Identification of Potential Trypanosoma cruzi Trans-Sialidase Inhibitors by Computational Drug Repositioning Approaches

Abstract

Chagas disease, caused by the parasitic protozoan Trypanosoma cruzi (T. cruzi), represents a worldwide public health issue. To date, there is no efficient treatment to combat this pathology, and the only drugs available are usually toxic to the patient. Through the enzyme trans-salidase, the parasite invades, infects, and multiplies intracellularly in the host cell. This protein has been considered an attractive target for developing or searching for compounds with potential trypanocidal activity. In this study, an in silico analysis was performed using a Food and Drug Administration-approved computational drug repositioning approach to identify compounds with anti-Chagas potential against two trans-sialidase proteins. Those compounds with potential inhibition were analyzed and selected through a molecular docking-based virtual screening. Forty-nine compounds were identified, of which forty-five are available on the market, and the rest were evaluated in silico. Our predicted results follow that these compounds are safe for human use and could be potential anti-trans-sialidase agents.

1. Introduction

Chagas disease, also known as American trypanosomiasis, is a neglected public health disease caused by the parasitic protozoan Trypanosoma cruzi (T. cruzi). It is transmitted to humans through triatomine insects that function as vectors. It is estimated that between 6 and 7 million people are infected with the T. cruzi parasite worldwide, with the highest prevalence in Latin America [1,2]. However, cases have been detected in Canada, the United States of America, and many European countries [1] (World Health Organization, 2023). Approximately 75 million people could become infected with T. cruzi if not diagnosed promptly. The symptoms of this pathology can vary (for example, fever, headache, enlarged lymph nodes, paleness, muscle pain, difficulty breathing, swelling, and abdominal or chest pain), although in most cases, there are no symptoms [1]. Chagas disease is the third most common cause of death from parasitic infections in Latin America [3].

To date, no effective vaccines or treatments have been developed to combat Chagas disease, and the only antiparasitics available are nifurtimox and benznidazole [1,4]. However, these drugs are highly toxic to people and only act in the acute phase of the disease [5,6,7]; thus, there is an urgent need to look for therapeutic alternatives against the parasite T. cruzi.

T. cruzi requires sialic acid (SAc) molecules to survive in the host cell; however, the parasite is unable to synthesize SAc de novo, so it acquires it from the host cell’s sialyl-glycoconjugates to transfer SAc monosaccharide to mucin-like receptors on the plasma membrane, conferring a strong negative charge on the parasite to evade the host cell’s immune system. This ingenious activity is carried out by an enzyme known as trans-sialidase (TcTS), a version of a modified sialidase from T. cruzi [8,9]. TcTS enzymes contain a domain of tandem repetitive sequences in the C-terminal region termed SAPA (Shed Acute Phase Antigen). The SAPA domain lengthens the half-life of the TcTS enzyme in the bloodstream and modulates the production of antibodies against the catalytic site [10]. The TS proteins of T. cruzi are encoded by a large family of genes and are classified into eight groups; Group I comprises those genes that translate proteins with catalytic activity (TcTS-active) and Group II includes genes that produce enzymatically inactive proteins (TcTS-inactive) [9,11,12]. It has been documented that TS provides T. cruzi the ability to invade, infect, and multiply intracellularly in the host cell [9,12,13,14]. These characteristics make TS an ideal therapeutic target to combat T. cruzi [9,12]. Furthermore, the TcTS protein is ideal as a target because humans have no analogues [15].

Despite significant efforts to design anti-TS inhibitors, no compound has been discovered that effectively inhibits enzymatic activity [13,15]. The compounds that have been reported as T. cruzi TS inhibitors are analogues to the enzyme-substrate or molecules other than the substrate, such as derivatives of flavonoids and anthraquinones; however, the inhibitory effects are very weak for the TcTS enzyme in millimolar concentration ranges [13,16]. Therefore, it is necessary to implement new drug search strategies against TcTS. Computational approaches such as molecular docking are becoming a powerful tool for searching for compounds with potential therapeutic use [17].

In this study, in silico analysis was performed with an Food and Drug Administration (FDA)-approved computational drug repositioning approach to identify compounds with inhibition potential against two T. cruzi TcTS proteins, TcTS-active and TcTS-inactive. We identified forty-nine compounds, of which forty-five are commercially available for use, and the rest were evaluated in silico. Our predicted results indicate that these compounds are safe for use in people and could function as potential anti-trans-sialidase agents against T. cruzi.

2. Materials and Methods

2.1. Molecular Modeling and Building of the 3D Structure of TcTS from T. cruzi

The three-dimensional structure of the TcTS protein with PDB code 1MS4 was retrieved from the Protein Data Bank (RCSB PDB: Homepage accessed on 6 January 2022; https://www.rcsb.org/). The amino acid sequence of another member of the TS family was recovered from the NCBI gene bank (accession: AAD13347.1) to construct the 3D protein by homology; both proteins correspond to the CL Brener strain. The 3D structure was built using SWISS-MODEL version 4.0 software, accessible through the Ex-PASy web server (https://swissmodel.expasy.org//; accessed on 22 January 2022). The best-predicted models were evaluated using global model quality estimation (GMQE) and evaluated after model construction using the global QMEAN score. The Chimera MatchMaker tool was used to compare the theoretically predicted TS structure with the TS 1MS4 protein. UCSF Chimera 1.14 Molecular Graphics Systems was used to model and visualize the 3D structures [18]. Additionally, a search for conserved motifs was performed using the Multiple EM for Motif Elicitation (MEME) online server (https://meme-suite.org/meme/tools/meme, accessed on 17 October 2022).

