Garbage in, garbage out: how reliable training data improved a virtual screening approach against SARS-CoV-2 MPro.

Introduction: The identification of chemical compounds that interfere with SARS-CoV-2 replication continues to be a priority in several academic and pharmaceutical laboratories. Computational tools and approaches have the power to integrate, process and analyze multiple data in a short time. However, these initiatives may yield unrealistic results if the applied models are not inferred from reliable data and the resulting predictions are not confirmed by experimental evidence. Methods: We undertook a drug discovery campaign against the essential major protease (MPro) from SARS-CoV-2, which relied on an in silico search strategy -performed in a large and diverse chemolibrary- complemented by experimental validation. The computational method comprises a recently reported ligand-based approach developed upon refinement/learning cycles, and structure-based approximations. Search models were applied to both retrospective (in silico) and prospective (experimentally confirmed) screening. Results: The first generation of ligand-based models were fed by data, which to a great extent, had not been published in peer-reviewed articles. The first screening campaign performed with 188 compounds (46 in silico hits and 100 analogues, and 40 unrelated compounds: flavonols and pyrazoles) yielded three hits against MPro (IC50 ≤ 25 µM): two analogues of in silico hits (one glycoside and one benzo-thiazol) and one flavonol. A second generation of ligand-based models was developed based on this negative information and newly published peer-reviewed data for MPro inhibitors. This led to 43 new hit candidates belonging to different chemical families. From 45 compounds (28 in silico hits and 17 related analogues) tested in the second screening campaign, eight inhibited MPro with IC50 = 0.12-20 µM and five of them also impaired the proliferation of SARS-CoV-2 in Vero cells (EC50 7-45 µM). Discussion: Our study provides an example of a virtuous loop between computational and experimental approaches applied to target-focused drug discovery against a major and global pathogen, reaffirming the well-known "garbage in, garbage out" machine learning principle.

Preclinical Studies and Drug Combination of Low-Cost Molecules for Chagas Disease.

Chagas disease is caused by the protozoan Trypanosoma cruzi (T. cruzi). It remains the major parasitic disease in Latin America and is spreading worldwide, affecting over 10 million people. Hundreds of new compounds with trypanosomicidal action have been identified from different sources such as synthetic or natural molecules, but they have been deficient in several stages of drug development (toxicology, scaling-up, and pharmacokinetics). Previously, we described a series of compounds with simple structures, low cost, and environmentally friendly production with potent trypanosomicidal activity in vitro and in vivo. These molecules are from three different families: thiazolidenehydrazines, diarylideneketones, and steroids. From this collection, we explored their capacity to inhibit the triosephosphate isomerase and cruzipain of T. cruzi. Then, the mechanism of action was explored using NMR metabolomics and computational molecular dynamics. Moreover, the mechanism of death was studied by flow cytometry. Consequently, five compounds, 314, 793, 1018, 1019, and 1260, were pre-clinically studied and their pharmacologic profiles indicated low unspecific toxicity. Interestingly, synergetic effects of diarylideneketones 793 plus 1018 and 793 plus 1019 were evidenced in vitro and in vivo. In vivo, the combination of compounds 793 plus 1018 induced a reduction of more than 90% of the peak of parasitemia in the acute murine model of Chagas disease.

Potent and Selective Inhibitors of Trypanosoma cruzi Triosephosphate Isomerase with Concomitant Inhibition of Cruzipain: Inhibition of Parasite Growth through Multitarget Activity.

Triosephosphate isomerase (TIM) is an essential Trypanosoma cruzi enzyme and one of the few validated drug targets for Chagas disease. The known inhibitors of this enzyme behave poorly or have low activity in the parasite. In this work, we used symmetrical diarylideneketones derived from structures with trypanosomicidal activity. We obtained an enzymatic inhibitor with an IC50 value of 86 nm without inhibition effects on the mammalian enzyme. These molecules also affected cruzipain, another essential proteolytic enzyme of the parasite. This dual activity is important to avoid resistance problems. The compounds were studied in vitro against the epimastigote form of the parasite, and nonspecific toxicity to mammalian cells was also evaluated. As a proof of concept, three of the best derivatives were also assayed in vivo. Some of these derivatives showed higher in vitro trypanosomicidal activity than the reference drugs and were effective in protecting infected mice. In addition, these molecules could be obtained by a simple and economic green synthetic route, which is an important feature in the research and development of future drugs for neglected diseases.

