Chemoinformatic tools applied for the screening of inhibitors of the enzyme DHODH from Leishmania major

Abstract

The application of informatics methods to solve chemical problems is called chemoinformatics. A large variety of chemoinformatic tools have been developed in the last years, covering from structural representation to data analysis methods. Drug discovery can be based on screening chemical compounds, regarding their physicochemical properties, applying chemoinformatic tools. Quantitative Structure-Activity Relationships (QSAR) modeling defines the activity/property as a function of calculated descriptors. Model building techniques can be unsupervised (e.g. PCA), supervised with regression (e.g. PLS) and supervised without regression (e.g. ANN). Neglected Tropical Diseases (NTDs) affect more than a billion people worldwide, yet few effective drugs are available to treat these diseases, especially those caused by trypanosomatids, such as Leishmaniasis (Leishmania spp.). The flavoenzyme dihydro-orotate dehydrogenase (DHODH) catalyzes the fourth reaction of the de novo pyrimidine biosynthetic pathway and by that it is a potential molecular target for antiprotozoal drugs. Secondary metabolites isolated from species of Asteraceae have antiprotozoal activity, highlighting the sesquiterpene lactones (STL), which are taxonomic markers of this plant family. The present BEPE proposal is grounded by the general objective from the main research project (2014/01443-6), that is the screening of plant extracts and isolated compounds from Asteraceae for in vitro inhibitory potential against L. major DHODH (LmDHODH). The objective of the current proposal is therefore to establish QSAR models to predict and explain the relationships between structural information of chemical compounds and their inhibitory activity against DHODH. The bioassays were already performed. The pIC50 values are distributed homogeneously within a proper range (2.5-4.5, delta = 2 log units, SD < 5%), thus providing representation and statistical reliability to generate QSAR models. These promising results for in vitro inhibition of LmDHODH for 59 natural products (NP) are unprecedented in the literature. The in silico methods for the QSAR study will involve: a) molecular modeling and geometry optimization; b) descriptors generation and selection; c) model building and validation. Until now, no QSAR study has been performed aiming this enzyme. Besides understanding the most important molecular properties that provides inhibitory activity against LmDHODH, what can lead to new candidates for the treatment of Leishmaniasis; the novelty of this study is that it can be used as a template for developing further models using different classes of NP and also other DHODHs. Applying widely used chemoinformatic tools, this 9 month-research internship abroad will provide significant progress to achieve part of the objective of the main research project and, beyond that, will raise the quality and dimension of its results. The foreign supervisor will be Prof. Dr. Thomas J. Schmidt, from Münster University (Germany). The collaboration with this researcher was chosen because of his great expertise and background in QSAR, NTDs and STL. (AU)