Text

theses

In this master thesis, exploration is done in employing molecular modelling methods to predict drug solubility in liquid solvents, compare the simulation results with experimental data to validate the molecular force field model, and explain certain solubility behaviour of the statin compounds. Experimental data shows that lovastatin solubility increases in a family of alkanols and reaches a peak as the non-polarity nature of the solvent increases from water (polar) to 1-butanol (nonpolar), and the trend reverses from 1-pentanol to 1-octanol. This study investigates this interesting behaviour and provides insight on why this can be occurring. In this study, the CHARMM General Force Field is utilized to model lovastatin and simvastatin in different liquid solvents. The free energy and thermodynamic properties of these systems are calculated using molecular dynamics techniques. Periodic boundary conditions are used with electrostatics treated with Particle-mesh Ewald (PME), using a short-range cutoff of 1.2 nm, while having van der Waals interactions switched off between 1.0 to 1.2 nm. The Bennet Acceptance Ratio (BAR) method is implemented as a means of estimating the Gibbs free energy of decoupling the drug molecule in the system. We report results obtained from two studies. In Study 1, the free energy of de-coupling of a drug molecule in different liquid solvents is computed. These results yield the ratio of the infinitely dilute activity coefficient and solubility of the drug compounds in two different solvents (including water and a family of alcohols up to 1-octanol) The simulation results are observed to predict a peak in the lovastatin solubility in alkanols, with a peak for 1-butanol. The simulated data is analyzed further by calculating the energies between polar and nonpolar groups between the drug and solvent. In Study 2, we employ a specific thermodynamic model that involves mutating the unique methyl group in simvastatin to hydrogen, essentially converting simvastatin to lovastatin in liquid solvents and vacuum. The free energy of mutation can be shown to be related to the activity coefficients of lovastatin and simvastatin in a solvent. Results from Study 1 showed that the simulated lovastatin solubility ratios are in agreement with experimental data in that a solubility peak in 1-butanol is estimated (Fig1). Further analysis of the energy of interactions of the polar and nonpolar groups between the drug and solvent molecules shows that the nonpolar interactions become stronger with increasing alcohol carbon chain length. However, the interactions between the polar functional groups of the drug and solvent appear to reach a peak in strength at 1-butanol and seem to reverse trend (Fig2). The polar and polar interactions contribution to the overall enthalpy thus changes direction and provides a weakening effect after the organic alcohol chain length surpasses that of 1-butanol. Results obtained from Study 2 show that the simulation provides a reasonably adequate prediction of the activity coefficient ratios between the two drugs within the same solvent. In this study, it was shown that CHARMM General Force Field is a reasonably good model for lovastatin, simvastatin and common liquid solvents. Our simulation study was found to give good estimates and capture the behavior of the solubility of lovastatin in liquid solvents. Once the drug solvent mixtures were further analyzed and broken into groups based on polarity, analysis on the energy of interactions provides scientific insights and explanation on why solubility of lovastatin reaches a peak with increasing organic alcohol chain length.

ETD_4966

M.S.

Includes bibliographical references

Includes vita

by Fabian F. Casteblanco

doi:10.7282/T3N58JDR

ETD graduate

The author owns the copyright to this work.