In Vitro and In Silico Analysis of New n-Butyl and Isobutyl Quinoxaline-7-carboxylate 1,4-di-N-oxide Derivatives against Trypanosoma cruzi as Trypanothione Reductase Inhibitors

American trypanosomiasis is a worldwide health problem that requires attention due to ineffective treatment options. We evaluated n-butyl and isobutyl quinoxaline-7-carboxylate 1,4-di-N-oxide derivatives against trypomastigotes of the Trypanosoma cruzi strains NINOA and INC-5. An in silico analysis of the interactions of 1,4-di-N-oxide on the active site of trypanothione reductase (TR) and an enzyme inhibition study was carried out. The n-butyl series compound identified as T-150 had the best trypanocidal activity against T. cruzi trypomastigotes, with a 13% TR inhibition at 44 μM. The derivative T-147 behaved as a mixed inhibitor with Ki and Ki’ inhibition constants of 11.4 and 60.8 µM, respectively. This finding is comparable to the TR inhibitor mepacrine (Ki = 19 µM).


Introduction
No efficient drug treatment has been developed in the more than 110 years since the discovery of the causal agent of American trypanosomiasis or Chagas disease. For this reason, it continues to be one of the most prevalent parasitic diseases, with 6 to 7 million people infected worldwide [1]. The increase in human migration from endemic countries and vector migration resulting from climate change have increased the incidence of this disease in non-endemic areas, making it a worldwide problem [2].
The variable efficacy of benznidazole (Bnz) and nifurtimox (Nfx) in the acute and chronic phases of Trypanosoma cruzi (T. cruzi) infection and their high human toxicity have resulted in treatment abandonment [3][4][5]. Therefore, more effective, and less toxic new treatments are necessary [6,7].

Biological Evaluation
Trypanocidal Activity against Trypomastigotes Quinoxaline derivatives were initially evaluated against trypomastigote at a single fixed concentration of 50 µg/mL (Table 1). NINOA strain mortality ranged from 8.7 to 78.3%. For the reference drugs, Nfx and Bnz, it was 76.6% and 68.8%, respectively. Four compounds, T-142, T-168, T-169, and T-170, had a similar or higher trypanocidal activity than the reference drugs. Table 1. Trypanocidal activity of n-butyl-and isobutyl quinoxaline-7-carboxylate 1,4-di-N-oxide derivatives against T. cruzi trypomastigotes of the NINOA and INC-5 strains. Figure 1. The synthetic pathway for forming the n-butyl and isobutyl quinoxaline 1,4-di-N-oxide series from the Beirut reaction.
The half-maximal lytic concentration (LC 50 ) analysis showed that multiple compounds derived from these two series had better trypanocidal activity than the reference drugs Nfx and Bnz against the two strains tested. Compounds T-141, T-150, and T-169 showed the best activity of both series with LC 50 values of 38.9 and 118, 64.3 and 81.5, and 56.9, and 62.2 µM for NINOA and INC-5, respectively, compared to 70.4 and 139.4, and 130.7 and 191.3 µM for Nfx and Bnz, respectively. These findings reinforce the importance of the presence of aromatic substituents in quinoxaline derivatives. T-141 featured a benzyl ester group, T-150, a naphthyl ketone, and T-169, a benzamide.

Molecular Docking on TcTR
Molecular docking analysis on the active site of TcTR showed that seven (T-149, T-150,  T-151, T-157, T-162. T-163, and T-167) of the thirty quinoxaline-7-carboxylate 1,4-di-N-oxide derivatives had a similar or lower predicted free energy of binding (FEB) than the natural ligand trypanothione disulfide (Table 2). Based on the structure of the substituents present in these quinoxalines, it may be suggested that the presence of large aromatic groups and protonable amines tends to increase the affinity toward the catalytic site of TcTR. Table S1) and 2-D representation (Supplementary  Table S2) for all evaluated ligands are included in the Supplementary Material. Table 2. Estimated free energy of binding (FEB) and 2D-interaction profile representation for top scored (≤−8.3 kcal/mol) n-butyl and isobutyl quinoxaline-7-carboxylate 1,4-di-N-oxide derivatives binding on the active site of TcTR.

