the specification of the search space including the binding site. Using vina helps us test the binding site of proteins and their affinity to Tb³⁺.

The free energy of Tb³⁺-Fusion Protein varied from -0.90 kal/mol to -1.97 kal/mol(details are showed below).

We counted the amino acids that appear most frequently at the docking interface, and selected the highest rated results for further analysis.

References