Activity of β-Caryophyllene Oxide and Benznidazole Mixture Against Trypanosoma cruzi and In Silico Prediction of Anti-Trypanocidal Interaction

Abstract

Trypanosoma cruzi is the protozoan parasite that causes Chagas disease, affecting approximately 6–7 million people worldwide. The current treatment lacks efficacy in the chronic phase of the disease. This study aims to determine the in vitro synergistic activity of concomitant therapy (benznidazole with β-caryophyllene oxide) against Trypanosoma cruzi, assess its cytotoxicity, and propose the mechanism of this synergism through in silico analysis. The tested concentrations of the treatment demonstrated hemocompatibility (<5% hemolysis) and no cytotoxicity (>80% cell viability). Additionally, synergistic activity against the parasite was confirmed, reducing epimastigote viability by up to 80%. In this work, in silico analysis revealed that β-caryophyllene oxide also binds to the T. cruzi ABC channel in regions localized to amino acids 108–271 and 399–558, suggesting this interaction could inhibit it. This treatment emerges as a promising candidate for Chagas disease therapy. It lacks cytotoxic and hemolytic activity while exhibiting synergism against the parasite, such as through the inhibition of ABC channels, as suggested in silico.

1. Introduction

American trypanosomiasis, also known as Chagas disease (CD), is a zoonotic disease caused by the parasite Trypanosoma cruzi (T. cruzi), which is transmitted by a vector of the Reduviidae family, subfamily Triatominae. It represents a significant public health problem, primarily in the Americas. It is estimated that approximately 6 to 7 million people are infected with T. cruzi. This parasite is a protozoan of the order Kinetoplastea, characterized by a single mitochondrion whose DNA constitutes the kinetoplast, comprising 20% of the parasite’s DNA [1].

CD is characterized by two phases: acute and chronic. The acute phase has an incubation period of 10–14 days. After 20–30 years of infection, the chronic phase or symptomatic stage can occur in approximately 30% of patients, manifesting in four forms: cardiac, digestive, nervous, and congenital [1]. The variation in disease severity is attributed to the natural diversity of T. cruzi, as demonstrated in biological, biochemical, and molecular aspects [2].

Currently, the treatment for CD is limited to two drugs: nifurtimox (8–10 mg/kg) and benznidazole (5–7 mg/kg) [1]. However, these drugs can induce side effects such as genotoxicity, peripheral neuropathy, skin reactions, granulocytopenia, abdominal pain, anorexia, nausea, vomiting, and weight loss. Additionally, resistance to benznidazole has been observed [3,4,5]. No treatments are currently available for the chronic phase of the disease; however, palliative treatment is provided [1].

New treatment alternatives for CD have been explored in recent years, such as ergosterol inhibitors, natural products, and combination therapy [3,4]. Combination therapy is an alternative for treating pathogens such as HIV, Mycobacterium tuberculosis, and even protozoa like Plasmodium spp. [6,7,8]. In CD, combined therapy with different drugs and metabolites has been tested. For instance, β-caryophyllene oxide with lupenone reduced amastigote nests in cardiac tissue by 80% in an in vivo murine model that mimics the chronic phase of disease, achieving a synergistic pharmacological interaction [9]. Similarly, the combination of β-caryophyllene oxide with limonene showed the same pharmacological interaction across all three parasite stages; however, cytotoxicity was reduced in Vero cells in vitro [10].

Currently, the mechanism of action of β-caryophyllene oxide in T. cruzi is unknown. However, in liver and pancreatic cancer cells, it has been shown to inhibit ATP-binding cassette (ABC) transporter and modulate ABC transporter expression through the inhibition of STAT3 signaling and the MAPK/ERK pathway, which are associated with drug resistance in cancer [11].

Drug elimination through ABC transporters is a mechanism of resistance in protozoa such as Leishmania, Trypanosoma, Plasmodium, Trichomonas, Entamoeba, and others [12]. Some T. cruzi strains can exhibit drug resistance, which has been associated with the presence of ABC transporters such as T. cruzi P-Glycoprotein (TcPgP) and T. cruzi ABCG1 (TcABCG1) [13,14]. TcABCG1 is an ABC transporter of the G family, comprising 665 residues. Its gene is a single copy located on chromosome 37 of T. cruzi (CL Brener strain) [14]. Additionally, recent studies have demonstrated that benznidazole increased the expression of ABC transporters in HepG2 cells; furthermore, ABC transporters can expel benznidazole, reducing the intracellular concentration of the drug [15].

