(747f) Sequestering of Toxins for the Reversal of Drug Toxicity: A Coarse Grained Molecular Dynamics Study

Eighty seven people die every day in the United States as a result of an unintentional poisoning.   Of poisoning mortalities, more than ~90% are due to drug intoxications.  In vivo experiments and rare clinical reports suggest that intravenous delivery of a lipid emulsion facilitates reversal of drug toxicity and rapid restoration of compromised cardiac function. Patients unresponsive to other resuscitation measures have regained a spontaneous and sustained cardiac output in as little as 60 seconds following a lipid intervention. However, the mechanism of action of this therapy is poorly understood. Consequently, it is not possible to predict which toxins would be susceptible to lipid therapy.

The most commonly cited mechanism of action – known as the 'lipid sink' – suggests that lipophilic toxins associate with lipid emulsion droplets via hydrophobic partitioning. However, many of the compounds in question exist in a predominantly charged state at physiological pH. Moreover, acidosis may occur during cardiac arrest, further reducing blood pH levels and increasing the degree of molecular protonation. Hence, electrostatic interactions likely play an important role in driving the sequestration of toxins. In this work, we aim to clarify the relative roles of hydrophobicity and charge in mediating the association of toxin molecules with lipid droplets.

Coarse grained molecular dynamics simulations utilizing the MARTINI force field are employed to develop molecular topologies for the multiphase system of interest (an aqueous/oil system with phospholipid monolayer interface). The MARTINI scheme has been shown to semi-quantitatively reproduce the fundamental structural and thermodynamic properties of biomacromolecules such as proteins, carbohydrates, and phospholipids. However, similar topologies for small molecules are not as well explored. Here, neutral and protonated coarse grained topologies are developed for several small molecules with varying levels of lipophilicity and differing degrees of protonation at physiological pH. These topologies are systematically developed based on the three-dimensional molecular structures of the toxins of interest. A parameterization scheme is developed and validated by computationally reproducing the experimentally determined octanol/water partition coefficients of toxin compounds. These topologies are then incorporated into the multiphase 'droplet' system for the purposes of our study.