2.2. Molecular Docking-Based Virtual Screening

The structures of the TcTS protein with PDB code 1MS4 [13] and the theoretically predicted structure (GenBank accession number: AAD13347.1) were used in all virtual screening calculations. Each protein (PDB format) and ligand (mol2 format) was prepared using the tool option located in the toolbar to select the Dock prep option of the UCSF Chimera version 1.14 software. Solvents/ions were removed. Hydrogen atoms, charges, and standard protonation states were added. The hydrogens added to the ligand were adjusted to a pH of 7.4 to simulate physiological conditions [18]. For virtual screening, the 3D structures were submitted separately to the Drug Discovery TACC server (https://drugdiscovery.tacc.utexas.edu/faq; accessed on 3 January 2023) of the Texas Advanced Computing Center, the medical branch of the University of Texas. This server uses AutoDock Vina version 2.0 software for docking. In total, 555,273 compounds were filtered, including some metabolic compounds from vegetable plants against the TS of T. cruzi. The compounds were selected based on a scoring criterion greater than −6.0 Kcal/mol. The 3D structure of the DANA (2-orexi-2,3-dehydro-n-acetyl-neuramine) inhibitor, previously reported in the literature, was used as a reference [13]. A redocking was performed to verify the selected compounds and to evaluate the binding properties of the ligands with the TS of T. cruzi using the DockThor server (https://dockthor.lncc.br/v2/; accessed on 22 February 2023). DockThor’s scoring function is based on the MMFF94S force field. Force field-based functions consist of a sum of energy terms of a classical force field, usually considering the interaction energies of the protein–ligand complex and the energy of the internal ligand. A grid box with blind molecular docking was used to define the docking region. The parameters are referred to as defaults in DockThor, and the structures with positional root mean square deviation (RMSD) or up to 3 Å were clustered together. All results were analyzed using UCSF Chimera 1.14 Molecular Graphics Systems [18].

2.3. Prediction of In Silico Activity and Toxicity of Selected Compounds by Virtual Screening against T. cruzi TcTS

The PASS (Prediction of Activity Spectra for Substances) online server was used to predict the biological activities (http://www.way2drug.com/passonline/; accessed on 11 May 2023) and the ProTox-II version 3.0 software was used to theoretically determine the endpoints of toxicity, hepatotoxicity, carcinogenicity, and immunotoxicity of the compounds selected as potential T. cruzi TcTS inhibitors (https://tox-new.charite.de/protox_II/index.php?site=home/; accessed on 17 May 2023). ProTox II version 3.0 software divided the tested compounds into six categories depending on their lethal dose (LD50 value), class 1 being lethal and class 6 non-toxic [19]. The PASS version 2024 software theoretically predicts biological activity from real and predicted molecular structures of charged compounds using Pa (active) and Pi (inactive) criteria. The maximum activity value is obtained from the ratio Pa/Pi = 1, while a value close to zero (Pa/Pi → 0) is considered inactive. PASS is based on molecular structure profiling analysis and Bayesian probabilistic modeling to predict the biological activity of molecular ligands [20].

3. Results

3.1. Structural Analysis and Identification of Motifs of Two TcTS Proteins from T. cruzi

Before performing the virtual screening experiments, it was necessary to model the 3D structures of the TcTS proteins of T. cruzi. In this study, two TcTS proteins from T. cruzi were analyzed: a TcTS-active with the code 1MS4 previously reported [13], retrieved from the Protein Data Bank, and another TcTS-inactive with the accession number AAD13347 retrieved from the GenBank; For the latter, the building of the 3D structure prediction was carried out (Figure 1). The length of the TcTS-active protein sequence is 648 amino acid residues, and it folds into two domains; the N-terminal region corresponds to the catalytic domain, and the C-terminal region is the LamG or lectin-like domain (Figure 1a). The catalytic site cleft of TcTS-active is formed by a triad of arginine residues (Arg35, Arg245, and Arg314), in addition to an Asp59 residue and a Tyr342 residue. The predicted model of TcTS-inactive consists of 786 amino acids; topologically, it folds into two domains similar to the previously reported 1MS4 protein [14]. Furthermore, the 1MS4 proteins and the predicted model are structurally conserved (Figure 1c). However, the residues in the N-terminal region of the predicted TcTS-inactive model differ concerning the residues identified in the catalytic site of the TcTS-active structure (Figure 1b). It was previously reported that the difference between TcTS-active and TcTS-inactive proteins is a Tyr342 residue: Tyr342 is present for TcTS-active, while His342 is present for TcTS-inactive, and the Tyr342 → His substitution abolished the enzymatic activity of TS in T. cruzi [21]. Furthermore, ten conserved motifs were identified in both sequences of the T. cruzi TcTS proteins. The sialidase motif is located in the N-terminal region, and two highly conserved motifs (VTV-NV-LYNR-LN and FTLVASVTI) are in the C-terminal region (Figure 1d–f). These motifs have been previously described as FLY and TS9 peptides for inactive TcTS, and they possibly play a crucial role in infection through host–parasite interaction [14,22]. Molecular modeling results indicate that although both T. cruzi TcTS proteins differ in enzymatic activity and residue sequence length, structurally, they are highly conserved, including the sialidase, FLY, and TS9 motifs identified in this work.

