Search for Zinc Complexes with High Affinity in Pyrazinamidase from Mycobacterium Tuberculosis Resistant to Pyrazinamide

• Jesus Antonio Alvarado-Huayhuaz Universidad Nacional de Ingeniería (UNI) / Universidade Federal do Rio Grande (FURG) https://orcid.org/0000-0003-4912-3178
• Daniel Alonso Talaverano-Rojas Universidad Nacional de Ingeniería (UNI) https://orcid.org/0000-0002-7963-0886
• Reneé Isabel Huamán Quispe Universidad Nacional de Ingeniería (UNI) https://orcid.org/0000-0003-2116-7478
• Maurício Dorneles Caldeira Balboni Universidade Federal do Rio Grande (FURG) https://orcid.org/0000-0001-7653-0962
• Oscar Emilio Arrúa Arce Universidade Federal do Rio Grande (FURG) https://orcid.org/0000-0003-2603-1940
• Adriano Velasque Werhli Universidade Federal do Rio Grande (FURG) https://orcid.org/0000-0001-7107-5024
• Karina dos Santos Machado Universidade Federal do Rio Grande (FURG) https://orcid.org/0000-0002-8966-5708
• Ana Cecilia Valderrama-Negrón Universidad Nacional de Ingeniería (UNI) https://orcid.org/0000-0001-7741-3207

Resumo

Tuberculosis is an ancient and current disease. Resistance to the prodrug pyrazinamide (PZA), one of the most important antituberculosis drug, is often associated with various mutations in the pncA gene that expresses the metalloenzyme pyrazinamidase (PZase). Some hard and intermediate acids, such as Co(II), Mn(II), and Zn(II), showed the ability to partially recover the susceptibility to PZA in resistant strains. In this work, we investigate the affinity that zinc complexes can achieve in the PZase protein with a low affinity for pyrazinamide. First, we select the PZase mutant with the best resistance profile to PZA using the web-server SUSPECT-PZA and a home-made script. Then we use the tmQM database, which contains 86,665 metal complexes with crystallographic structures and quantum descriptors, to search for zinc complexes with high affinity for PZase. Out of 5867 Zn complexes, 100 with lower dipole moment, higher hardness and lower HOMO energy were selected. Molecular docking studies using (a) empirical scoring functions (SF) and (b) SF based on machine learning, allowed us to find complexes such as BUXZUQ, FEQTUS or DOSQUA that have higher affinity for PZase than PZA. These Zn complexes not only exhibit higher global reactivity compared to PZA, but are also very similar to each other, and to a lesser degree their organic part is also similar to that of PZA. The compounds we have reported can serve as a basis for the design of new antituberculosis metallodrugs.

Palavras-chave: Zn complexes, Pyrazinamidase, Molecular docking, Machine learning