Molecular Level Insights into Taste Perception and Beyond

Taste perception is a complex and multi-layered process which involves several components from the molecular level up to subcellular, cellular, and tissue levels. At the molecular scale, taste perception is triggered by interactions between food compounds and taste receptors located on the tongue's papillae. Taste is the primary driver of food intake since the five basic taste sensations, i.e. sweet, bitter, umami, sour and salty, are related to specific nutritional needs or control strategies. Understanding the molecular features and mechanisms that determine the activation of taste receptors and the resulting gustatory sensation is crucial to comprehend why specific substances are perceived with a particular taste and how food consumption is regulated. Moreover, due to the interconnected relationship between food intake and health status, the investigation of the molecular effects of food tastants on secondary actors involved in specific diseases or interested functions appears particularly fascinating. In a broader context, this comprehensive knowledge has the potential to pave the way towards, for example, designing the taste of foods, creating ingredients that are less harmful to health, controlling food quality, improving the intake of drug treatments by enlightening their taste, or engineering personalized diets coupled with traditional pharmacological treatments to target molecular types of machinery involved in specific diseases. In recent years, molecular modelling and machine learning have emerged as fundamental methodologies for elucidating the molecular properties that underlie specific macromolecular functions. In the context of this doctoral thesis, we have applied these methodologies to investigate taste-related molecular actors. Molecular modelling has been employed herein to establish a computational framework aimed at characterizing the interactions between a receptor and its agonist and search for similar binding pockets in protein databases of interest. We applied our methodology to a human bitter taste receptor bound with its agonist to screen the complete repertoire of solved human proteins for potential off-targets that possess similar binding sites. Starting from the methodology just mentioned, we subsequently explored the effects of natural compounds on the molecular structure and dynamics of specific proteins implicated in neurodegenerative diseases. Conversely, machine learning has been employed within a ligand-centred perspective to comprehend the physiochemical properties of food tastants that underlie their taste. To this end, we have developed specialized machine learning- based tools to predict three fundamental taste modalities, namely umami, bitter, and sweet, of a given molecule based on its molecular structure. These findings have been or will soon be incorporated into the web platform (https://virtuous.isi.gr) which has been developed as a component of the VIRTUOUS project (https://virtuoush2020.com). The project, funded by the European Union (EU), strives to establish a comprehensive platform that amalgamates various levels and methodologies of investigation to predict the organoleptic profile of Mediterranean ingredients based on their chemical composition. This project aims to advance our comprehension of how the chemical structure of food influences our perception of taste, encompassing both the molecular realm and intricate sensory encounters that contribute to the overall taste profile. In summary, employing a computational approach that integrates molecular modelling and machine learning, the current doctoral thesis has yielded insights into the molecular foundations of taste perception and its potential impact on secondary targets. This work serves as a foundational step towards a comprehensive, multi-level, and interdisciplinary exploration of taste, with the overarching goal of unravelling the intricate processes that link taste perception, food intake, and overall health status.