Enzymatic Activity Evaluation
Three (T-147, T-148, and T-150) n-butyl derivatives bearing a trifluoromethyl group were selected ( Figure 3) to test their capacity to inhibit TR. These compounds showed a high mortality percentage and structural similarity to the known quinoxaline-derived TR inhibitor, T-085 [44].  (Table 3).

Enzymatic Activity Evaluation
Three (T-147, T-148, and T-150) n-butyl derivatives bearing a trifluoromethyl group were selected ( Figure 3) to test their capacity to inhibit TR. These compounds showed a high mortality percentage and structural similarity to the known quinoxaline-derived TR inhibitor, T-085 [44].

Enzymatic Activity Evaluation
Three (T-147, T-148, and T-150) n-butyl derivatives bearing a trifluoromethyl group were selected ( Figure 3) to test their capacity to inhibit TR. These compounds showed a high mortality percentage and structural similarity to the known quinoxaline-derived TR inhibitor, T-085 [44].   (Table 3). Compound T-147 ( Figure 4) was studied at two fixed inhibitor concentrations (5 and 20 µM) to evaluate the type of inhibition in more detail, varying the trypanothione disulfide concentrations (22,44,88,110, and 220 µM) and saturating NADPH. A Lineweaver-Burk plot revealed a mixed-type inhibition with respect to trypanothione disulfide. The inhibitor constants, Ki and Ki' of 11.4 and 60.8 µM, respectively, were calculated by non-linear regression using Graph Pad Prism software (version 5). Compound T-147 ( Figure 4) was studied at two fixed inhibitor concentrations (5 and 20 μM) to evaluate the type of inhibition in more detail, varying the trypanothione disulfide concentrations (22,44,88,110, and 220 μM) and saturating NADPH. A Lineweaver-Burk plot revealed a mixed-type inhibition with respect to trypanothione disulfide. The inhibitor constants, Ki and Ki' of 11.4 and 60.8 μM, respectively, were calculated by non-linear regression using Graph Pad Prism software (version 5).

Molecular Dynamics on TcTR
A molecular dynamics analysis was performed for the compound that behaved as the TR inhibitor, T-147, and the known TR inhibitor, mepacrine. The dynamics were

Molecular Dynamics on TcTR
A molecular dynamics analysis was performed for the compound that behaved as the TR inhibitor, T-147, and the known TR inhibitor, mepacrine. The dynamics were analyzed with three measurements: root-mean-square differentiation (RMSD), root-mean-square fluctuation (RMSF), and radius of gyration. Figure 5 shows the RMSD fluctuations for the T-147-TcTR complex (green), mepacrine-TcTR complex (yellow), and free TcTR (red). The mepacrine-TcTR complex shows a maximum fluctuation of 8.1 Å and maintains a fluctuation in the range of 3-6 Å with a mean of 6.00 ± 1.53 Å. While, the T-147-TcTR complex shows a maximum fluctuation of 12.28 Å, which can be observed at the end of the dynamics; still, the dynamics remain stable for the first 88 ns with a fluctuation of around 2 Å. The overall fluctuation had a mean of 2.68 ± 2.01 Å.
2.5.1. RMSD Analysis Figure 5 shows the RMSD fluctuations for the T-147-TcTR complex (green), mepacrine-TcTR complex (yellow), and free TcTR (red). The mepacrine-TcTR complex shows a maximum fluctuation of 8.1 Å and maintains a fluctuation in the range of 3-6 Å with a mean of 6.00 ± 1.53 Å. While, the T-147-TcTR complex shows a maximum fluctuation of 12.28 Å, which can be observed at the end of the dynamics; still, the dynamics remain stable for the first 88 ns with a fluctuation of around 2 Å. The overall fluctuation had a mean of 2.68 ± 2.01 Å.

RMSF Analysis
RMSF fluctuations for the complexes, T-147-TcTR (green), mepacrine-TcTR (red), and free TcTR (yellow), are shown in Figure 6. The fluctuations have only minor differences with respect to the apoprotein, suggesting stable binding without affecting residues outside the active site.  Figure 7 shows the radius of gyration for the mepacrine-TcTR and T-147-TcTR complexes, and apo-TcTR. This graph shows a mean radius of gyration between 31 and 32 Å.