In this work, we report on the in vitro antiparasitic activity of concomitant benznidazole and β-caryophyllene oxide treatment against T. cruzi. Furthermore, we obtained the T. cruzi ABCG1 channel structure in silico by homology modeling, and docking of the ABC channel with β-caryophyllene oxide was performed to analyze whether this protein could be a target to enhance the effectiveness of the concomitant treatment and consequently decrease benznidazole cytotoxicity.

2. Materials and Methods

2.1. Benznidazole and β-Caryophyllene Oxide

Benznidazole (BNZ) and β-caryophyllene oxide (BCPO) were purchased from Sigma-Aldrich. Mixtures were prepared by combining 7 or 3.5 µg/mL BNZ (therapeutic-use concentration) and BCPO at 30 or 15 µg/mL (reported IC₅₀), dissolved in 0.5% dimethylformamide (DMF) [9,16].

2.2. Cell Culture

Macrophages (approximately 3 × 10⁶ cells/mice) were extracted from 3–6-week-old BALB/c mice weighing approximately 28 ± 2 g under basal conditions. The animals were maintained according to the principles and guidelines for the care and use of laboratory animals by the National Institutes of Health (NIH) and the Official Mexican Standard (NOM-062-ZOO-1999) [17,18]. The mice were euthanized, and 10 mL of cold phosphate-buffered saline (PBS) was injected into the peritoneal cavity. The peritoneal fluid was aspirated and centrifuged for 10 min (1109× g at 4 °C). The obtained cells were washed with PBS for 5 min (493× g at 4 °C) and resuspended in modified Dulbecco’s Modified Eagle Medium (DMEM) with 10% fetal bovine serum (FBS) and 1% penicillin–streptomycin. Then 2 × 10⁴ cells were seeded in each well of a 96-well plate and incubated for 72 h at 37 °C with 5% CO₂. To determine macrophage viability (>95%), the trypan blue method was used [19].

Epimastigotes of the Y (reference strain isolated from a patient in Brazil) and H4 (isolated in Yucatan, Mexico) T. cruzi strains were used in this study [20,21]. The parasites were cultured in liver-infusion tryptose (LIT) medium with 10% FBS at 27 °C [22].

2.3. Cell Viability Assay in Murine Macrophages

Murine peritoneal macrophages were placed in 96-well plates and treated with 100 µL of different mixtures (7 or 3.5 µg/mL BNZ with BCPO at 30 or 15 µg/mL). Subsequently, the macrophages were incubated for 24 h at 37 °C with 5% CO₂. The supernatant was eliminated by aspiration, and 100 µL of culture medium containing 0.5 mg/mL of MTT was added. Cell cytotoxicity and viability were assessed using 100% dimethyl sulfoxide (DMSO) and the vehicle (DMF), respectively. The plates were incubated for 24 h at 37 °C with 5% CO₂. After that, the supernatants were removed, and 100 µL of 100% DMSO was added to dissolve the formazan crystals. The plates were read at 550 nm using an ELISA plate spectrophotometer xMarK™ BIO-RAD, and the absorbance values were expressed as cell viability by comparing them to the control. Cells maintained without treatment were used as a negative control, and cells treated with dimethyl sulfoxide were considered a positive control for cytotoxicity [19,23].

2.4. Hemolysis Assay

Three milliliters of human blood samples was centrifuged for 5 min at 239× g. The plasma was removed, the blood cells were washed three times with PBS, and then the erythrocyte solution was diluted to 2%. In a 96-well plate, 100 µL of the BNZ-BCPO mixtures was added with 100 µL of the erythrocyte solution. The controls included distilled water as a positive control and PBS with the vehicle (5% DMF) as a negative control. The plate was gently shaken for 2 h and then centrifuged, and the supernatants were collected. The absorbance was read at 340 nm using an ELISA plate spectrophotometer (xMark™ BIO-RAD). The percentage of hemolysis was calculated with the absorbance values [24].

