Applying Computational Scoring Functions to Assess Biomolecular Interactions in Food Science: Applications to the Estrogen Receptors

Abstract

During the last decade, computational methods, which were for the most part developed to study protein-ligand interactions and especially to discover, design and develop drugs by and for medicinal chemists, have been successfully applied in a variety of food science applications [1,2]. It is now clear, in fact, that drugs and nutritional molecules behave in the same way when binding to a macromolecular target or receptor, and that many of the approaches used so extensively in medicinal chemistry can be easily transferred to the fields of food science. For instance, nuclear receptors are common targets for a number of drug molecules and could be, in the same way, affected by the interaction with food or food-like molecules. Thus, key computational medicinal chemistry methods like molecular dynamics can be used to decipher protein flexibility and to obtain stable models for docking and scoring in food-related studies, and virtual screening is increasingly being applied to identify molecules with potential to act as endocrine disruptors, food mycotoxins, and new nutraceuticals [3,4,5]. All of these methods and simulations are based on protein-ligand interaction phenomena, and represent the basis for any subsequent modification of the targeted receptor's or enzyme's physiological activity. We describe here the energetics of binding of biological complexes, providing a survey of the most common and successful algorithms used in evaluating these energetics, and we report case studies in which computational techniques have been applied to food science issues. In particular, we explore a handful of studies involving the estrogen receptors for which we have a long-term interest.