3.2. In Silico Identification of Potential Inhibitors against TcTS of T. cruzi

A universe of commercially available and FDA-approved compounds is used for conditions other than those initially intended [23]. In this sense, drug repositioning through molecular docking is of great importance, which consists of making computational predictions of compounds that bind to a protein at specific sites [24]. Molecular docking allows us to make this type of prediction to identify new biological activities of drugs for clinical use [25]. In this work, we execute the molecular docking-based virtual screening method (Figure 2) directed at the TcTS-active and TcTS-inactive proteins of T. cruzi to identify potential anti-Chagas compounds. We use the Drug Discovery TACC server of the Texas University Advanced Computing Center to do this. This server extracts thousands of FDA-approved drugs from the ZINC database, including some natural compounds (https://drugdiscovery.tacc.utexas.edu/; accessed on 3 January 2023).

The molecular docking of the inhibitor DANA, which was used as a reference compound, was carried out [13]. The affinity energies of DANA against TcTS-active and TcTS-inactive proteins were −6.1 (Kcal/mol) and −6.5 (Kcal/mol), respectively. The docking of 546,000 FDA-approved compounds was performed against TcTS-active and TcTS-inactive proteins of T. cruzi. The molecular docking-based virtual screening predicted 1010 compounds with an affinity energy score of −9.5 (Kcal/mol) and −5.0 (Kcal/mol). From this number of predicted compounds, redocking was performed again, and finally, only those compounds with a score equal to or greater than the DANA inhibitor were selected. In total, 49 compounds were selected for each protein, TcTS-active, and TcTS-inactive, including a plant metabolite, sulforaphane; affinity energies ranged from −9.5 (Kcal/mol) to 6.1 (Kcal/mol) (

Table 1. Compounds identified by molecular docking-based virtual screening against the TcTS-active and TcTS-active proteins of T. cruzi.

TcTS-Active		TcTS-Inactive
Compounds	Affinity (Kcal/mol)	Compounds	Affinity (Kcal/mol)
N-[(4-methyl-6-oxo-1h-pyrimidin-2-yl)methyl]benzamide	−9.5	N-[(4-methyl-6-oxo-1h-pyrimidin-2-yl)methyl]benzamide	−9.3
Elbasvir	−9.5	6-hydroxybenzanthrone	−9.2
6-hydroxybenzanthrone	−9.4	N-[(4-fluorophenyl)methyl]tetrazolo[1,5-b]pyridazin-6-amine	−9.2
N-[(4-fluorophenyl)methyl]tetrazolo[1,5-b]pyridazin-6-amine	−9.3	Cobicistat	−9.1
Ivermectin	−9.3	Lopinavir	−9.0
Cobicistat	−9.2	3-(benzylcarbamoylamino)-3-oxopropanoic acid	−9.0
Ledipasvir	−9.2	Ledipasvir	−9.0
3-(benzylcarbamoylamino)-3-oxopropanoic acid	−9.1	Daclatasvir	−8.9
Daclatasvir	−8.8	Elbasvir	−8.8
Lopinavir	−8.7	Piperaquine	−8.8
Digitoxin	−8.6	Ciclesonide	−8.4
Ciclesonide	−8.5	Thiethylperazine	−8.3
Trametinib	−8.4	Remdesivir	−8.2
Remdesivir	−8.2	Imatinib	−8.2
Ivacaftor	−8.2	Ritonavir	−8.1
Imatinib	−8.2	Trametinib	−8.0
Loperamide	−8.1	Abemaciclib	−8.0
Piperaquine	−8.1	Darunavir	−8.0
Fluphenazine	−8.0	Sildenafil	−8.0
Abemaciclib	−8.0	Sofosbuvir	−7.9
Dasatinib	−7.9	Gilteritinib	−7.9
Promethazine	−7.9	Salinomycin	−7.8
Sildenafi	−7.8	Hydroxyprogesterone	−7.8
Diosmin	−7.8	Umifenovir	−7.8
Triflupromazine	−7.7	Hexachlorophene	−7.8
Niclosamide	−7.5	Digitoxin	−7.8
Umifenovir	−7.4	Ruxolitinib	−7.7
Bazedoxifene	−7.3	Tranilast	−7.7
Proscillaridin	−7.3	Ivacaftor	−7.7
Thiethylperazine	−7.2	Tilorone	−7.6
Ibuprofen	−7.1	Niclosamide	−7.5
Fingolimod	−7.0	Baloxavir	−7.5
Ruxolitinib	−6.9	Azithromycin	−7.5
Hydroxychloroquine	−6.9	Chloroquine	−7.5
Leflunomide	−6.8	Methylprednisolone	−7.5
Nitazoxanide	−6.7	Proscillaridin	−7.4
Chloroquine	−6.7	Dexamethasone	−7.4
Gemcitabine	−6.6	Ivermectin	−7.4
Dexamethasone	−6.6	Loperamide	−7.3
Thalidomide	−6.6	Bazedoxifene	−7.2
Oseltamivir	−6.5	Nitazoxanide	−7.1
Penciclovir	−6.5	Ibuprofen	−7.0
Methylprednisolone	−6.5	Dasatinib	−6.9
Nafamostat	−6.3	Oseltamivir	−6.9
Triazavirin	−6.2	Penciclovir	−6.7
Acetylcysteine	−6.1	Thalidomide	−6.6
Ascorbic acid	−6.1	Sulforaphane	−6.5
Sulforaphane	−6.1	Favipiravir	−6.1
Ribavirin	−6.1	Ascorbic acid	−6.1
DANA	−6.1	DANA	−6.5