RMSF Analysis
RMSF fluctuations for the complexes, T-147-TcTR (green), mepacrine-TcTR (red), and free TcTR (yellow), are shown in Figure 6. The fluctuations have only minor differences with respect to the apoprotein, suggesting stable binding without affecting residues outside the active site.
2.5.1. RMSD Analysis Figure 5 shows the RMSD fluctuations for the T-147-TcTR complex (green), mepacrine-TcTR complex (yellow), and free TcTR (red). The mepacrine-TcTR complex shows a maximum fluctuation of 8.1 Å and maintains a fluctuation in the range of 3-6 Å with a mean of 6.00 ± 1.53 Å. While, the T-147-TcTR complex shows a maximum fluctuation of 12.28 Å, which can be observed at the end of the dynamics; still, the dynamics remain stable for the first 88 ns with a fluctuation of around 2 Å. The overall fluctuation had a mean of 2.68 ± 2.01 Å.

RMSF Analysis
RMSF fluctuations for the complexes, T-147-TcTR (green), mepacrine-TcTR (red), and free TcTR (yellow), are shown in Figure 6. The fluctuations have only minor differences with respect to the apoprotein, suggesting stable binding without affecting residues outside the active site.

In Vitro Selectivity Assessment of Enzymatic Activity
The effect of T-147 on human glutathione reductase (hGR) was determined to assess the selectivity of compound T-147 towards the parasite TR. The degree of inhibition of human GR was measured at two fixed concentrations (5 μM and 20 μM) of T-147 in the presence of 37 or 92 μM glutathione disulfide (GSSG) saturating NADPH. The results are shown in Table 4.

In Vitro Selectivity Assessment of Enzymatic Activity
The effect of T-147 on human glutathione reductase (hGR) was determined to assess the selectivity of compound T-147 towards the parasite TR. The degree of inhibition of human GR was measured at two fixed concentrations (5 µM and 20 µM) of T-147 in the presence of 37 or 92 µM glutathione disulfide (GSSG) saturating NADPH. The results are shown in Table 4. A second experiment was conducted to determine the type of inhibition and the inhibition constant at two fixed concentrations of the inhibitor (5 and 20 µM), varying the glutathione disulfide concentrations (21.5, 43, 86, 107.5, 372, and 930 µM) and saturating NADPH. The compound was a non-competitive inhibitor of hGR with a Ki value of 25 µM. The kinetic constant was obtained by non-linear regression of the experimental data using Graph Pad Prism (version 5). In addition, a Lineweaver-Burk plot that directly visualized the non-competitive type of inhibition was prepared (Supplementary Figure S1).

Molecular Dynamics on hGR
To further analyze selectivity, a molecular dynamics analysis was performed with hGR for the compound found to behave as a TR inhibitor, T-147, and for the known TR inhibitor, mepacrine. The dynamics were analyzed with three measurements: RMSD, RMSF, and radius of gyration.

RMSF Analysis
RMSF fluctuations for the complexes, T-147-hGR (green) and mepacrine-TcTR (red), and free hGR (yellow) are shown in Figure 9. The fluctuations had minor differences with respect to the apoprotein in the residue with ranges of 118-158, 288-298, and 328-348, suggesting unstable binding since it minorly affects residues outside the intended binding site.

RMSF Analysis
RMSF fluctuations for the complexes, T-147-hGR (green) and mepacrine-TcTR (red), and free hGR (yellow) are shown in Figure 9. The fluctuations had minor differences with respect to the apoprotein in the residue with ranges of 118-158, 288-298, and 328-348, suggesting unstable binding since it minorly affects residues outside the intended binding site.