2.5. Trypanocidal Assay

Exponential-phase epimastigotes (1 × 10⁷ epimastigotes/mL) were added to each well of the 96-well plates and then treated with the mixtures (7 or 3.5 µg/mL BNZ with BCPO at 30 or 15 µg/mL). Benznidazole, 7 and 3.5 µg/mL, was used as a positive control, while parasites in LIT medium with 5% DMF were used as a negative control. The plates were incubated at 28 °C for 48 h. Then, the plates were centrifuged at 1613× g for 10 min, and the activity was evaluated using the resazurin method. Subsequently, 200 µL of a resazurin/PBS solution (0.3 M) was added. The plates were incubated under the same conditions for 24 h, and absorbance was measured at 560 nm using an ELISA plate spectrophotometer xMarK™ BIO-RAD. The activity was expressed as a viability percentage and as IC₅₀ (µg/mL) [9].

2.6. Drug Combination Analysis

Exponential-phase epimastigotes (1 × 10⁷ epimastigotes/mL) were added to each well of the 96-well plates and were then treated with mixtures of BNZ at 7 µg/mL with 120, 60, 30, 15, 7.5, 3.75, or 1.875 µg/mL BCPO, and BCPO at 30 µg/mL with 28, 14, 7, 3.5, 1.75, 0.875, or 0.4375 µg/mL BNZ. The individual concentrations of BNZ and BCPO were also evaluated. Parasites in LIT medium with 5% DMF were used as a negative control. The plates were incubated at 28 °C for 48 h. Then, the plates were centrifuged at 1613× g for 10 min, and the activity was evaluated using the resazurin method previously mentioned.

The activity was expressed in isobolograms and as IC₅₀ (µg/mL) [9,25].

The synergistic activity of the mixtures was calculated using the fractional inhibitory concentration (FIC), which is obtained as follows [9,26]:

FIC = Fractional Effect of compound A (FEa) + Fractional Effect of compound B (FEb);

FEa = IC₅₀a in mixture/IC₅₀a individual;

FEb = IC₅₀b in mixture/IC₅₀b individual.

Values of FIC < 1 indicate synergistic activity, FIC values of 1 represent an additive effect, and FIC > 1 indicates antagonistic activity [27].

2.7. Bioinformatics Tools

Bioinformatics analysis was conducted using various public databases and servers, including the National Center for Biotechnology Information, BLAST 2.16.0, SMART, Prosite, Swiss-Model, and Molecular Operating Environment 2022 (MOE; www.chemcomp.com (accessed on 26 March 2024)), which was used for docking analysis and TcABCG1-BCPO complex visualization [28,29,30,31,32,33].

2.8. Ligands

The PubChem database was used to download the 2D coordinates of BCPO (access code 1742210) and adenosine triphosphate (ATP) (access code 5957) [34]. BCPO meets some of the criteria of “Lipinski’s Rule of Five”: it has a molecular weight of 220.35 atomic mass units (amu) (<500 amu), one hydrogen bond acceptor (<10), and no hydrogen bond donors (<5) [35]. In MOE, polar and nonpolar hydrogens were added to the ligands (BCPO and ATP) and converted into a 3D structure.

2.9. Molecular Modeling of TcABCG1

The 3D model of the T. cruzi TcABCG1 transporter (ID Q4DW56, UniProt www.uniprot.org (accessed on 5 March 2024)) was obtained using the Swiss-Model software [32]. The crystallographic structure of the homologous Homo sapiens ABCG2 transporter (ID 7OJH, Protein Data Bank www.pdb.org (accessed on 6 March 2024)) was used as a template to generate numerous modeled structures. Among the generated models, the structure with the best packing index (QMEAN and GMQE ≥ 0.4) and with the position of the phi and psi torsion angles in the Ramachandran plot corresponding to the highest score was selected for further analysis. An energy minimization step was performed with the AMBER99 internal package program of MOE to obtain the final model [33]. The ligand and protein structures are protonated at pH = 7.4, T = 300 K, and salt = 0.1 M.

2.10. Docking Analysis

Binding sites or pockets on the surface were identified using an alpha-shape site-identification algorithm in MOE [36]. The pockets found on the surface were analyzed, and the putative binding sites were defined as regions of the structure that would interfere with the ABC transporter channel and ATPase recognition. Docking was performed using the Dock subroutine in the MOE package. Two hundred and fifty conformations per molecule were generated with a maximum torsion energy of 4 kcal/mol, and at least 15,000 different orientations were tested for the ligand conformer in each selected pocket within the interface region. Bases (such as amines) were protonated, acids (such as carboxyls) were deprotonated, and the double bonds C=C remained in their initially defined spatial configuration (cis or trans). The protonation state of histidine residues was verified using the AMBER99 force field under a pH of 7.4 implemented in the MOE software suite.