Note. The TcTS-active protein corresponds to the structure with PDB code 1MS4, and the TcTS-inactive protein is the theoretically predicted structure. The 2-deoxy-2,3-dehydro-N-acetyl-neuraminic acid (DANA) acid compound was used a reference. The drug names in red indicate those unique compounds that were docked for each T. cruzi TcTS protein.

).

In this study, the 49 compounds had a heterogeneous pattern in their chemical structure, intended for different conditions by the FDA. For example, these compounds included antiparasitics, anti-inflammatories, antineoplastics, antivirals, steroids, antipsychotics, antihistamines, antibiotics, and a few compounds without available or sufficient information on their commercial or biological use. Of the 49 compounds docked in both TcTS proteins of T. cruzi, only 12 and 11 different or unique drugs were identified for TcTS-active and TcTS-inactive, respectively; while the rest of the identified compounds shared binding for both proteins, even the affinity energies were similar, as is the case of the drug cobicistat with a score of −9.2 (Kcal/mol) for the TcTS-active protein and −9.1 (Kcal/mol) for the TcTS-inactive protein (Table 1). Most compounds identified against the two TcTS proteins of T. cruzi are known drugs on the market. However, we identified five compounds with a score greater than or equal to the DANA inhibitor that are not completely characterized or about which very little is known. These compounds are N-[(4-methyl-6-oxo-1H-pyrimidin-2-yl)methyl]benzamide, 6-Hydroxybenzanthrone, N-[(4-fluorophenyl)methyl]tetrazolo[1,5-b] pyridazin-6-amine, 3-(benzylcarbamoylamino)-3-oxopropanoic acid, and the plant metabolite sulforaphane. Regarding their physicochemical characteristics, the four compounds are small molecules with a molecular mass of less than 500 kDa, number of H-bond donors (H-BD) < 5, and number of H-bond acceptors (H-BA) < 10 (

Table 2. Physical and chemical properties of the compounds N-[(4-methyl-6-oxo-1H-pyrimidin-2-yl)methyl]benzamide, 6-Hydroxybenzanthrone, N-[(4-fluorophenyl)methyl]tetrazolo[1,5-b]pyridazin-6-amine, 3-(benzylcarbamoylamino)-3-oxopropanoic acid, sulforaphane, including the DANA inhibitor.

ZINC ID	Compounds	Molecular Formula	Molecular Mass (g/Mole)	H-Bond Donors	H-Bond Acceptors	2D Structure
ZINC20031597	N-[(4-methyl-6-oxo-1H-pyrimidin-2-yl)methyl]benzamide	C13H13N3O2	243.26	2	3
ZINC000022636338	6-Hydroxybenzanthrone	C17H10O2	246.26	1	2
ZINC000008724771	N-[(4-fluorophenyl)methyl]tetrazolo[1,5-b]pyridazin-6-amine	C11H9FN6	244.23	1	6
ZINC000004375877	3-(benzylcarbamoylamino)-3-oxopropanoic acid	C11H12N2O4	236.22	2	4
ZINC3875035	Sulforaphane	C6H11NOS2	177.3	0	4
ZINC4096466	DANA	C11H17NO8	291.2	5	8

Note. The physical and chemical data and 2D structure were obtained from ZINC (https://zinc.docking.org/; accessed on 5 January 2023) and PubChem (https://pubchem.ncbi.nlm.nih.gov/; accessed on 5 January 2023) databases. DANA inhibitor as reference.

).

