Computational repositioning of bioactive compounds from large chemogenomic screens: identification of conserved druggable modules between yeasts and trypanosomes

MERCEDES DIDIER GARNHAM; LIONEL URAN LANDABURU; EMIR SALAS SARDUY; AGÜERO, FERNÁN

Capital Federal

Simposio; XXI Simposio Internacional Mundo Sano; 2023

Fundación Mundo Sano

Introduction. Detailed characterization of the cellular response to chemicals is fundamental to understand the mechanism of action of drugs. One strategy to do this is to assess growth (fitness) of gene mutants exposed to different drugs. Recently, a number of genome‐wide fitness profiling assays were performed on Saccharomyces cerevisiae. These screens were based on whole-genome collections of heterozygous and homozygous deletions (Barazandeh M et al, 2022), and constitute a rich source of pharmacogenomic associations between drugs and genes. In contrast, these associations are scarce in trypanosomes. In this work we propose the use of these yeast chemogenomic data as starting points to guide repurposing opportunities (Urán Landaburu et al, 2023). Materials and methods. We have curated and standardized yeast-based chemogenomic assays from published studies, and developed an orthology mapping pipeline. For drugs, the pipeline contained filters to retain compounds that were drug-like, novel, commercially available, and with low potential promiscuity. For genes, we selected those with orthologs in T. brucei that carry significant fitness phenotypes when knocked down (through an orthology mapping between T. cruzi genes and T. brucei whole-genome RNAi essentiality assays in Alsford et al, 2011). Selected compounds were purchased to assess their antiparasitic activity and cytotoxicity in vitro.Results: We obtained 271,955 gene-drug associations, with a set of 5,811 unique genes and 2,935 unique drugs. After standardization and application of filters in the pipeline we obtained a final library of 194 compounds, linked to 73 T. cruzi candidate protein targets. We purchased 21 of these compounds and found 3 that exhibited micromolar activity on T. cruzi amastigotes and negligible cytotoxicity on Vero host cells. The developed chemogenomic pipeline provided a significant enrichment in bioactive compounds.Conclusions: A modular repositioning pipeline was built to leverage chemogenomic data from yeast. This pipeline can be re-used for other chemogenomic datasets, and for other diseases. References: Alsford S (2011) Genome Research 21: 915; Barazandeh M (2022) BMC Genomics 23: 197; Urán Landaburu L (2023) Biochem. Soc. Trans. 51: 195.