A database was created containing the coordinates and scores of the top five orientations for the compound. The evaluation criteria for determining the best orientations relied on a scoring function with three factors: the ligand’s compatibility with the binding site, the contact energy between the two molecules, and the internal conformational energy of the bound ligand. Docking was performed using a rigid docking approach, where the protein backbone and side chains were kept fixed during the simulation. Finally, the TcABCG1–compound complexes with the most favorable binding energies were selected and assessed. ATP was used as a control during the docking process, and it was fixed at P2TcABCG1, as expected.

2.11. Statistical Analyses

The statistical analyses were performed using GraphPad Prism 8.0. Data are presented as mean ± standard deviation (S.D.) and IC₅₀. For statistical analyses, one-way/two-way ANOVA and the Bonferroni post hoc test were used to compare treatments with the control (p < 0.05).

3. Results

3.1. Cytotoxic Activity

To assess the cytotoxic effects of concomitant treatment with benznidazole (BNZ) and β-caryophyllene oxide (BCPO), hemolysis and 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assays were conducted. The results from the hemolysis assay showed that all mixtures were non-hemolytic, with hemolysis less than 2% (p < 0.0001 compared to the positive control) (Figure 1a). The mixtures were considered non-hemolytic based on the criteria reported by Luna-Vazquez et al. (2021), which define low hemolysis as 5–10%, and values less than 5% are declared non-hemolytic [37]. Similarly, the results from the MTT assay for murine macrophage cell viability confirmed that all the different mixtures are non-cytotoxic, showing cell viability values greater than 83% (CC₅₀ values: BNZ: 41.53.7 ± 0.20 µg/mL, BCPO: 173.5.5 ± 0.35 µg/mL) (Figure 1b). Statistical analysis revealed no significant differences among the mixtures (p > 0.05). According to the ISO-109935:2009 standard, a treatment that exhibits a reduction in cell viability of no more than 30% can be considered a non-cytotoxic compound [38].

3.2. Trypanocidal Activity

The trypanocidal effect of the mixture was evaluated against the T. cruzi Y strain, and the results showed that the mixture of BNZ at 3.5 µg/mL with BCPO at 30 µg/mL exhibited the highest trypanocidal activity, superior to the compounds alone (BCPO and the reference drug). It inhibited viability by more than 85% (p < 0.05 compared with the negative control and the BNZ control at 7 µg/mL, which is the current treatment concentration). However, the interaction value of p = 0.1279 in two-way ANOVA indicates the absence of statistical synergy (Figure 2a).

The results obtained with the H₄ strain were similar, showing that the best trypanocidal activity was observed with treatment of the BNZ mixtures at 7 µg/mL and 3.5 µg/mL with 30 µg/mL BCPO. This treatment resulted in inhibition of more than 55% (p < 0.0001 compared with the negative control and the BNZ control at 3.5 µg/mL) and had statistical synergy in two-way ANOVA (interaction value of p = 0.0222) (Figure 2b). Conversely, BCPO alone had no effect on this strain compared to the Y strain, and the activity of BNZ was lower for the H4 strain of T. cruzi than for the Y strain.

The isobologram study demonstrated a decrease in the viability of T. cruzi in the BNZ-with-BCPO mixture compared to monotherapy activity. For the Y strain, the IC₅₀ values obtained for the mixtures were 0.90 ± 0.5 µg/mL for BNZ and 4.82 ± 0.17 µg/mL BCPO, and for monotherapies, 9.17 ± 0.02 µg/mL for BNZ and 36.20 ± 0.19 µg/mL for BCPO (Figure 3a,b). For the H₄ strain, the IC₅₀ values obtained for the mixtures were 4.67 ± 0.21 µg/mL for BNZ and 8.25 ± 0.32 µg/mL for BCPO, and for monotherapies, 42.56 ± 0.21 µg/mL for BNZ and 125.9 ± 0.32 µg/mL for BCPO (Figure 3c,d).

Although, statistically, the mixture of BCPO with BNZ only showed synergy in the H₄ strain, the synergism of each strain was confirmed (Figure 3e) with an FIC of 0.23 in the Y strain and 0.16 in the H₄ strain, respectively (FIC < 1) (

Table 1. IC₅₀, CC₅₀, and pharmacological interaction of evaluated mixtures against T. cruzi and macrophages.