3.3. In Silico Prediction of Biological Activity and Toxicity

As mentioned above, most of the compounds (44 of the 49) identified by virtual screening are FDA-approved; that is, they can be administered to humans, while a minority (only 5 compounds) identified in this study have not been fully characterized. Therefore, we performed the biological activity and toxicity prediction analyses for the compounds N-[(4-methyl-6-oxo-1H-pyrimidin-2-yl)methyl]benzamide, 6-Hydroxybenzanthrone, N-[(4-fluorophenyl)methyl]tetrazolo[1,5-b]pyridazin-6-amine, 3-(benzylcarbamoylamino)-3-oxopropanoic acid, and sulforaphane identified by virtual screening in this study. The PASS server (see Materials and Methods) used for this analysis predicted that the compound N-[(4-methyl-6-oxo-1H-pyrimidin-2-yl)methyl]benzamide may have leukopoiesis stimulant activity. In contrast, the compound 6-Hydroxybenzanthrone can function as a substrate of CYP2C12, and sulforaphane as an inhibitor of CYP2E1 and a chemoprotectant. The compound 3-(benzylcarbamoylamino)-3-oxopropanoic acid has different predicted activities, such as mucomembranous protective (

Table 3. Theoretical prediction of biological activity of five compounds identified by virtual screening.

Compounds	Predicted Activity	Score	2D Structure
N-[(4-methyl-6-oxo-1H-pyrimidin-2-yl)methyl]benzamide	Leukopoiesis stimulant	0.67
	Gastrin inhibitor	0.64
	Kidney function stimulant	0.63
6-Hydroxybenzanthrone	CYP2C12 substrate	0.95
	Histidine kinase inhibitor	0.89
	NAD(p)+arginine ADP-ribosyltransferase inhibitor	0.89
	Alkane 1-monooxygenase inhibitor	0.87
	Cytochrome P450 stimulant	0.72
N-[(4-fluorophenyl)methyl]tetrazolo[1,5-b]pyridazin-6-amine	NAI	NAI
3-(benzylcarbamoylamino)-3-oxopropanoic acid	Mucomembranous protector	0.89
	Polyporopepsin inhibitor	0.87
	Phobic disorders treatment	0.84
	Peptidyl-dipeptidase Dcp inhibitor	0.82
	Chymosin inhibitor	0.82
	Protein-glutamate methylesterase inhibitor	0.73
Sulforaphane	Glutathione S-transferase substrate	0.93
	Chemoprotective	0.87
	Apoptosis agonist	0.87
	CYP2E1 inhibitor	0.70

NAI = Not available information.

). For the compound N-[(4-fluorophenyl)methyl]tetrazolo[1,5-b]pyridazin-6-amine, the PASS server did not return information.

One of the crucial requirements of those compounds discovered for human use is that they must be non-toxic. Therefore, the prediction of the toxicity of the compounds N-[(4-methyl-6-oxo-1H-pyrimidin-2-yl)methyl]benzamide, 6-Hydroxybenzanthrone, N-[(4-fluorophenyl)methyl]tetrazolo[1,5-b]pyridazin-6-amine, 3-(benzylcarbamoylamino)-3-oxopropanoic acid, and sulforaphane was evaluated in silico. The five compounds analyzed in this study were not toxic (

Table 4. Theoretical prediction of toxicity of five compounds identified by virtual screening.

Compounds	Hepatotoxicity	Carcinogenicity	Immunotoxicity	Acute Oral Toxicity (LD50: mg/kg−1)	Toxicity Class
N-[(4-methyl-6-oxo-1H-pyrimidin-2-yl)methyl]benzamide	Inactive	Inactive	Inactive	200	3
6-Hydroxybenzanthrone	Inactive	Inactive	Active	3570	5
N-[(4-fluorophenyl)methyl]tetrazolo[1,5-b]pyridazin-6-amine	Inactive	Inactive	Inactive	500	4
3-(benzylcarbamoylamino)-3-oxopropanoic acid	Inactive	Inactive	Inactive	800	4
Sulforaphane	Inactive	Inactive	Inactive	1000	4

Note: Toxicity is classified from 1 to 6. Class 1 is the most toxic and lethal and class 6 is non-toxic.

). The theoretically predicted data of those compounds did not show hepatotoxicity, carcinogenicity, or immunotoxicity effects. The ProTox-II software used in this study to evaluate the toxicity of the compounds classified the substances into six toxicity classes, with class 1 being the most lethal and toxic (LD50 ≤ 5) and class 6 designating the non-toxicity of the compound (LD50 > 5000). The compound N-[(4-methyl-6-oxo-1H-pyrimidin-2-yl)methyl]benzamide falls into class 3 with an LD50 value of 200 mg/kg; the compound 6-Hydroxybenzanthrone falls into class 5 with an LD50 value of 3.570 mg/kg; the N-[(4-fluorophenyl)methyl]tetrazolo[1,5-b]pyridazin-6-amine compound falls into class 4 with an LD50 value of 500 mg/kg; while 3-(benzylcarbamoylamino)- 3-oxopropanoic acid and sulforaphane fall into class 4 with an LD50 value of 800 and 1000, respectively (Table 4). Theoretical data of biological activity and toxicity indicate that these four compounds may be safe for humans because they are not toxic; even the compound 3-(benzylcarbamoylamino)-3-oxopropanoic acid may have beneficial effects as a mucomembranous protector.

