IN SILICO TARGET IDENTIFICATION FOR QUINOLINE DERIVATIVES WITH POTENTIAL ANTILEISHMANIAL ACTIVITY

Quinolines form a group of compounds present in several natural products which have been researched in numerous works in the area of organic synthesis, having demonstrated high antimalarial, antibacterial, antifungal, antitumor and antituberculosis potential, in addition to being pointed out as promising antileishmanial agents. However, the literature on structure-based studies for leishmania targets in general and specifically for these ligands is still quite incipient, so that most of the studies that seek to identify possible biological activities for these compounds are based on trial and error approaches. The present work aims to identify biological targets with high affinities for a set of eight 2-aryl-quinoline-4-carboxylic acids synthesized in our research group with potential activity against Leishmania donovani strains. This parasite species causes one of the most severe forms of leishmaniasis, known as visceral leishmaniasis, responsible for high mortality rates around the world. Two approaches were used: (i) the search in target fishing servers based on similarity principles (SwissTargetPrediction), molecular fingerprints (SEA), molecular 3D similarity (ChemMapper), and deep learning-based algorithm (TargetNet), as well as (ii) molecular docking studies of the structures in four L. donovani enzymes with determined crystallographic structures available at PDB. Docking simulations were performed using the programs Autodock, AutoDock Vina, Gold and DockThor. The results of both approaches indicated the enzyme Dihydroorotate dehydrogenase as the target with the highest affinity for the ligands under study. Redocking studies were performed with the same parameters for L. major enzymes, which are the only ones with crystallographic complexes available in PDB, sharing more than 90% of sequence identity to the L. donovani enzymes under study. These studies indicated that AutoDock Vina provides the results with lower RMSD values for predicted ligand poses. Also, the ligand-receptor affinities estimated by this software presented the most consistent results in agreement with molecular docking data available in the literature. These findings will guide our undergoing structure-based development of derivatives with improved chemical properties to further optimize the therapeutic potential of quinolines as antileishmanial agents.

This work was supported by National Institute of Science and Technology on Molecular Sciences (INCT-CiMol), Grant CNPq 406804/2022-2.