Receptor de Canabinoide 2. novo alvo terapêutico para tratamento da osteoporose : uma análise da interação com ligantes através da aplicação de ferramentas da bioinformática

Abstract

Cannabinoid receptors are widely distributed in animal tissues and are classified as types 1 and 2. It is known that these receptors are responsible for numerous functions in healthy tissues and their structure is preserved over time and species. The cannabinoid receptor 1, which recently had its structure resolved by Xray diffraction crystallography, is present largely in the central nervous system - being responsible for functions such as managing energy expenditure, memory and learning. We are also aware of its role in various disorders of the systems such as the manifestations of chronic pain, epilepsy and neurodegenerative diseases. Although also present in the central nervous system, the cannabinoid receptor 2 is more frequently found in hematopoietic tissue and in the immune system, being responsible for controlling some of the processes related to bone metabolism and in routes of inflammatory activity. Both receptors belong to the family of receptors linked to the G protein in a widely studied system and today known as the endocannabinoid system. The use of bioinformatics tools helps to elucidate therapeutic targets to control the expression of these receptors in order to solve some deficiencies in the treatment of complex chronic diseases. Recently, the crystallographic structure of the Cannabinoid Receptor 1 was resolved with high resolution. Because it is a membrane receptor, there are intrinsic difficulties for structural determination of this protein. In a short period, two structures of the receptor linked to agonists and two others linked to antagonists were made available in the Protein Data Bank database. Most recently, the three-dimensional structure of the type 2 cannabinoid receptor has been resolved and this advance made room for, with some certainty, the use of resolved structure to find the active protein cavities for docking and determining possible therapeutic targets for the treatment of osteometabolic diseases, like osteoporosis, and chronic inflammatory diseases. We used a computational methodology to identify protein binding sites and docking to test small molecules as receptor agonists and antagonists. Molecular docking approaches and machine learning were used to investigate protein-ligand interactions and to create a proteinspecific score function to accurately predict other possible interactions should they be evaluated in the future.