	T. cruzi Epi/Cell (μg/mL)
	Y (IC50)	H4 (IC50)	Macrophages (CC50)
BNZ Mixture	0.90 ± 0.5	4.67 ± 0.21	41.53 ± 0.20
BCPO Mixture	4.82 ± 0.17	8.25 ± 0.32	173.5 ± 0.35
ΣFIC	0.23	0.16	1
Pharmacological interaction	Synergy	Synergy	No interaction or addition

Epi: Epimastigotes; IC50: inhibitory concentration 50; CC50: cytotoxic concentration 50; FIC: fractional inhibitory concentration; FIC < 1.0: synergistic activity; FIC = 1.0: additive effect; FIC > 1.0: antagonist activity.

). In contrast, the FIC against macrophages was 1, representing an additive effect [26].

3.3. Molecular Modeling Analysis of TcABCG1

To explore the potential binding interactions of BCPO with TcABCG1, a homology model of the target protein was constructed. The modeling of TcABCG1 was based on the homologous homodimer crystal structure of human ABCG2, which shares a sequence similarity of 48.3% with TcABCG1 and has the lowest resolution (3.10 Å) among other transporter structures with high similarity (the resolution range varies from 3.10 to 4.0 Å). The resulting model exhibited acceptable structural quality, with 93.36% of residues in the Ramachandran favored regions, a GMQE score of 0.51, and a QMEANDisCo Global score of 0.54, parameters that support the reliability of the model. Structurally, TcABCG1 closely resembled the human ABCG2 template, except for a few additional loops, indicated by arrows in Figure 4. The two pockets with the highest protein–ligand binding (PLB) scores were selected, shown in yellow and green in Figure 5a.

These pockets were also the widest openings and were identified as potential binding sites for docking analysis. The first pocket, labeled P1TcABCG1 (Figure 5b), showed a PLB score of 4.31. Both protein chains form this cavity, which is located just in the middle of the channel pore and is composed of the following amino acids: chain A (indicated in orange): Met 447, Phe 451, Val 543, Val 544, Ser 546, Gly 547, Thr 548, Pro 550, Val 551, Leu 551, Pro 555, Met 558; chain B (indicated in purple): Asn 399, Tyr 402, Ser 405, His 406, Ile 408, Gln 449, Phe 412, Phe 413, Ile 416, Asn 444, Met 447, Gly 448, Gln 449, Phe 451, Ile 452, Asn 455 (Figure 5b).

The second selected pocket, P2TcABCG1 (Figure 5c), was also formed by both chains, with a PLB score of 3.96. However, the ATP-binding region of the ABC channel, indicated in blue, was found in chain B. The amino acids that form this cavity are as follows: in chain A (shown in orange): Ser 108, Thr 241, Ser 242, Gly 243, Leu 244, Asp 245, Ser 246, Ser 249, His 271, Gln 272, Pro273, Thr 274; in chain B, indicated in purple color: Ser 102, Ser 108, Gln 109, Ala 110, Lys 112, Thr 113, Gln 155, Asp 238, Glu 239, Thr 241, Ser 242, Gly 243, Leu 244, Asp 245, ser 249, His 271.

3.4. Docking of β-Caryophyllene Oxide in the TcABCG1 Pockets

To predict the potential binding mode of BCPO within the modeled TcABCG1 transporter, molecular docking simulations were performed. The conformers of the TcABCG1-BCPO complex with the lowest energy were selected. For P1TcABCG1, the conformation with an energy of −5.89 Kcal and RMSD of 1.24 was chosen (Figure 6a). BCPO binds to the protein through hydrophobic interactions with amino acids Met A447, Pro A550, Phe A451, Met A447, Pro B550, Phe B541, and Leu B554. Additionally, it has polar interactions with Gly A448 and Ser A546 (Figure 6b). In P1TcABCG1, BCPO is located in the channel conduit (formed by chains A and B) (Figure 6c).

