(169ae) Molecular Dynamics Simulations of the Tear Film Lipid Layer to Elucidate the Causes of Dry Eye Syndrome

Dry Eye Syndrome (DES) affects nearly 20 million people in the United States, leading to a loss in quality of life due to constant irritation of the eyes. The most common cause of DES is defects in the tear film lipid layer (TFLL) which is a <100 nm thick layer of lipids that separates tears from the ambient air. The lipids in the TFLL originate in the meibomian glands in the eyelids. The TFLL is hypothesized to disperse the tear over the surface of the eye, to prevent airborne particles from entering the tears, and to prevent tear vaporization. When the TFLL is malformed due disfunctions of the meibomian glands, the TFLL no longer functions as intended, and patients experience a dry sensation in the eyes. The composition, properties, and precise balance of the TFLL structure directly leads to DES. Thus, understanding the structural and functional attributes of TFLL is essential for grasping the functioning of the tear film under both normal and pathological circumstances. While the chemical composition of the TFLL is well-established, the layer-by-layer structure of the TFLL is unknown. The structuring of these layers is key to developing DES drugs and therapies. Experimental methods for probing the TFLL are limited due to ethical considerations of experimenting on living eyes. Therefore, we have used molecular dynamics (MD) simulations to test hypothetical TFLL structural arrangements and support or refute prevailing published hypotheses.

Structures of 100 different lipids from the lipid classes such as wax esters, cholesterol esters, OAHFAs, phospholipids, sphingolipids, and glycerides. CHARMM36m and the CHARMM General Force Field was used to model common and uncommon lipids. Lipids were packed together using Packmol to follow hypothetical TFLL structures. Five different arrangements of the TFLL were constructed, and the models contained nearly 1 million atoms. MD simulations of at least 1 μs were conducted to allow for structural relaxation and to provide estimations of properties that have been measured experimentally. Simulations were thus validated against experimental measurements of surface tension, viscosity, rotamer distribution, and lipid crystallinity. Based upon these simulations, some of the prevailing hypotheses have been refuted, leading to a reduced set of plausible TFLL structures. These findings provide plausible atomistic structures that allow for future in silico testing of drug molecules that can modify the properties of the TFLL, thereby restoring function and providing relief to millions of patients.