RMSF Analysis
RMSF fluctuations for the complexes, T-147-hGR (green) and mepacrine-TcTR (red), and free hGR (yellow) are shown in Figure 9. The fluctuations had minor differences with respect to the apoprotein in the residue with ranges of 118-158, 288-298, and 328-348, suggesting unstable binding since it minorly affects residues outside the intended binding site.  Figure 10 shows the radius of gyration for the mepacrine-hGR and T-147-hGR complexes, and apo-hGR. This graph shows a mean radius of gyration between 30.5 and 31.5 Å. Figure 10. The radius of gyration graph for molecular dynamics over time for mepacrine-hGR (yellow) and T-147-hGR complexes (green), and apo-hGR (red).  Figure 10 shows the radius of gyration for the mepacrine-hGR and T-147-hGR complexes, and apo-hGR. This graph shows a mean radius of gyration between 30.5 and 31.5 Å.

Radius of Gyration
RMSF fluctuations for the complexes, T-147-hGR (green) and mepacrine-TcTR (red), and free hGR (yellow) are shown in Figure 9. The fluctuations had minor differences with respect to the apoprotein in the residue with ranges of 118-158, 288-298, and 328-348, suggesting unstable binding since it minorly affects residues outside the intended binding site.  Figure 10 shows the radius of gyration for the mepacrine-hGR and T-147-hGR complexes, and apo-hGR. This graph shows a mean radius of gyration between 30.5 and 31.5 Å. Figure 10. The radius of gyration graph for molecular dynamics over time for mepacrine-hGR (yellow) and T-147-hGR complexes (green), and apo-hGR (red).  Regarding the isobutyl series against the NINOA strain, LC 50 values were 33.3 to 556 µM. Four of the fifteen compounds were under 100 µM, comparable to or lower than the reference drugs. The LC 50 values obtained for the isobutyl series against the INC5 strain were 62.2 to 233 µM. Nine of the fifteen compounds were below 136 µM, lower than the reference drugs. The LC 50 ranges found herein were comparable to or lower than the reference drugs.

Trypanocidal Activity
Reports of the half-maximal cytotoxic concentration values (CC 50 ) for quinoxaline derivatives suggest that multiple derivatives from both series may act as trypanocidal agents at concentrations safe for mammalian cells. Quinoxaline-1,4-di-N-oxide derivatives reported by Chacon-Vargas et al., 2017, and analogous to those reported here, showed mean cytotoxic concentrations (CC 50 ) against macrophages in the range of 6.7-325.61 µM on equivalence with reference drugs Nfx and Bnz with CC 50 201.05 and 352.01 µM, respectively. These derivatives had LC 50 values of 2.42-238.28 µM with a desired higher trypanocidal effect rather than a cytotoxic effect for multiple derivatives [44]. Perez-Silanes et al., 2016 reported quinoxaline derivatives with IC 50 values of 0.6 to 12.1 µM for the highest trypanocidal compounds compared to CC 50 values for VERO cells of 3-454.7 µM, with all these compounds having higher antiparasitic than cytotoxic activity [45]. Quinoxaline derivatives reported by Estevez et al. showed a trypanocidal effect with an IC 50 of 11.5 to >25 µM and cytotoxic activity against VERO cells of 7.6 to >254 µM [36]. Taken together, these studies suggest that quinoxalines can potentially be used safely against parasites as they have lower cytotoxicity than trypanocidal activity.

Structure-Activity Relationship
The n-butyl derivative with the highest mortality percentage against trypomastigote form against both strains of the parasite was T-150, and for isobutyl derivatives, compounds T-167 and 169. On the one hand, T-150 has a naphthyl group at the 2-position and a trifluoromethyl group at the 3-position. In contrast, T-167 and T-169 bear a benzamide substituent at the 2-position, and phenyl and methyl at 3-position respectively. These results support the importance of bulky, aromatic, and halogenated substituents to favor biological activity. In contrast to these results, the analogs of these compounds with methyl and ethyl esters at the 7-position in the study by Villalobos-Rocha et al. [43] and isopropyl in the study by Chacon-Vargas et al. [39] did not show outstanding results suggesting that perhaps a longer chain for the ester at the 7-position allows enhanced activity.
Derivatives with bulky aromatic or trifluoromethyl groups were among the most active compounds against trypomastigotes. A similar comparison revealed that some compounds and T-167 vs. T-155) were favored against NINOA strain trypomastigotes. Compared to the presence of alkyl vs. alkoxide, a slight increase was observed in the activity of ester-containing compounds compared to ketone-containing compounds for the NINOA stain. Aliphatic esters at the 2-position, in a trypomastigote assay of NINOA strain, had similar activity, with tert-butyl substituted T-140 being the lowest compared to aromatic benzyl ester T-141 which has the highest activity among all the esters. This tendency is different in the INC-5 strain, where the tert-butyl-substituted T-140 was the most potent ester, and the aromatic benzyl ester T-141 had slightly less activity. Compounds T-143 and T-144, which have benzamides at the 2-position, had activities higher than 50% for the NINOA strain. These compounds had a notable trypanocidal activity against this parasite form. In contrast, their isobutyl analogs did not show an activity higher than 30% against NINOA, yet T-163 displayed an activity higher than 50% against the INC-5 strain.