BCPO exclusively interacts with chain B in the P2TcABCG1 pocket, even though this pocket is formed by both dimer chains (Figure 7a). The energy obtained for the selected conformation was −4.85 kcal, and the RMSD was 1.62. This protein–ligand binding interaction also includes polar interactions with the amino acids Ser 107, Ser 108, Thr 113, Gln 155, Ser 242, and His 271 of chain B. It also features other interactions, such as hydrophobic interactions with Ala 110, basic interactions with Lys 112, and acidic interactions with Asp 238 and Glu 239 of chain B (Figure 7b). In this binding mode, His271 was modeled in its neutral HID tautomeric form, with the proton located at the ND1 position of the imidazole ring. This assignment is consistent with its pKa value of 6.29, indicating that at physiological pH (7.4) the residue remains predominantly uncharged. Under these conditions, the imidazole ring can act as a proton acceptor in a weak π-H hydrogen bond, formed between the aromatic system of His271 and one of the donor hydrogens of BCPO.

The ATP used as a control for P2TcABCG1 generated an energy of −5.94 kcal and an RMSD of 1.45, interacting with Thr113 and Thr114 (Figure 7c). The TcABCG1 residues involved in contact with BCPO, such as Thr113, are essential for ATP phosphorylation, acting as the side-chain donor (Figure 7d).

Figure 7. Interaction between P2TcABCG1 and BCPO: (a) The residues interacting with BCPO are in chain B of the protein. The hydrogen bond (green dashed line) between His 271 and C2 of BCPO can be observed. ATP binding is highlighted in red; (b) β-caryophyllene oxide (BCPO) exhibits polar interactions with the residues Ser 107, Ser 108, Thr 113, Gln 155, Ser 242, and His 271; acidic interactions with the residues Asp 238 and Glu 239; basic interactions with Lys 112; and hydrophobic interactions with Ala 110. These interactions occur with carbons 7, 8, and 11 and all methyl groups of BCPO. Carbon 2 forms a hydrogen bond with the ring of His 271; (c) ATP (lemon color) interacts with the residues from Ser 108 to Thr 113 (the ATP-binding motif, shown in red). Thr 113 and Thr 114 have clashes with ATP (yellow dashed line); (d) ATP presents polar interactions with Ser 107, Ser 108, Gly 109, Gly 111, Thr 113, Thr 114, Gln 155, Ser 242, and His271; hydrophobic interactions with Ile 88 and Ala 110; basic interactions with Lys 112; and acidic interactions with Glu 239. Thr 113 is a side-chain donor to the first phosphate group of ATP.

Figure 7. Interaction between P2TcABCG1 and BCPO: (a) The residues interacting with BCPO are in chain B of the protein. The hydrogen bond (green dashed line) between His 271 and C2 of BCPO can be observed. ATP binding is highlighted in red; (b) β-caryophyllene oxide (BCPO) exhibits polar interactions with the residues Ser 107, Ser 108, Thr 113, Gln 155, Ser 242, and His 271; acidic interactions with the residues Asp 238 and Glu 239; basic interactions with Lys 112; and hydrophobic interactions with Ala 110. These interactions occur with carbons 7, 8, and 11 and all methyl groups of BCPO. Carbon 2 forms a hydrogen bond with the ring of His 271; (c) ATP (lemon color) interacts with the residues from Ser 108 to Thr 113 (the ATP-binding motif, shown in red). Thr 113 and Thr 114 have clashes with ATP (yellow dashed line); (d) ATP presents polar interactions with Ser 107, Ser 108, Gly 109, Gly 111, Thr 113, Thr 114, Gln 155, Ser 242, and His271; hydrophobic interactions with Ile 88 and Ala 110; basic interactions with Lys 112; and acidic interactions with Glu 239. Thr 113 is a side-chain donor to the first phosphate group of ATP.

4. Discussion

As the disease progresses, treatment efficacy decreases, and in older individuals, side effects become more aggressive. Therefore, it is essential not only to enhance treatment effectiveness but also to minimize potential adverse effects [1,3,4].

When erythrocyte and macrophage cells were treated with BNZ-BCPO, hemolysis and viability were <5% and more than 80%, respectively. This result confirms that the combination does not exhibit hemolytic activity or cytotoxicity. According to Luna-Vazquez et al. [37] hemolytic activity lower than 5% is considered non-hemolytic, meaning it does not induce red blood cell membrane damage. However, this assay does not demonstrate intracellular damage [37,39]. Thus, the cytotoxicity of the concomitant treatment was evaluated through the mitochondrial viability of murine macrophages using the MTT assay, which involves the reduction of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium by the mitochondrial enzyme succinate dehydrogenase to produce formazan. A treatment that does not damage mitochondria shows cell viability greater than 80%. According to the ISO 10993-5:2009 standard, these can be considered non-cytotoxic treatments. When BNZ and BCPO were mixed, no significant changes in cytotoxicity were observed. In the report by Moreno et al. (2018), the mixture of limonene with β-caryophyllene oxide reduced the toxicity of limonene in Vero cells compared to the control [10,23,40].

