Benchmarking pK_a prediction methods for Lys115 in acetoacetate decarboxylase

Area of Science:

• Biochemistry
• Computational Chemistry

Background:

• Acetoacetate decarboxylase (AADase) is a key enzyme in metabolic pathways.
• Understanding the ionization states of amino acid residues is critical for enzyme function.
• Lys115 and Glu76 are important residues in the active site of AADase.

Purpose of the Study:

• To accurately predict the pK_a of Lys115 in AADase using multiple computational methods.
• To investigate the influence of the protonation state of nearby residues, particularly Glu76, on the pK_a of Lys115.
• To reconcile computational predictions with experimental data and mutagenesis studies.

Main Methods:

• Utilized three pK_a prediction methods: PROPKA, Multiconformation Continuum Electrostatics (MCCE), and Molecular Dynamics/Thermodynamic Integration (MD/TI) with implicit solvent.
• Manually sampled protonation states of ionizable residues, focusing on Glu76.
• Performed MD/TI calculations with Glu76 in its protonated state.

Main Results:

• All three methods yielded incorrect pK_a values for Lys115 when Glu76 was deprotonated.
• When Glu76 was manually set to a protonated state, MD/TI predicted a pK_a of 5.3 for Lys115, closely matching the experimental value of 5.9.
• Site-directed mutagenesis data supports the conclusion that Glu76's protonation state, not its charge, influences Lys115's pK_a.

Conclusions:

• The protonation state of Glu76 significantly impacts the predicted pK_a of Lys115 in AADase.
• Accurate pK_a prediction necessitates explicit consideration of the protonation states of neighboring residues.
• It is postulated that Glu76's own pK_a is shifted, leading to its protonated (neutral) state within AADase under physiological conditions.