LISTEN

MSc. Thesis Defense: Nazlı Kocatuğ11-07-2018

COMPUTATIONAL ASSESSMENT OF THE EFFECT OF ALLOSTERIC MUTATIONS ON THE DYNAMICS OF PDZ DOMAINS

 

Nazlı Kocatuğ
Molecular Biology, Genetics and Bioengineering Program

MSc. Thesis, 2018

 

Thesis Jury

Prof. Dr. Canan Atılgan (Thesis Advisor), Prof. Dr. Ali Rana Atılgan,

Assist. Prof. Mehdi Sahihi

 

 

Date & Time: July 13th, 2018 – 10.00 AM

Place: FENS L067

Keywords: PDZ domains, allostery, molecular dynamics simulations, free energy calculations

 

Abstract

 

PDZ domain-containing proteins are involved in intercellular interactions such as trafficking, signaling, cell to cell communication and organization of signaling complexes. PDZ domains are themselves small proteins which typically consist of 90 to 100 amino acids. However, the extra α helix structure at the carboxyl terminus introduces a selective structural feature to the third PDZ domain of PSD-95 which has a stabilizing effect and participates in allosteric communication. PDZ domains are the most commonly studied models to understand single domain allostery without resulting in significant structural changes. One change triggers another change at distal site, and the source of the ‘changes’ are localized perturbations such as a binding event, posttranslational modification, a mutation or light absorption. Mutations can alter the stabilization of the protein and result ON or OFF state for ligand binding. They can also cause a change in the active site and affect the ligand preference. Here we investigate the reasons leading to the allosteric regulation of mutations and their effect on the ligand preferences.

By using third PDZ domain of postsynaptic density 95 (PSD-95) as a model system H372 directly connected to the binding site and G330 with a somewhat removed position were selected to assess the effect of allosteric mutations on the dynamics. In the literature, it was observed that the H372A and G330T/H372A mutations change ligand preferences from class I (T/S residue preference at position 2 of the ligand) to class II (hydrophobic amino acid preference at position 2 of the ligand). On the other hand, the G330T mutation leads to the recognition of both class I and class II types of ligands. Therefore, H372A is a ‘switching mutation’ while G330T mutation is ‘class bridging’. We have performed 200 ns molecular dynamics simulations for wild-type, H372A, G330T single mutants and a double mutant of third PDZ domain in the absence and presence of both types of ligands. The comparative study helps to identify the changes in the dynamics that are effective in the onset and prevention of allosteric communication. With the combination of free energy difference calculations and a detailed analysis of MD trajectories, the behavior of the PDZ domain under the mutations, which are ‘class bridging’(G330T) and ‘class changing’(H372A), and their effects on the ligand preferences and binding affinities are explained. We show that the ensemble view of allostery provides a better description of site-to-site couplingrather than a pathway view that assumes a direct connection between the effector and binding sites.