Computational pharmaceutics using molecular simulations
Title: Computational pharmaceutics using molecular simulations
DNr: SNIC 2014/1-349
Project Type: SNIC Medium Compute
Principal Investigator: Christel Bergström <christel.bergstrom@farmaci.uu.se>
Affiliation: Uppsala universitet
Duration: 2014-12-01 – 2015-10-01
Classification: 30101 10407 30108
Homepage: http://www.farmfak.uu.se/farm/lmformul/research.html#ddr8
Keywords:

Abstract

Oral administration of drugs requires sufficient solubility of the active ingredient in gastric or intestinal media to allow complete absorption. Modern drug discovery techniques often generate lead compounds with low aqueous solubility, resulting in an inefficient and expensive drug development process, since time and resources are spent on drug candidates that in the end cannot easily be administered in vivo. If lipophilic, these drug compounds can for instance be solubilized in endogenous and dietary aggregates such as micelles and vesicles, consisting of mixed lipids such as lecithin and bile, resulting in a significant increase of the in vivo solubility. This study aims to increase the understanding for the molecular basis in solubilisation of poorly soluble compounds in physiological fluids containing lipids, using molecular dynamics (MD) simulations with the Gromacs code. In detail, we will use the Berger-Edholm force field for lipids (Berger, Edholm, Jähnig,1997) in order to study mixed lecithin/taurocholate micelles and liposome aggregates, which are also present in simulated and aspirated human intestinal fluids. These simulated intestinal fluids (SIFs) have been characterized extensively experimentally in terms of lipid aggregate composition, concentration and size, and are commonly used in experimental solubility studies as a part of the drug development process, both by industry and academia. If the mechanisms behind the solubilisation in these lipid aggregates could be understood on the molecular level, knowledge-driven decisions about solubility-related issues could be made much earlier during the drug development process than the current setting is allowing. As a continuation from our first 6 month allocation (40000 core-h/month) we will probe the phase stability of mixed micelles and bilayers of specific compositions of the lecithin molecule DLiPC (1,2-dilinoleoyl-sn-glycero-3-phosphocholine) as well as the bile salt taurocholate, since they are the main lipid constituents in SIFs. The intention is to explore the effect of variation in composition on phase stability. In a second subproject we will explore the interaction of poorly soluble drug molecules and bile molecules themselves with the characterized micelles and bilayers assemblies, using free-energy calculations through umbrella sampling simulations. This will be done by calculating potential of mean force (PMF) along the relevant reaction coordinate using the gromacs utility g_wham (Hub, de Groot & van der Spoel, 2010), using the methodology and work-flow developed during our previous allocation. These calculations will hopefully yield the kinetic barriers to loading/unloading of drugs from the assemblies as well as the partitioning of the drug molecules in the interior of the lipid assemblies. In parallel, we will also calculate the theoretical water/octanol partitioning from similar Umbrella sampling simulations of biphasic octanol/water systems, as well from FEP simulations using the multi-state Bennet acceptance ratio (Shirts & Chodera, 2008) on octanol resp. water systems. As a final goal, the results from these MD simulations will be compared with available experimental data on both the lipid assemblies alone and assemblies equilibrated with the drug.