Reposicionamento de fármacos para Leishmania spp: estratégias “in silico” e avaliação experimental

• Autor(a) principal: Silva, Diego Cabral
• Data de Publicação: 2015
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Institucional da UFG
• Texto Completo: http://repositorio.bc.ufg.br/tede/handle/tede/11605
• topic: Leishmaniose; Reposicionamento de fármacos; Bioinformática; QSAR; Triagem virtual; Ensaios in vitro; Leishmaniasis; Drug repositioning; Bioinformatics; QSAR; Virtual screening; In vitro assays; CIENCIAS BIOLOGICAS::PARASITOLOGIA::PROTOZOOLOGIA DE PARASITOS

Leishmaniases are diseases caused by more than 20 protozoan parasites belonging to the genus Leishmania, and transmitted through the bite of infected female Phlebotomine and Lutzomya. An estimated 20,000 to 30,000 deaths and 1.3 million new cases occur annually. Currently available drugs have serious limitations regarding their efficacy and especially their toxicity, side effects and costs. Thus, there is a pressing need for new therapies that are safer and more effective. Due to the high costs of traditional process of drug discovery and development, alternative strategies have been developed to speed up this process, and reducing its costs. Among them, is drug repositioning, which is the discovery of new therapeutic applications for drugs already on the market. The aim of this work was to search and identify approved and clinically available drugs with potential antileishmanial activity, using bio- and cheminformatics approaches, and experimental validation of these drugs using in vitro assays. Initially, we generated a dataset of Leishmania genes with orthologs in four species (L. major, L. braziliensis, L. infantum e L. mexicana), specific to the genus Leishmania, trying to explore potential molecular targets that could be effective against all species and essential to the process of development and differentiation of the parasite. This dataset was used to interrogate three databases of approved drugs (DrugBank and TTD) aiming to identify homologues of validated targets for other diseases. Furthermore, binary QSAR models were generated from phenotypic assay data, using different descriptors, and two machine learning methods, and then consensus models were built. Homology search allowed the identification of 36 new potential molecular targets that need to be validated experimentally, and 122 drugs. Of these 122 compounds, 28 were previously reported on the literature as actives. Five drugs not yet tested were selected for biological screening in vitro against promastigotes (lansoprazole, ibuprofen, sertaconzole, nilutamide and clomifen). Three of them showed activity at 100 μM and we determined their IC50. Ibuprofen showed an IC50 of 55.08 μM, sertaconazole IC50 < 15 μM and clomifen (IC50 5,75 μM, more potent than the standard drug (pentamidine IC50 = 7,24), suggesting a potential activity. Besides that, the QSAR models generated had adequate statistical parameters, especially for consensus models. One of the models generated by consensus was employed to predict the activity of the drugs identified by the bioinformatics approach. The best models can be used as filters in a virtual screening process. In vitro assays in the promastigote form of L. amazonensis were standardized, and used to successfully identify new potential candidates for drug repositioning.