The BNZ-BCPO mixtures showed synergistic activity against T. cruzi epimastigotes in vitro in two parasite strains (H₄ and Y). The best activity in both T. cruzi strains was obtained with the mixture of BNZ at a dose of 3.5 µg/mL with BCPO at 30 µg/mL. Even at the lowest BNZ concentration, inhibition of the Y strain was more than 85%, compared to 55% for BNZ and 21% for BCPO as monotherapies. However, with the H₄ strain, the mixture inhibited more than 55% of the parasites. In contrast, the BNZ and BCPO monotherapies inhibited only 27% and 5%, respectively, indicating that BCPO enhances the BNZ efficacy, yielding better results with a lower dose.

It is important to mention that although T. cruzi epimastigotes are considered a non-infective stage, studies have demonstrated that they can exhibit infective potential through receptors characteristic of infective stages, such as trypomastigote trans-sialidase (tTS) [41,42,43]. Different immunological and molecular studies utilize epimastigotes due to their exponential growth in axenic cultures. Additionally, the sequences of some proteins, such as TcABCG1, were obtained from this stage [44,45,46].

The mixture of BNZ-BCPO at doses of 3.5 µg/mL and 30 µg/mL did not exhibit either cytotoxicity or hemolysis in macrophages. It showed antiparasitic activity, suggesting that this mixture is a potential treatment targeted at parasites with fewer effects on vertebrate cells. This parasite-targeting selectivity is supported by a selectivity index (CC₅₀/IC₅₀) greater than 10.

Although synergistic activity was observed in both strains, it was greater in H₄ than in the Y strain. This variability in treatment efficacy between these two strains may be attributed to pleomorphism. Currently, six discrete units (DTUs) are described (TcI, TcII, TcIII, TcIV, TcV, and TcVI). The Y strain belongs to TcII, and the H₄ strain to TcI [5,47]. According to Moreno et al. [48] strains are considered susceptible to BNZ if they have an IC₅₀ ≤ 4.3 µg/mL, whereas they are considered resistant if they have an IC₅₀ ≥ 8.3 µg/mL. In our study, the Y strain showed an IC₅₀ of 9.17 ± 0.02 µg/mL, classifying it as a resistant strain, and the H4 strain showed an IC₅₀ of 42.56 ± 0.21 µg/mL, considered highly resistant [48]. This difference in susceptibility may imply that the effect of the mixture could vary among untested parasite strains.

It has been demonstrated that strains belonging to TcI are less susceptible to BNZ, with an LC₅₀ twice as high as strains belonging to TcII [2,5]. Therefore, the reduced effect of the mixture against the H₄ strain compared to the Y strain could be due to inter-DTU factors; however, it is important to mention that despite the partial resistance of the H₄ strain to BNZ, a synergistic effect was observed in both strains, confirming the potential of this mixture.

The drug resistance observed in different DTUs may be related to ABC transporters. El Awady et al. [12] described how some protozoa, such as Leishmania spp., Trypanosoma brucei, and Trypanosoma cruzi, utilize ABC transporters as a resistance mechanism against several drugs by expelling them from the cell. Nevertheless, this is not the only mechanism of drug resistance [12]. Recently, the TcABCG1 gene, which encodes an ABC transporter, was reported to be overexpressed in T. cruzi strains that are resistant to BNZ [46].

Resistance to treatment through ABC channels has also been observed in other cells, such as cancer cells, and drugs to inhibit these channels are under investigation [49]. Recently, Di Giacomo et al. [11] reported the mechanism of ABC channel inhibition by BCPO and some signal transduction involved in the synthesis of these transporters in the liver, bile, and pancreatic cancer cells [11]. However, it is unknown how BCPO acts on T. cruzi. Therefore, molecular docking was performed to observe whether BCPO binds to inhibit one of the ABC channels (TCABG1) that promote resistance to BNZ treatment.