Molecular Docking on Trypansoma cruzi Trypanothione Reductase (TcTR)
The interaction predicted by molecular docking with the PLIP software showed that the n-butyl and isobutyl series interact with previously reported TcTR residues important for binding the natural ligand trypanothione disulfide (see PLIP interaction profile in Table 2, Supplementary Tables S1 and S2).
The most recurring predicted interactions for n-butyl-quinoxaline-7-carboxylate 1,4di-N-oxide derivatives were with active site residues, Val-59, Ile107, Ile339, Phe-396, Leu-399, His-461, and Glu-466. Similarly, isobutyl quinoxaline-7-carboxylate 1,4-di-N-oxide derivatives had recurring predicted interactions with Ile-339, Asn-340, Phe-369, His-461, and Glu-466. The residues Leu-399 and Phe-396 are part of the denominated Z-site (subsite within the interphase catalytic site) reported to participate in the binding of the hydrophobic part of chlorpromazine, a known TR inhibitor; likewise, the interactions with Glu-466 and His-461 are considered important to the binding of the natural ligand [46,47]. According to the interaction profile of trypanothione disulfide, it can be predicted that both n-butyl and isobutyl quinoxaline derivatives may behave as TR inhibitors. Figure 2 shows the interaction profile predicted for the new quinoxaline-1,4-di-N-oxide derivative that behaves as a TR inhibitor. No predicted interactions are shared with the natural ligand, which is congruent with the mixed-inhibitory activity found.

Trypanothione Reductase Inhibition
In accordance with previous reports [46,47], we suggest that n-butyl-quinoxaline-7carboxylate-1,4-di-N-oxide derivatives may inhibit TR and, through this activity, contribute to the molecular action mechanism of trypanocidal activity observed against T. cruzi. As such, three (T-148, T-147, and T-150) derivatives bearing a trifluoromethyl group ( Figure 5) were selected to test their capacity to inhibit TR. These compounds showed a high mortality percentage and structural similarity to the known quinoxaline-derived TR inhibitor, T-085 [44].
Firstly, the compounds were analyzed at a single sub-saturating substrate concentration of 44 µM TS 2 (Km TS 2 = 23 µM) [48]. At 20 µM (the highest non-precipitating concentration), T-148 showed an inhibition of 36%, followed by T-147 with a percentage inhibition of 35%; T-150 was insoluble (Table 3). Each compound was then evaluated at a concentration of 5 µM. Under these conditions, T-147 and T-148 yielded 18% inhibition, whereas T-150 lowered the activity by 13% (Table 3). In the presence of 100 µM TS 2 , T-148 and T-147 displayed a degree of inhibition similar to that seen at 44 µM TS 2 , suggesting that the inhibition may not be purely competitive. Though all compounds chosen for the enzymatic inhibition assay bear hydrophobic substituents, it is noteworthy that the structures of the two compounds with the most inhibition, T-147, and T-148, had the least bulky substituents, suggesting that they fit better at the inhibition site than the bulkier T-150. At the same time, T-150 showed a slight inhibition at a 44 µM concentration of TS 2 but no inhibition at 100 µM TS 2, suggesting that if T-150 were to behave as an inhibitor, it would have a more competitive nature than T-147 and T-148. Even though T-148 and T-150 share an aromatic substituent at the 2-position, their behaviors are considerably different, which is probably related to the bulkiness of the latter.