By separately analyzing the five compounds in both TcTS proteins, surprisingly, we found that the compounds N-[(4-methyl-6-oxo-1H-pyrimidin-2-yl)methyl]benzamide (ZINC20031597) and sulforaphane (ZINC3875035) bind in specific regions of both TcTS proteins of T. cruzi. Docking predicted that the compound N-[(4-methyl-6-oxo-1H-pyrimidin-2-yl)methyl]benzamide can bind in both the pocket and the C-terminal region of the TcTS-active protein (Figure 3a,b). In the pocket, this compound interacts through four H-bonds with residues Arg35, Arg314, and Asp59, while in the C-terminal region, it possibly interacts with three H-bonds with residues Pro450, Asn603, and Asn604 of TS9 and FLY peptides (Figure 3a,b).

The sulforaphane compound also showed two poses in the pocket and the C-terminal region of the TcTS-active protein (Figure 3c,d). The residues that interact in the pocket with the ligand (sulforaphane) through five H-bonds are Asp51, Tyr113, Arg245, Arg134, Tyr 342, and Tyr 364, while in the C-terminal region, it forms two H-bonds with Asn451 and Gly628 (Figure 3c,d). Similarly, the results obtained after docking prediction indicate that the plant metabolite sulforaphane also has a binding preference against the TcTS-inactive protein of T. cruzi in the N-terminal region and the C-terminal region, except the compound N-[(4-methyl-6-oxo-1H-pyrimidin-2-yl)methyl]benzamide, which only showed binding in the C-terminal region. Sulforaphane interacts with His167, Glu383, and Thr415 through three H-bonds in the N-terminus and with Ser469, Ser544, Thr667, Ile684, and Arg687 in the C-terminal region through four H-bonds in the TcTS-inactive. By comparison, the compound N-[(4-methyl-6-oxo-1H-pyrimidin-2-yl)methyl]benzamide only forms an H-bond with the Lys688 residue in the C-terminal region in the TcTS- inactive of T. cruzi (Figure 4c). The molecular docking-based virtual screening results indicate that 49 potential compounds bind against two TcTS proteins of T. cruzi. The data shown in this work suggest that these compounds could function as potential anti-TcTS compounds of T. cruzi. We encourage the scientific community to implement in vitro assays to validate the drugs identified in this study.

4. Discussion

The search for or design of bioactive compounds is increasing in the pharmaceutical industry. FDA-approved computational drug repositioning approaches have innovated the discovery of bioactive molecules by evaluating large libraries of compounds in silico, thus leading to the reduction in costs, infrastructure, space, and time involved in the de novo design process [23,26,27] In this study, we implemented a molecular docking-based virtual screening strategy using a database (see Materials and Methods) of FDA-approved compounds to identify molecules with inhibition potential against two proteins, TcTS-active and TcTS-inactive, in T. cruzi.

Docking is a computational prediction technique that analyzes how ligands and receptors fit together and how proteins interact with the ligand [28,29]. A study carried out by molecular docking of derivatives of 3-amino-3-aryl propionic acid against TcTS enzymes showed that this compound had a correlation between the docking method and enzymatic inhibition. Furthermore, it had trypanocidal activity for the NINOA and INC-5 strains [30]. Compounds derived from benzamides as inhibitors of the glycine transporter type 1 (GlyT1) were recently identified as candidate drugs to treat schizophrenia using computational tools such as molecular modeling and molecular docking [31]. Through this approach, plant metabolites (flavonoids) have also been identified as potential inhibitors of the enzyme indoleamine 2,3-dioxygenase-1 (IDO1). Plant flavonoids are promising compounds against IDO1 as an interesting target for anticancer action [32]. In the present work, our in silico predicted results showed 49 compounds, including a plant metabolite, as potential anti-TS candidates in two proteins (TcTS-active and TcTS-inactive) in T. cruzi (Table 1). These compounds were selected based on the affinity values of the reference inhibitor DANA (−6.1 kcal/mol and −6.5 kcal/mol). The affinity energies of the selected compounds against the two TcTS proteins varied between −9.5 kcal/mol and −6.1 kcal/mol (Table 1). The vast majority of the compounds identified in this study are drugs known to treat different medical indications, such as antiparasitics, anti-inflammatories, antineoplastics, antivirals, steroids, antipsychotics, antihistamines, and antibiotics. However, there is insufficient information about the biological activity and toxicity of five compounds that were significantly docked in both TcTS proteins of T. cruzi. Those compounds are N-[(4-methyl-6-oxo-1H-pyrimidin-2-yl)methyl]benzamide, 6-Hydroxybenzanthrone, N-[(4-fluorophenyl)methyl]tetrazolo[1,5-b]pyridazin-6-amine, 3-(benzylcarbamoylamino)-3-oxopropanoic acid, and sulforaphane. The predicted results suggest that these compounds are not toxic, may be safe for human use, and could even have beneficial properties such as a leukopoiesis stimulant, mucomembranous protector, and chemoprotectant (Table 3).

