Computer-aided search of novel inhibitors of n-myristoyl transferase with trypanocidal effects

LUCAS NICOLÁS ALBERCA; MARÍA SBARAGLINI; MARÍA DANIELA RUIZ; SALOMÉ VILCHEZ LARREA; FERNÁNDEZ VILLAMIL, SILVIA H.; CAROLINA CARRILLO; ALAN TALEVI

Buenos Aires

Congreso; 18th International Congress on Infectious Diseases; 2018

International Society for Infectious Diseases

PurposeThe aim of the project is to find new treatments for the American and AfricanTrypanosomiasis using an interesting strategy: drug repositioning. It consists in the search of new therapeutic indications for already known drugs. N-myristoyl transferase(NMT) has been identified as a druggable drug target in trypanosomatids. We report the experimental assay against Trypanosoma cruzi and Trypanosoma brucei of potential NMT inhibitors emerging from a virtual screening campaign.Methods and MaterialsAn ensemble of ligand-based linear classifiers capable of discriminating within NMT inhibitors and non-inhibitors was inferred from a 224 training set. After internal and external validation, the model ensemble was applied to screen the DrugBank 4.0 database, using positive predictive value (PPV) surfaces to optimize the score threshold and applicability domain estimations. Three hits were acquired and experimentally tested: dicyclomine (previously used for the treatment of irritable bowel syndrome), quinestrol (used in hormone replacement therapy) and danazol (for the treatment of endometriosis and fibrocystic breast disease). The effect of different concentrations of the candidates on T. cruzi epimastigotes was assayed and the correspondent EC50 calculated. Also, theviability of T. cruzi trypomastigotes at 20µM of the candidates were tested. Finally, the effects of the compounds on T. brucei trypomastigotes was measured.ResultsThe three tested drugs showed a strong trypanocidal activity on T. cruzi epimastigotes. Dicyclomine, quinestrol and danazol showed an EC50 of 4.2µM, 4.8µM and 1.8µM respectively. Conversely, the inhibition of T. cruzi trypomastigotes was moderate: only quinestrol reduced the viability more than 50% at 20 µM. The calculated EC50 for quinestrol on these forms of the parasites was 8µM. Finally, these drugs showed EC50s against T. brucei procyclic and bloodstream trypomastigotes between 10-20µM.ConclusionStarting from a database of compounds previously evaluated against trypanosomatid N-Miristoil Transferase (NMT), we have used an in silico strategy to generate models capable of identifying drugs with potential trypanocidal activity. The assembly of models has shown an improvement in the predictive capacity with respect to individual models. Keeping in mind the drug repositioning strategy, we have applied the best model assembly in the Virtual Screening of the DrugBank database, finding 33 approved drugs as potential trypanocidal drugs. Three of these candidates were acquired and tested over T. cruzi and T. brucei. All of them, Danazol, Dicyclomine and Quinestrol possesstrypanocidal activity against these parasites. Our results demonstrate the value of the in silico screening since we detected 3 posible candidates with minium investment of time and resourses. This strategy is a very valuable tool that could provide solutions for Neglected Diseases.