In Silico Analysis of Inhibitor Binding TcTR
RMSD results ( Figure 5) for mepacrine are consistent with the information that this compound is a confirmed trypanothione reductase inhibitor with a stable trajectory at about 6 Å. The behavior observed for T-147 is consistent with its inhibitory activity, maintaining an RMSD around 2 Å, close to the binding site of the natural ligand and bearing interactions with residues at the hydrophobic clefts, the Z-site and γ-Glu site. Analysis of RMSF graphs ( Figure 6) for TR show minor fluctuations located at loop regions prone to fluctuations; still, proteins remain mostly stable through the dynamics suggesting that a ligand interaction did not considerably affect the protein. The most notable change in RMSF may be seen in the region 102-152, where the RMSF increased slightly for T-147 with respect to the apoprotein. This RMSF supports the notion that this complex formed with T-147 is stable at its binding site and does not considerably alter other regions. It is noteworthy in the graph of the radius of gyration (Figure 7) that complexes have a stable radius throughout the 120 ns of molecular dynamics. When comparing receptors to complexes, there is no major difference. This finding suggests that protein remains compact during its dynamics analysis. Altogether it is possible to consider this in silico prediction to be a close representation of the inhibitory activity of these compounds on TcTR.

Glutathione Reductase Inhibition
The results obtained for the inhibitory assay of T-147 on human glutathione reductase show that the substrate concentration is not very relevant to the enzyme inhibition. This finding suggests that the compound may not directly interfere with disulfide binding.
Thus, both enzymes showed comparable Ki values for T-147, indicating that this derivative is not a selective TR inhibitor. Still, this result, together with the reported T-085 inhibitor [42], provides the basis for new chemical modifications to obtain a TR inhibitor with higher selectivity.
Modifications to R7 (from isopropyl to n-butyl) and R2 (from isopropyl to tert-butyl) on the quinoxaline ring of T-085 and T-147 modified the type of inhibition they had on TR. Considering this information, it is worthwhile to further investigate modifications to these positions to potentially attain a modification of the type of inhibition and selective inhibition of TR.

In Silico Analysis of Inhibitor Binding hGR
Additional molecular dynamics analysis permitted the finding that the stability of T-147 on hGR is significantly lower than on TcTR since it has a mean RMSD of 8.62 Å (Figure 8) in contrast to 2 Å with TcTR ( Figure 5), suggesting better inhibition. These results support that T-147 may not directly interfere with disulfide binding since it moves considerably from the binding site. The results observed for RMSF ( Figure 9) further emphasize the idea that T-147 moves from the initial binding site since there are fluctuations on at least three regions, residues in the ranges 118-158, 288-298, and 328-348; whereas there are nearly no fluctuation differences for TcTR. Just as for TcTR, the radius of gyration for hGR ( Figure 10) remains constant for each complex and apoprotein, supporting the idea that the ligand does not mainly destabilize protein, thus remaining compact throughout the dynamics analysis.

Chemical Synthesis
All reagents were purchased from chemical vendor Sigma-Aldrich (Mexico). The n-butyl and isobutyl quinoxaline-7-carboxylate1,4-di-N-oxide derivatives (from T-137 to T-170) were synthesized using the Beirut reaction described by Gomez-Caro et al [49]. The corresponding β-diketone (10.6 mmol) was added to the solution of the appropriate benzofuroxan-N-oxide (2.4 mmol) in dry chloroform (35 mL) while on an ice bath. Triethylamine (TEA) was added (1 mL), and the reaction mixture was stirred at room temperature for 4-7 days. The infrared (IR) spectra were obtained using the PLATINUM-ATR Bruker Alpha FT-IR spectrometer. The 1 H-NMR spectra were obtained in DMSO-d 6  strains were used for this study. These strains were isolated in Mexico from humans [50]. Epimastigotes were cultured in vitro in brain-heart infusion (BHI) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin (500 µg/mL (Gibco, Thermo Fisher, Ciudad de Mexico, Mexico) at 28 • C. The cultures were kept at exponential growth. They were transferred to fresh media every 7 days. CD1 mice, 6 to 8 weeks old, were infected with T. cruzi trypomastigotes to obtain trypomastigote samples in blood. The experiments were conducted in accordance with NOM-062-Z00-1999, published on 22 August 2009.

