C. Satheesan Babu

Solvation free energies of polar molecular solutes: Application of the two-sphere Born radius in continuum models of solvation

A two-sphere description of the effective Born radius for spherical ions was found in previous work to yield accurate free energies for spherical ions. This effective Born radius (Reff) was identified as the mean of the ionic radius (Rion) and the distance to the first peak of the ion–oxygen/hydrogen radial charge or number density distribution function (Rgmax); i.e., Reff=(Rion+Rgmax)/2. To see whether this prescription also applies to the solvation of nonspherical polar molecules, it was used in finite-difference Poisson methods as well as in Kirkwood and generalized Born models to compute solvation free energies of model diatomic molecules of varying interatomic bond distances. Hydration free energies for the same model systems were also derived from free energy simulations in the presence of explicit water molecules. The good agreement between explicit solvent results and continuum solvent results with the two-sphere Born radius indicates that the latter description provides the required solute–solven...

On the charge and molecule based summations of solvent electrostatic potentials and the validity of electrostatic linear response in water.

Solvent-induced electrostatic potentials and field components at thesolute sites of model Na+q–Cs-q molecules were computed bysumming over either solvent charges (q-summation) or solventmolecular centers (M-summation) from molecular dynamics simulations.These were compared with values obtained by solving Poisson equation withthe dielectric boundary defined by Reff = (Ratom+Rgmax )/2.q-summation using cut-offs that are ≤ 10 Å generallyunderestimates or overestimates the magnitude of (a) the potentials and field components atNa+q and Cs-q relative to the theoretical values and (b)electrostatic solvation free energies of the dipolar solutes assuminglinear solvent response relative to the respective values from free energysimulations. Furthermore, the q-summed electric potentials showedsignificant oscillations even beyond the second hydration shell. Incontrast, the corresponding M-summed potentials plateaued after thefirst hydration shell. Although the different water molecular centersyielded different converged potential values, the dipole center producedvalues in remarkable agreement with the theoretical values for solutecharges ranging from 1 to 0.1e, indicating the existence of an a convenient molecular center for computing these quantities. In contrast to theM-summed potentials, the electrostatic field components andelectrostatic solvation free energies from linear response relationshipswere found not to be sensitive to the choice of the molecular centerfor typical cut-off distances (8 to 12 Å) used in most simulations.

Protein/Solvent Medium Effects on Mg2+-Carboxylate Interactions in Metalloenzymes

We employed umbrella sampling molecular dynamics simulations in explicit water to study the binding of the Mg(2+) cofactor to ribonuclease H (RNase H) from three different organisms. We show that the enzyme can differentiate between different Mg(2+)-binding modes that are nearly equally stable by creating a free-energy barrier between a water-rich mode and a water-depleted mode. Through a comparison with the corresponding free-energy barrier in water, this effect is shown to emanate from the enzymess three-dimensional architecture and its associated environment. Implications of these protein medium effects in RNase H function and in structure-based drug/molecular design are discussed.

Modeling Zn2+ Release From Metallothionein

Mammalian metallothioneins (MTs) comprise a Zn3Cys9 cluster in the β domain and a Zn4Cys11 cluster in the α domain. They play a crucial role in storing and donating Zn(2+) ions to target metalloproteins and have been implicated in several diseases, thus understanding how MTs release Zn(2+) is of widespread interest. In this work, we present a strategy to compute the free energy for releasing Zn(2+) from MTs using a combination of classical molecular dynamics (MD) simulations, quantum-mechanics/molecular-mechanics (QM/MM) minimizations, and continuum dielectric calculations. The methodology is shown to reproduce the experimental observations that (1) the Zn-binding sites do not have equal Zn(2+) affinity and (2) the isolated β domain is thermodynamically less stable and releases Zn(2+) faster with oxidizing agents than the isolated α domain. It was used to compute the free energies for Zn(2+) release from the metal cluster in the absence and presence of the protein matrix (protein architecture and coupled protein-water interactions) to yield the respective disulfide-bonded product. The results show the importance of the protein matrix as well as protein dynamics and coupled conformational changes in accounting for the differential Zn(2+)-releasing propensity of the two domains with oxidizing agents.