Haluk Resat

Comparative molecular dynamics simulations of amphotericin B–cholesterol/ergosterol membrane channels

Amphotericin B (AmB) is a very effective anti-fungal polyene macrolide antibiotic whose usage is limited by its toxicity. Lack of a complete understanding of AmBs molecular mechanism has impeded attempts to design less toxic AmB derivatives. The antibiotic is known to interact with sterols present in the cell membrane to form ion channels that disrupt membrane function. The slightly higher affinity of AmB toward ergosterol (dominant sterol in fungal cells) than cholesterol (mammalian sterol) is regarded as the most essential factor on which antifungal chemotherapy is based. To study these differences at the molecular level, two realistic model membrane channels containing molecules of AmB, sterol (cholesterol or ergosterol), phospholipid, and water were studied by molecular dynamics (MD) simulations. Comparative analysis of the simulation data revealed that the sterol type has noticeable effect on the properties of AmB membrane channels. In addition to having a larger size, the AmB channel in the ergosterol-containing membrane has a more pronounced pattern of intermolecular hydrogen bonds. The interaction between the antibiotic and ergosterol is more specific than between the antibiotic and cholesterol. These observed differences suggest that the channel in the ergosterol-containing membrane is more stable and, due to its larger size, would have a higher ion conductance. These observations are in agreement with experiments.

Enzyme-inhibitor association thermodynamics: explicit and continuum solvent studies.

Studying the thermodynamics of biochemical association reactions at the microscopic level requires efficient sampling of the configurations of the reactants and solvent as a function of the reaction pathways. In most cases, the associating ligand and receptor have complementary interlocking shapes. Upon association, loosely connected or disconnected solvent cavities at and around the binding site are formed. Disconnected solvent regions lead to severe statistical sampling problems when simulations are performed with explicit solvent. It was recently proposed that, when such limitations are encountered, they might be overcome by the use of the grand canonical ensemble. Here we investigate one such case and report the association free energy profile (potential of mean force) between trypsin and benzamidine along a chosen reaction coordinate as calculated using the grand canonical Monte Carlo method. The free energy profile is also calculated for a continuum solvent model using the Poisson equation, and the results are compared to the explicit water simulations. The comparison shows that the continuum solvent approach is surprisingly successful in reproducing the explicit solvent simulation results. The Monte Carlo results are analyzed in detail with respect to solvation structure. In the binding site channel there are waters bridging the carbonyl oxygen groups of Asp189 with the NH2 groups of benzamidine, which are displaced upon inhibitor binding. A similar solvent-bridging configuration has been seen in the crystal structure of trypsin complexed with bovine pancreatic trypsin inhibitor. The predicted locations of other internal waters are in very good agreement with the positions found in the crystal structures, which supports the accuracy of the simulations.

Conformational properties of amphotericin B amide derivatives – impact on selective toxicity

Even though it is highly toxic, Amphotericin B (AmB), an amphipathic polyene macrolide antibiotic, is used in the treatment of severe systemic fungal infections as a life-saving drug. To examine the influence of conformational factors on selective toxicity of these compounds, we have investigated the conformational properties of five AmB amide derivatives. It was found that the extended conformation with torsional angles (φ,ψ)=(290°,180° ) is a common minimum of the potential energy surfaces (PES) of unsubstituted AmB and its amide derivatives. The extended conformation of the studied compounds allows for the formation of an intermolecular hydrogen bond network between adjacent antibiotic molecules in the open channel configuration. Therefore, the extended conformation is expected to be the dominant conformer in an open AmB (or its amide derivatives) membrane channel. The derivative compounds for calculations were chosen according to their selective toxicity compared to AmB and they had a wide range of selective toxicity. Except for two AmB derivatives, the PES maps of the derivatives reveal that the molecules can coexist in more than one conformer. Taking into account the cumulative conclusions drawn from the earlier MD simulation studies of AmB membrane channel, the results of the potential energy surface maps, and the physical considerations of the molecular structures, we hypothesize a new model of structure-selective toxicity of AmB derivatives. In this proposed model the presence of the extended conformation as the only well defined global conformer for AmB derivatives is taken as the indicator of their higher selective toxicity. This model successfully explains our results. To further test our model, we also investigated an AmB derivative whose selective toxicity has not been experimentally measured before. Our prediction for the selective toxicity of this compound can be tested in experiments to validate or invalidate the proposed model.

Free energy simulations: Correcting for electrostatic cutoffs by use of the Poisson equation

The use of electrostatic cutoffs in calculations of free energy changes by molecular dynamics or Monte Carlo simulation is known to introduce errors, which can be quite large when the net charge of the system is changed. The Born equation has often been used to correct for such errors, but this and other analytical methods cannot be used for many systems with complicated structures. Here, we show that numerical methods for solving the Poisson equation, which have been extensively developed recently for studies of solvation thermodynamics, provide a more generally applicable alternative to the traditional Born‐type corrections.

Ion passage pathways and thermodynamics of the amphotericin B membrane channel

Abstract. Amphotericin B is a polyene macrolide antibiotic used to treat systemic fungal infections. Amphotericin Bs chemotherapeutic action requires the formation of transmembrane channels, which are known to transmit monovalent ions. We have investigated the ion passage pathways through the pore of a realistic model structure of the channel and computed the associated thermodynamic properties. Our calculations combined the free energy computations using the Poisson equation with a continuum solvent model and the molecular simulations in which solvent molecules were present explicitly. It was found that there are no substantial structural barriers to a single sodium or chloride ion passage. Thermodynamic free energy calculations showed that the path along which the ions prefer to move is off center from the channels central axis. In accordance with experiments, Monte Carlo molecular simulations established that sodium ions can pass through the pore. When it encounters a chloride anion in the channel, the sodium cation prefers to form a solvent-bridged pair configuration with the anion.

Correcting for electrostatic cutoffs in free energy simulations: Toward consistency between simulations with different cutoffs

The use of electrostatic cutoffs in calculations of free energy differences by molecular simulations introduces errors. Even though both solute–solvent and solvent–solvent cutoffs are known to create discrepancies, past efforts have mostly been directed toward correcting for the solute–solvent cutoffs. In this work, an approach based on the generalized reaction field formalism is developed to correct for the solvent–solvent cutoff errors as well. It is shown using a series of simulations that when the cutoff lengths are significantly smaller than the half unit cell size, and the solute–solvent cutoff is not much larger than the solvent–solvent cutoff, the new algorithm is able to yield better agreement among simulations employing different truncation lengths.

Correcting for solvent–solvent electrostatic cutoffs considerably improves the ion-pair potential of mean force

A recently developed algorithm based on the continuum treatment of the solvent molecules beyond the electrostatic cutoff sphere is applied to the potential of mean force results between sodium and chloride ions to study the effects of the solute–solvent and solvent–solvent cutoff errors. The results show that although the solute–solvent correction improves the thermodynamic results slightly, physically realistic results are obtained only when the solvent–solvent correction is applied. This further supports past findings that proper treatment of solvent–solvent interactions is as important as that of the solute interactions, and should not be ignored.