Trypanocidal Activity against Trypomastigotes
CD1 female mice, 6 to 8 weeks old, were infected with T. cruzi bloodstream trypomastigotes. The course of infection was followed for 4 to 6 weeks. At the peak of parasitemia (tracked every few days under a microscope), blood was drawn by cardiac puncture, and the sample was supplied with sodium heparin as an anticoagulant. The blood sample was adjusted to 1 × 10 6 bloodstream trypomastigotes/mL.
In a 96-well microplate, 195 µL of infected blood was added to each well along with 5 µL of a solution at different testing concentrations of n-butyl and isobutyl quinoxaline-7-carboxylate 1,4-di-N-oxide, and the reference drugs to reach a final volume of 200 µL. Each concentration (3.25-50 µg/mL) was assayed in triplicate. Initially, all the compounds were tested at a final concentration of 50 µg/mL (103-157 µM). Wells with untreated trypomastigotes, DMSO 5%, served as a negative control. Wells with reference drugs were used as positive controls. The microplates were refrigerated at 4 • C for 24 h. The bloodstream trypomastigotes were quantified by the Brenner-Pizzi method [51]; 5 µL were placed on the microscope slide and covered with a 13 × 13 mm glass cover. The motile parasites were counted in 15 microscope fields at 40× with an optical microscope. The percentage of mortality was calculated by comparing untreated wells with treated wells. The half-maximal lethal concentration (LC 50 ) was determined by Probit analysis.

Molecular Docking Analysis
Molecular docking analysis was conducted to evaluate the possible interaction of the n-butyl and isobutyl quinoxaline-7-carboxylate 1,4-di-N-oxide derivatives against T. cruzi TR. The ligand structures were drawn with Marvin Sketch 18.10, energy minimized with Chimera 1.12, and saved in pdb format. The protein receptor trypanothione reductase from Trypanosoma cruzi TcTR structure was obtained from the Protein Data Bank (PDB), access code 1BZL (accessed August 2018). Water molecules, co-crystalized ligands, and ions were removed from the protein structure. It was then energy-minimized using the Yasara Minimization Server (accessed August 2018) before docking simulation. All docking simulations were performed with AutoDock Tools 4.2 [52].
The molecular docking analysis was done on the interphase catalytic site coordinate space (X = 65.805, Y = 5.055, and Z = 0.955) using 20 Å in each axis with 1 Å spacing in the Grid Box. The results were analyzed considering the highest estimated free energy of binding (FEB) for each protein-ligand complex, which was later analyzed to determine the molecular interactions between protein and ligand utilizing the Protein-Ligand Interaction Profiler (PLIP) version 2.2.2 [53].

