Progress in protein pK_a prediction

Abstract

pH represents solution acidity and plays a key role in many life events that are associated with human diseases. For instance, the β-site amyloid precursor protein cleavage enzyme, BACE1, which is a major therapeutic target of treating Alzheimer’s disease, functions within a narrow pH region around 4.5. In addition, the sodium-proton antiporter NhaA from Escherichia coli is activated only when the cytoplasmic pH is higher than 6.5 and the activity reaches the maximal around pH 8.8. To explore the molecular mechanism of a protein regulated by pH, it’s of importance to measure, typically by NMR, the binding affinities of protons to ionizable key residues, namely pK_a’s, which determine the deprotonation equilibria under a pH condition. However, web-lab experiments are often expensive and time consuming. In some cases, due to the structural complexity of a protein, pK_a measurements become difficult, making theoretical pK_a predictions in a try lab more advantageous.In the past thirty years, many efforts had been made for accurate and fast protein pK_a predictions with physics-based methods. Theoretically, constant pH molecular dynamics (CpHMD) methods that take conformational fluctuations into account give the most accurate predictions, especially the explicit-solvent CpHMD model proposed by Huang and coworkers (J. Chem. Theory Comput. 2016, 12, 5411-5421) which in principle is applicable to any system that a force field can describe. However, lengthy molecular simulations are usually necessary for extensive sampling in conformation. In particular, the computational complexity increases significantly if water molecules are included explicitly in the simulation systems. Thus, CpHMD is not suitable for high-throughout computing requested in industry. To accelerate pK_a prediction, Poisson-Boltzmann (PB) or empirical equation-based schemes, such as H++ and PropKa, have been developed and widely applied where pK_a’s are obtained via one-structure calculations. Recently, artificial intelligence (AI) is applied to the area of protein pK_a prediction, which leads to the development of DeepKa by Huang lab (ACS Omega 2021, 6, 34823-34831), the first AI-driven pK_a predictor. In this paper, we review the advances in protein pK_a prediction contributed mainly by CpHMD methods, PB or empirical equation-based schemes, and AI models. Notably, the modeling hypotheses explained in the review would shed light on future developments of more powerful protein pK_a predictors.