The results obtained in molecular docking showed that in the P1TcABG1 pocket, BCPO interacts with Phe451 in both chains. In the HsABCG2 protein used as a template, Phe 439 is utilized to excrete xenobiotics into the extracellular space [50]. Therefore, BCPO may occupy residues necessary for transportation, acting as a competitive inhibitor.

In the case of the P2TcABCG1 pocket, BCPO interacts with the residues Ser 108, Gln 109, Ala 110, Lys 112, and Thr 113 in chain B. All these residues belong to the ATP-binding region. Therefore, the binding of BCPO to this site could prevent ATP binding. Furthermore, BCPO forms a weak hydrogen bond with the imidazole ring of His 271. Weak hydrogen bonds are generally defined as those with energies less than 4 Kcal/mol, such as π-H interactions, which offer several advantages in molecular recognition. They exhibit improved desolvation characteristics, have enhanced specificity due to the retention of their electrostatic character, and enable water molecules to occupy stable positions even in otherwise challenging environments, in contrast to conventional hydrogen bonds. These weak interactions can be functionally interchangeable with conventional hydrogen bonds. In the context of drug design, this interchangeability is highly advantageous, as it allows for the modification of a ligand’s non-binding-related properties, such as cell permeability or metabolic stability, without compromising binding affinity [51]. The nucleotide-binding domain (NBD) of the protein can exhibit variations in amino acid residues among resistant parasite strains, according to Zingales et al. (2015); however, these residues do not affect the BCPO interaction [14].

The docking analysis suggests that BCPO could inhibit TcABCG1 transporters at two potential sites: a pocket formed in the transmembrane region and the ATP-binding region. Both sides could be inhibited, and when the pocket is empty, BCPO acts primarily on the P1TcABCG1 pocket due to its higher affinity. However, a key limitation of this study is the absence of a crystallographic structure for the target protein, which required the use of a homology model for docking simulations. Although the selected template presented high-resolution structural data, the reliability of the model is inherently constrained by the sequence identity and coverage between the template and the target. Consequently, the predicted binding interaction is necessary to confirm the biological relevance of the interaction proposed in silico.

The binding sites for BCPO identified in our modeling differ from those proposed by Di Sotto et al. [52], who modeled BCPO on the P-gp transporter of HepG2 cells. They propose that BCPO binds to the same site as verapamil (an inhibitor of the P-gp transporter), which is a region near the ATP-binding domain. However, this pocket was not observed in TCABCG1. Additionally, in the P-gp channel of HepG2 cells, BCPO exhibits a higher affinity with a binding energy of −19.73 Kcal compared to TcABCG1, where the highest binding energy was −5.89 Kcal [52]. Our model proposes BCPO as a potential inhibitor of the ABC transporter in T. cruzi.

BNZ is also metabolized by the host cell; however, BCPO would increase the production of glutathione peroxidase 4 (GPX4) and the concentration of glutathione (GSH) through glutathione synthetase. GPX4 reduces ferroptosis, and GSH, an endogenous antioxidant, neutralizes free radicals and oxidizing species like glyoxal, a metabolite produced by BNZ. Since the parasite lacks the glutathione synthetase gene, it likely would not be able to take advantage of BPO’s antioxidant effect [13,53]. Therefore, the mixture of BNZ-BCPO could potentially reduce the cytotoxic effect of BNZ on the host cell, allowing it to target the parasite primarily.

The results obtained in our study demonstrate that treating BNZ with BCPO is an important combination therapy strategy for Chagas disease treatment. This combined therapy has been used since 2001 in other protozoan diseases, such as malaria, and it is supported by the World Health Organization (WHO) [7,54].

5. Conclusions

The mixture of 3.5 µg/mL BNZ with 30 µg/mL BCPO exhibited the highest trypanocidal activity against T. cruzi. Neither cytotoxic activity nor a hemolytic effect was detected, indicating synergism. Thus, this concomitant treatment is a potential candidate for combined therapy to combat the parasite. The reduction in BNZ concentration could be attributed to the inhibition of the parasite’s TcABCG1 channel, thereby reducing drug elimination through these channels. However, further studies are necessary to determine this interaction and the efficacy of the mixture as a treatment in the acute and chronic stages of Chagas disease in vivo. Efflux and crystallization studies are also required to confirm the mechanism of BCPO on TcABCG1 channels.