Ye Mei

Developing Polarized Protein-Specific Charges for Protein Dynamics: MD Free Energy Calculation of pKa Shifts for Asp26/Asp20 in Thioredoxin

Ab initio quantum mechanical calculation of protein in solution is carried out to generate polarized protein-specific charge(s) (PPC) for molecular dynamics (MD) stimulation of protein. The quantum calculation of protein is made possible by developing a fragment-based quantum chemistry approach in combination with the implicit continuum solvent model. The computed electron density of protein is utilized to derive PPCs that represent the polarized electrostatic state of protein near the native structure. These PPCs are atom-centered like those in the standard force fields and are thus computationally attractive for molecular dynamics simulation of protein. Extensive MD simulations have been carried out to investigate the effect of electronic polarization on the structure and dynamics of thioredoxin. Our study shows that the dynamics of thioredoxin is stabilized by electronic polarization through explicit comparison between MD results using PPC and AMBER charges. In particular, MD free-energy calculation using PPCs accurately reproduced the experimental value of pK(a) shift for ionizable residue Asp(26) buried inside thioredoxin, whereas previous calculations using standard force fields overestimated pK(a) shift by twice as much. Accurate prediction of pK(a) shifts by rigorous MD free energy simulation for ionizable residues buried inside protein has been a significant challenge in computational biology for decades. This study presented strong evidence that electronic polarization of protein plays an important role in protein dynamics.

A new quantum method for electrostatic solvation energy of protein

A new method that incorporates the conductorlike polarizable continuum model (CPCM) with the recently developed molecular fractionation with conjugate caps (MFCC) approach is developed for ab initio calculation of electrostatic solvation energy of protein. The application of the MFCC method makes it practical to apply CPCM to calculate electrostatic solvation energy of protein or other macromolecules in solution. In this MFCC-CPCM method, calculation of protein solvation is divided into calculations of individual solvation energies of fragments (residues) embedded in a common cavity defined with respect to the entire protein. Besides computational efficiency, the current approach also provides additional information about contribution to protein solvation from specific fragments. Numerical studies are carried out to calculate solvation energies for a variety of peptides including alpha helices and beta sheets. Excellent agreement between the MFCC-CPCM result and those from the standard full system CPCM calculation is obtained. Finally, the MFCC-CPCM calculation is applied to several real proteins and the results are compared to classical molecular mechanics Poisson-Boltzmann (MM/PB) and quantum Divid-and-Conque Poisson-Boltzmann (D&C-PB) calculations. Large wave function distortion energy (solute polarization energy) is obtained from the quantum calculation which is missing in the classical calculation. The present study demonstrates that the MFCC-CPCM method is readily applicable to studying solvation of proteins.

Folding of a Helix at Room Temperature Is Critically Aided by Electrostatic Polarization of Intraprotein Hydrogen Bonds

We report direct folding of a 17-residue helix protein (pdb:2I9M) by standard molecular dynamics simulation (single trajectory) at room temperature with implicit solvent. Starting from a fully extended structure, 2I9M successfully folds into the native conformation within 16 ns using adaptive hydrogen bond-specific charges to take into account the electrostatic polarization effect. Cluster analysis shows that conformations in the native state cluster have the highest population (78.4%) among all sampled conformations. Folding snapshots and the secondary structure analysis demonstrate that the folding of 2I9M begins at terminals and progresses toward the center. A plot of the free energy landscape indicates that there is no significant free energy barrier during folding, which explains the observed fast folding speed. For comparison, exactly the same molecular dynamics simulation but carried out under existing AMBER charges failed to fold 2I9M into native-like structures. The current study demonstrates that electrostatic polarization of intraprotein hydrogen bonding, which stabilizes the helix, is critical to the successful folding of 2I9m.

