Thermodynamics of Ligand Binding to Glutamate Dehydrogenase

The role of glutamate dehydrogenase (GDH) in disease has been exhibited in congenital hyperinsulinism, specifically, hyperinsulinism/hyperammonemia syndrome (HHS). GDH catalyzes the reversible deamination of glutamate to 2-oxoglutarate. Mutations in GDH can lead to GDH over activity, causing increased ATP production via the Krebs Cycle and excess insulin release. In addition, GDH over activity leads to depletion of glutamate, which is the source of the urea cycle precursor N-acetylglutamate. Reduction of N-acetylglutamate leads to reduced urea cycle activity and increased accumulation of ammonium. There are currently no treatments that directly target GDH and HHS patients are only treated symptomatically. To address the need to develop HHS therapeutics that directly target GDH, we began a computational investigation of the mechanism of allosteric ligand binding to GDH. During our computational investigation, we discovered a 40-year-old sequence error at the NADH/ADP/ECG binding site. Residue 387 was mistakenly identified as asparagine rather than the correct amino acid identity, lysine. The free energy penalty for the endogenous NADH ligand binding at the site having the incorrect residue was +5 kcal/mol per binding site. On correcting the sequence error our collaborators noticed improved refinement and electron density of the ligands at the NADH/ADP/ECG site, which led us to continue our investigation of the difference in NADH (inhibitor) versus ADP (activator) binding at the NADH/ADP/ECG site. The computed binding free energy difference using thermodynamic integration is -0.3 kcal/mol, which is within the -0.275 and -1.7 kcal/mol experimental binding free energy difference range thereby allowing for postulation of how the structural changes induced in GDH between the two ligands causes the switch from inhibitor to activator. Visual analysis of the structural conformations agree with some structural findings of ligand-GDH interactions but also challenge other interactions. Our computational findings can serve as a hypothesis generator for experimentalists and guide them in both drug design and in prioritizing mutations when further investigating allosteric ligand binding to GDH.