When analyzing these molecules separately, interestingly, we found that N-[(4-methyl-6-oxo-1H-pyrimidin-2-yl)methyl]benzamide and sulforaphane bind in the pocket and the C-terminal region of both TcTS proteins of T. cruzi. Experimental data have documented that a Tyr342 residue in the pocket in TcTS-active proteins is necessary for enzyme functionality [21]. Crystallographic studies of the active enzyme TcTS indicate that the triad of residues Arg35, Arg245, and Arg314 is required for binding to the carboxylate group of sialic acid, and two conserved residues of Asp59 and Tyr342 are essential for catalysis, while the H bonding interaction between Tyr342 and Glu230 is crucial for stabilizing the transition state [13]. In this same study, it was reported that the substrate (sialic acid) induces conformational changes in the TcTS enzyme. The most notable change was observed in the side chain of residue Tyr119, in which the aromatic side chain moves toward the hydrophobic pocket and makes H-bonding interactions with the glycerol moiety of the ligand. Furthermore, the movement of Tyr119 generates a direct link between sialic acid binding and increased affinity for the acceptor substrate, according to crystallographic studies [13]. The compound N-[(4-methyl-6-oxo-1H-pyrimidin-2-yl)methyl]benzamide was docked into the pocket (N-terminal) of the TcTS-active protein through four H-bonds (with residues Asp59, Arg35, and Arg314) very close to the Tyr119 and Try342 residues (see Figure 3a), in such a way that it possibly prevents the binding of the substrate, inhibiting the enzymatic activity of the enzyme. On the other hand, the sulforaphane metabolite docked with greater stability in the pocket of the TcTS-active protein, interacting with six residues, including Tyr342, through six H-bonds (see Figure 3c). Sulforaphane (isothiocyanate) is a plant metabolite found in broccoli, a plant in the Brassicaceae family [33,34]. Previous studies have shown that sulforaphane can protect against several types of cancer and cardiovascular disease risk, making this compound a potential anti-cancer agent [35,36]. Additionally, a recent study showed that sulforaphane has antioxidant and protective effects on endothelial cells [37]. Although sulforaphane was tested in phase I and II clinical trials, with excellent anticancer and antibacterial results, it was never approved as a drug [38]. Only two reports of sulforaphane showed inhibition of the intracellular growth of the T. cruzi parasite in Hela cells and macrophages [39,40]. There is no information available in the literature on the mode of action of this metabolite. In this study, we predicted in silico that sulforaphane was docked in the catalytic site and the N-terminus region of two TcTS proteins from T. cruzi, a previously crystallized structure (TcTS-active) and another protein predicted by homology (TcTS-inactive) (Figure 3 and Figure 4).

On the other hand, these two molecules (N-[(4-methyl-6-oxo-1H-pyrimidin-2-yl)methyl]benzamide and sulforaphane) bind in the C-terminal region of both TcTS proteins, as suggested by molecular docking data. As shown in Figure 3d, the sulforaphane compound formed the complex (protein–ligand) with the TcTS-active protein and interacted with Asn451 and Gly628 through two H-bonds, and possibly also with Asn603, Asn604, and Ser487; these residues are located in the sequence of the FLY and TS9 peptide motifs, respectively. In the TcTS-inactive protein, sulforaphane forms five H-bonds, and it interacts directly with Ser544 and Thr667; these amino acid residues are located in the beta-sheet sequence of the FLY and TS9 peptide motifs, respectively, whereas the compound N-[(4-methyl-6-oxo-1H-pyrimidin-2-yl)methyl]benzamide was similarly docked to sulforaphane, forming only one H-bond with the oxygen of Lys688 (Figure 4b,c).

It has been documented that the FLY and TS9 peptides located in the LamG domain (C-terminal region) of the TcTS enzyme in T. cruzi present binding properties to the cytokeratin of the host cell. FLY and TS9 have been proposed to be essential in initiating cell adhesion and invasion [14]. Both peptides were found in group II protein sequences that are conserved in all members of the eight TS superfamily groups in T. cruzi [14]. To date, no inhibitor has been reported that efficiently reduces the enzymatic activity of the TS of T. cruzi, and no inhibitor has been directed at the C-terminal region. The first reports of monoclonal antibodies (H1A10 and 6A2) against a surface glycoprotein of T. cruzi were published decades ago. However, only partial inhibition of invasion of LLC-MK2 cells by T. cruzi was observed [41]. Since then, significant efforts have been made in drug design strategies, and currently, in combination with a computational approach, some promising compounds have been identified. Unfortunately, they have only shown a 50% inhibitory effect at millimolar concentrations [42]. Compounds that have been reported as T. cruzi TcTS inhibitors are analogous to the enzyme substrate or molecules other than the substrate. The DANA inhibitor binds to the catalytic site of the recombinant sialidase enzymes of T. rangeli and the TcTS of T. cruzi [13,15]. However, DANA showed a low affinity for the TcTS enzyme of T. cruzi and, consequently, a weak effect as a TcTS enzymatic inhibitor [13,15]. Other analogues of the TcTS substrate are 2,3-difluorosialic acid and 2-[2-(difluoromethyl)-4-nitrophenyl]-3,5-dideoxy-d-glycero-α-D-galacto-2-nonulopyranoside acid (Neu5AcFNP); these compounds had an inhibitory effect of 20 and 10 mM, respectively. Surprisingly, 2,3-difluorosialic acid covalently interacts with the hydroxyl group of residue Tyr342. However, the TcTS enzymatic inhibition of T. cruzi is temporary, while Neu5AcFNP formed covalent bonds with Arg245 and Asp247 and inhibited trypomastigote infection in cells of mammals [43,44]. Neu5AcFNP derivatives (dansyl-Neu5AcFP and Neu5Aca2–3-Galβ-O-octyl) also had a poor effect as T. cruzi TS inhibitors [45].