Simulation of NMR Data Reveals That Proteins’ Local Structures Are Stabilized by Electronic Polarization

Molecular dynamics simulations of NMR backbone relaxation order parameters have been carried out to investigate the polarization effect on the proteins local structure and dynamics for five benchmark proteins (bovine pancreatic trypsin inhibitor, immunoglobulin-binding domain (B1) of streptococcal protein G, bovine apo-calbindin D9K, human interleukin-4 R88Q mutant, and hen egg white lysozyme). In order to isolate the polarization effect from other interaction effects, our study employed both the standard AMBER force field (AMBER03) and polarized protein-specific charges (PPCs) in the MD simulations. The simulated order parameters, employing both the standard nonpolarizable and polarized force fields, are directly compared with experimental data. Our results show that residue-specific order parameters at some specific loop and turn regions are significantly underestimated by the MD simulations using the standard AMBER force field, indicating hyperflexibility of these local structures. Detailed analysis of the structures and dynamic motions of individual residues reveals that the hyperflexibility of these local structures is largely related to the breaking or weakening of relevant hydrogen bonds. In contrast, the agreement with the experimental results is significantly improved and more stable local structures are observed in the MD simulations using the polarized force field. The comparison between theory and experiment provides convincing evidence that intraprotein hydrogen bonds in these regions are stabilized by electronic polarization, which is critical to the dynamical stability of these local structures in proteins.

Quantum Computational Analysis for Drug Resistance of HIV-1 Reverse Transcriptase to Nevirapine Through Point Mutations

Quantum chemical calculation has been carried out to analyze binding interactions of nevirapine to HIV‐1 reverse transcriptase (RT) and single point mutants Lys103 → Asn (K103N) and Tyr181→ Cys (Y181C). In this study, the entire system of HIV‐1 RT/nevirapine complex with over 15,000 atoms is explicitly treated by using a recently developed MFCC (molecular fractionation with conjugate caps) approach. Quantum calculation of protein–drug interaction energy is performed at Hartree‐Fock and DFT levels. The RT‐nevirapine interaction energies are computed at fixed geometries given by the crystal structures of the HIV‐1 RT/nevirapine complexes from protein data bank (PDB). The present calculation provides a quantum mechanical interaction spectrum that explicitly shows interaction energies between nevirapine and individual amino‐acid fragments of RT. Detailed interactions that are responsible for drug resistance of two major RT mutations are elucidated based on computational analysis in relation to the crystal structures of binding complexes. The present result provides a qualitative molecular understanding of HIV‐1 RT drug resistance to nevirapine and gives useful guidance in designing improved inhibitors with better resistance to RT mutation. Proteins 2005.

Quantum study of mutational effect in binding of efavirenz to HIV‐1 RT

Full quantum mechanical computational study has been carried out to study binding of efavirenz (EFZ), a second generation FDA approved nonnucleoside inhibitor, to HIV‐1 reverse transcriptase (RT) and its K103N and Y181C mutants using the MFCC (molecular fractionation with conjugate caps) method. The binding interaction energies between EFZ and each protein fragment are calculated using a combination of HF/3‐21G, B3LYP/6‐31G* and MP2/6‐31G* ab initio levels. The present computation shows that Efavirenz binds to HIV‐1 RT predominantly through strong electrostatic interaction with the Lys101 residue. The small loss of binding to K103N mutant by Efavirenz can be attributed to a slightly weakened attractive interaction between the drug and Lys101 due to a conformational change of mutation. The small loss of binding to Y181C mutant by efavirenz can be attributed to the Glu698 residue moving closer to EFZ due to conformational change, which results in an increase of repulsive energy relative to the wild type (WT). The binding of efavirenz‐derived DPC961 to HIV‐1 RT is enhanced by an additional attractive interaction to residue Hid235 and reduced repulsion to Glu698, resulting in an increase of binding energy by about 4 kcal/mol. Proteins 2005.

Discovery and Characterization of Novel, Potent, and Selective Cytochrome P450 2J2 Inhibitors