Molecular Dynamics
Molecular dynamics analysis was performed with GROMACS version 2018.4 [54]. First, the topologies for T-147 and mepacrine were generated with ACPYPE with the general amber force field (GAFF). Solvation was done with water molecules in a dodecahedron with a minimum distance from the wall of 10 Å, using the TIP3P water model. The necessary ions (Na + and Cl − ) were added to have a neutral charge in the system. The system was energy-minimized by the steepest descent algorithm. Then, two equilibrium steps were performed. First, the compound was simulated at NVT conditions (constant number of particles, volume, and temperature). For the second step, the compound was simulated at NPT conditions (constant number of particles, pressure, and temperature). Each stage achieved a duration of 100 ps. Finally, the simulation was performed at a temperature of 300 K for a trajectory of 120 ns [55,56]. The stability of the complex was determined using GROMACS built-in tools. The RMSD for the α carbons and the ligand were obtained. The RMSD matrix was calculated for the 120 ns. The RMSF for the α carbons was done to understand the effect of the compound on the secondary structure of TcTR and HsGR. Finally, the radius of gyration of each complex was obtained to determine the stability of the complex and the tridimensional compactness of TcTR and HsGR. 4.5. Enzymatic Activity Evaluation 4.5.1. Trypanothione Reductase Natural ligand trypanothione disulfide (TS 2 ) and recombinant Trypanosoma brucei TR were prepared following a previously reported methodology [57]. Stock solutions of inhibitors at a concentration of 5 mM were prepared in DMSO. Recombinant human glutathione reductase (hGR) was provided by Dr. Heiner Schirmer, Heidelberg, Germany. The kinetic analyses were conducted using a Jasco V 650 spectrophotometer and optical polystyrene cuvettes (10 × 4 × 45 mm). TR activity was determined at 25 • C in a total volume of 1 mL of TR assay buffer (40 mM HEPES, 1 mM EDTA, pH 7.5 containing 100 µM NADPH) and 5-10 mU of enzyme in the presence and absence of inhibitors [57,58]. Each assay contained, at most, a total of 5% DMSO, which is reported to be innocuous for enzymatic activity. The reaction was initiated by adding various concentrations of TS 2, and the consumption of NADPH at 340 nm was monitored. The activity was determined for TR inhibition in the absence and presence of two or three constant concentrations of the inhibitor at varying TS 2 concentrations (20, 40, 60, 100, and 200 µM). The type of inhibition was evaluated from double reciprocal Lineweaver-Burk plots. The inhibition constants Ki were calculated by non-linear regression using Graph Pad Prism software (version 5) [42,59,60].

Glutathione Reductase
GR activity was determined at 25 • C in a total volume of 1 mL of GR buffer assay (20.5 mM KH 2 PO 4 , 26.5 mM K 2 HPO 4 , 200 mM KCl, and 1 mM EDTA, pH 6.9). The assays contained 100 µM of NADPH, 5-10 mU of GR, and varying inhibitor concentrations. Each assay contained, at most, a total of 5% DMSO. The reaction was initiated by adding glutathione disulfide (GSSG) (37 and 92 µM), and the decrease of absorption at 340 nm was monitored. The activity of GR inhibition was determined in the absence and presence of two or three constant concentrations of the inhibitor at varying GSSH concentrations (21.5, 43, 86, 107.5, 372, and 930 µM). The type of inhibition was evaluated from double reciprocal Lineweaver-Burk plots. The inhibition constants Ki were calculated by nonlinear regression using Graph Pad Prism software (version 5) [42,59,60].

Conclusions
In this study, novel n-butyl and isobutyl quinoxaline-7-carboxylate 1,4-di-N-oxide derivates were obtained. Compounds T-141, T-142, and T-170 had the lowest LC 50 values against trypomastigotes of the NINOA strain. Compounds T-150, T-163, and T-169 had LC 50 values against INC-5 strain lower than the reference drugs Bnz and Nfx. SAR analysis showed that aromatic substituents are predominant as a substitution at the 2-position to enhance biological activity. Interestingly, compound T-147 showed good trypanocidal activity against both parasite forms.
In silico analysis identified the importance of the presence of aromatic groups as substituents for potential TR inhibition. Eleven of the thirty compounds analyzed had a FEB lower than −8 kcal/mol (−8.1 to −9.9 kcal/mol). Ten of these eleven compounds had an aromatic ring as part of their structures, supporting the importance of this structure. The top derivative was T-150 with FEB −9.9 kcal/mol. Most interactions were seen with Z-site (Phe396, Pro398, and Leu399) and γ-Glu site (His461, Glu466, and Glu467), which are both subsites within the interphase catalytic site, portraying them as potential inhibitors.
The enzymatic analysis of T-148, T-147, and T-150 against TR revealed that T-147 behaves as a mixed-type inhibitor with inhibition constants Ki and Ki' values of 11.4 and 60.8 µM, respectively, comparable to the known TR inhibitor mepacrine (Ki = 19 µM) [48]. However, T-147 showed a non-competitive inhibition of hGR with a Ki of 25 µM, which renders the compound rather non-selective towards TR. These results warrant further research on new quinoxaline 1,4-di-N-oxide derivatives with trypanocidal activity that may act as specific TR inhibitors.