Analysis performed by molecular docking and virtual screening identified benzoic acid derivatives as potential inhibitors of TcTS in T. cruzi. The 4-acetylamino-3-hydroxymethylbenzoic acid derivative had TcTS inhibition with a Ki value of 300 μM and an IC50 of 0.54 mM [46]. Akioca and colleagues analyzed a library of natural compounds and found derivatives of flavonoids and anthraquinones as promising inhibitors of T. cruzi TcTS and mentioned that these compounds might serve as chemotherapeutic agents against Chagas disease [16]. A computational study that analyzed 3180 FDA-approved drugs against the TcTS of T. cruzi identified the anti-inflammatory sulfasalazine as a potential inhibitor of the TcTS enzyme. In vitro experiments demonstrated that this compound had a moderate inhibition of 37% on the enzymatic activity of TcTS [47]. On the other hand, a molecular docking study identified that the alkaloid montanine from Amaryllidaceae plants binds to the active site of TcTS T. cruzi. The authors note that montanine was the only alkaloid stable to TcTS; however, kinetic studies and in vivo tests are required [48]. The multiple roles of TcTS proteins are still not fully understood. Recently, in a study using CRISPR-Cas9 gene editing technology, the TcTS-active genes of the CL Brener strain were disrupted [49]. It was determined that the interrupted TS-active genes did not affect the parasite’s ability to invade cells or escape from the parasitophorous vacuole, but it did affect the conversion of amastigotes to trypomastigotes and also the exit of T. cruzi from the cell. In mice inoculated with T. cruzi cells mutated in TS-active, the parasite did not have the capacity to infect, even in the highly susceptible interferon gamma (IFN-γ) knockout mice. The authors of this study concluded that TS-active of T. cruzi plays an important role during the late stages of intracellular development and parasite egress [49].

5. Conclusions

In this work, a molecular docking-based virtual screening method was used to identify potential T. cruzi anti-TcTS compounds using a database of FDA-approved drugs. Based on the reference inhibitor DANA, those compounds with a docking score ≥ 6.0 (affinity in kcal/mol) were selected. In total, 49 compounds were identified, of which 44 are known drugs on the market and 5 have little information on their biological activity. These compounds are N-[(4-methyl-6-oxo-1H-pyrimidin-2-yl)methyl]benzamide, 6-hydroxybenzoantrone, N-[(4-fluorophenyl)methyl]tetrazolo[1,5-b]pyridazine -6-amine, 3-(benzylcarbamoylamino)-3-oxopropanoic acid, and sulforaphane. Of the 44 compounds identified, only ascorbic acid in combination with benznidazole has been reported to have trypanocidal effects; the rest of the compounds remain to be discovered on T. cruzi. The results of the docking data showed that the compounds N-[(4-methyl-6-oxo-1H-pyrimidin-2-yl)methyl]benzamide and sulforaphane were docked in the N-terminal and C-terminal regions of the two TcTS proteins from T. cruzi (TcTS-active and TcTS-inactive). Currently, several protein–ligand docking programs have been developed. In our study, we implemented a method to identify potential anti-TcTS compounds from T. cruzi. The docking calculations were run on supercomputers designed and maintained by the Texas Advanced Computing Center (TACC) at the University of Texas. The workflow presented in this methodology was based on the use of the AutoDock Vina program for docking calculations. The advantage of using this method is that AutoDock Vina yields reliable results because it combines knowledge-based empirical scoring functions with empirical parameters derived from the experimental binding affinities of proteins and ligands. Consequently, the methodology shown in this study is easy to execute, does not require too much computational capacity, and does not require much experience; nor is it required to be an expert in the area. Another advantage is that the server (TACC) is completely free and uses virtual libraries with thousands of ligands, with excellent results in a short time, to identify potential inhibitors.

In conclusion, our in silico study predicts that a single compound, or a combination, can be used to affect the enzymatic activity of T. cruzi TcTS. Therefore, we hope that the compounds identified in this work will be able to inhibit/block both TcTS proteins, to thus provide an alternative to combat Chagas disease. Although the results presented in this study are promising, it is necessary to perform laboratory experiments to validate the inhibitory effect of these compounds. However, this was not the objective of this study, so we encourage the scientific community to test this hypothesis.