Cytochrome P450 (CYP) 2J2 is one of the human CYPs involved in phase I xenobiotics metabolism. It is mainly expressed in extrahepatic tissues, including intestine and cardiovascular systems. The general role of CYP2J2 in drug metabolism is not yet fully understood, and the recent discovery that CYP2J2 can metabolize a wide range of structurally diverse drugs and its primary distribution in the intestine suggest its potentially indispensable role in first-pass intestinal metabolism and involvement in drug-drug interaction. To fully characterize its role in drug metabolism, selective and potent inhibitors of CYP2J2 are necessary tools. In the current study, 69 known drugs were screened for the inhibition of CYP2J2, and we discovered a number of marketed drugs as potent and selective CYP2J2 inhibitors. In particular, telmisartan and flunarizine have CYP2J2 inhibition IC50 values of 0.42 μM and 0.94 μM, respectively, which are at least 10-fold more selective against all other major metabolizing CYPs; moreover, they are not substrates of CYP2J2 and show no time-dependent inhibition toward this CYP. The results of enzyme kinetics studies, supported by molecular modeling, have also elucidated that telmisartan is a mixed-type inhibitor, and flunarizine competitively inhibits CYP2J2. The Ki for telmisartan is 0.19 μM, with an α value, an indicator of the type of inhibition mechanism, of 2.80, and flunarizine has a Ki value of 0.13 μM. These newly discovered CYP2J2 inhibitors can be potentially used as a tool to study CYP2J2 in drug metabolism and interaction in a clinical setting.

A numerically stable restrained electrostatic potential charge fitting method

Inspired by the idea of charge decomposition in calculation of the dipole preserving and polarization consistent charges (Zhang et al., J. Comput. Chem. 2011, 32, 2127), we have proposed a numerically stable restrained electrostatic potential (ESP)‐based charge fitting method for protein. The atomic charge is composed of two parts. The dominant part is fixed to a predefined value (e.g., AMBER charge), and the residual part is to be determined by restrained fitting to residual ESP on grid points around the molecule. Nonuniform weighting factors as a function of the dominant charge are assigned to the atoms. Because the residual part is several folds to several orders smaller than the dominant part, the impact of ill‐conditioning is alleviated. This charge fitting method can be used in quantum mechanical/molecular mechanical (QM/MM) simulations and similar studies, where QM calculated electronic properties are frequently mapped to partial atomic charges.

Polarization of intraprotein hydrogen bond is critical to thermal stability of short helix.

Simulation result for protein folding/unfolding is highly dependent on the accuracy of the force field employed. Even for the simplest structure of protein such as a short helix, simulations using the existing force fields often fail to produce the correct structural/thermodynamic properties of the protein. Recent research indicated that lack of polarization is at least partially responsible for the failure to successfully fold a short helix. In this work, we develop a simple formula-based atomic charge polarization model for intraprotein (backbone) hydrogen bonding based on the existing AMBER force field to study the thermal stability of a short helix (2I9M) by replica exchange molecular dynamics simulation. By comparison of the simulation results with those obtained by employing the standard AMBER03 force field, the formula-based atomic charge polarization model gave the helix melting curve in close agreement with the NMR experiment. However, in simulations using the standard AMBER force field, the helix was thermally unstable at the temperature of the NMR experiment, with a melting temperature almost below the freezing point. The difference in observed thermal stability from these two simulations is the effect of backbone intraprotein polarization, which was included in the formula-based atomic charge polarization model. The polarization of backbone hydrogen bonding thus plays a critical role in the thermal stability of helix or more general protein structures.

Calculations of Solvation Free Energy through Energy Reweighting from Molecular Mechanics to Quantum Mechanics

In this work, the solvation free energies of 20 organic molecules from the 4th Statistical Assessment of the Modeling of Proteins and Ligands (SAMPL4) have been calculated. The sampling of phase space is carried out at a molecular mechanical level, and the associated free energy changes are estimated using the Bennett Acceptance Ratio (BAR). Then the quantum mechanical (QM) corrections are computed through the indirect Non-Boltzmann Bennetts acceptance ratio (NBB) or the thermodynamics perturbation (TP) method. We show that BAR+TP gives a minimum analytic variance for the calculated solvation free energy at the Gaussian limit and performs slightly better than NBB in practice. Furthermore, the expense of the QM calculations in TP is only half of that in NBB. We also show that defining the biasing potential as the difference of the solute-solvent interaction energy, instead of the total energy, can converge the calculated solvation free energies much faster but possibly to different values. Based on the experimental solvation free energies which have been published before, it is discovered in this study that BLYP yields better results than MP2 and some other later functionals such as B3LYP, M06-2X, and ωB97X-D.