Expanding the family of heteroleptic oxidovanadium(IV) compounds with salicylaldehyde semicarbazones and polypyridyl ligands showing anti-Trypanosoma cruzi activity.

Searching for prospective vanadium-based drugs for the treatment of Chagas disease, a new series of heteroleptic [V(IV)O(L-2H)(NN)] compounds was developed by including the lipophilic 3,4,7,8-tetramethyl-1,10-phenanthroline (tmp) NN ligand and seven tridentate salicylaldehyde semicarbazone derivatives (L1-L7). The compounds were characterized in the solid state and in solution. EPR spectroscopy suggests that the NN ligand is bidentate bound through both nitrogen donor atoms in an axial-equatorial mode. The EPR and (51)V-NMR spectra of aerated solutions at room temperature indicate that the compounds are stable to hydrolysis and that no significant oxidation of V(IV) to V(V) takes place at least in 24h. The complexes are more active in vitro against Trypanosoma cruzi, the parasite responsible for Chagas disease, than the reference drug Nifurtimox and most of them are more active than previously reported [V(IV)O(L-2H)(NN)] complexes of other NN co-ligands. Selectivity towards the parasite was analyzed using J-774 murine macrophages as mammalian cell model. Due to both, high activity and high selectivity, L2, L4, L5 and L7 complexes could be considered new hits for further drug development. Lipophilicity probably plays a relevant role in the bioactivity of the new compounds. The [V(IV)O(L-2H)(NN)] compounds were designed aiming DNA as potential molecular target. Therefore, the novel L1-L7 tmp complexes were screened by computational modeling, comparing their DNA-binding features with those of previously reported [V(IV)O(L-2H)(NN)] compounds with different NN co-ligands. Whereas all the complexes interact well with DNA, with binding modes and strength tuned in different extents by the NN and semicarbazone co-ligands, molecular docking suggests that the observed anti-T. cruzi activity cannot be explained upon DNA intercalation as the sole mechanism of action.

Molecular docking and molecular dynamics simulation studies of Trypanosoma cruzi triosephosphate isomerase inhibitors. Insights into the inhibition mechanism and selectivity.

Trypanosoma cruzi (T. cruzi) triosephosphate isomerase (TcTIM) is a glycolytic enzyme essential for parasite survival and has been considered an interesting target for the development of new antichagasic compounds. The homodimeric enzyme is catalytically active only as a dimer. Interestingly, significant differences exist between the human and parasite TIMs interfaces with a sequence identity of 52%. Therefore, compounds able to specifically disrupt TcTIM but not Homo sapiens TIM (hTIM) dimer interface could become selective antichagasic drugs. In the present work, the binding modes of 1,2,4-thiadiazol, phenazine and 1,2,6-thiadiazine derivatives to TcTIM were investigated using molecular docking combined with molecular dynamics (MD) simulations. The results show that phenazine and 1,2,6-thiadiazine derivatives, 2 and 3, act as dimer-disrupting inhibitors of TcTIM having also allosteric effects in the conformation of the active site. On the other hand, the 1,2,4-thiadiazol derivative 1 binds into the active site causing a significant decrease in enzyme mobility in both monomers. The loss of conformational flexibility upon compound 1 binding suggests that this inhibitor could be preventing essential motions of the enzyme required for optimal activity. The lack of inhibitory activity of 1 against hTIM was also investigated and seems to be related with the high mobility of hTIM which would hinder the formation of a stable ligand-enzyme complex. This work has contributed to understand the mechanism of action of this kind of inhibitors and could result of great help for future rational novel drug design.