Repurposing of rabeprazole as an anti-Trypanosoma cruzi drug that targets cellular triosephosphate isomerase

Abstract

Trypanosoma cruzi is the causative agent of American trypanosomiasis, which mainly affects populations in Latin America. Benznidazole is used to control the disease, with severe effects in patients receiving this chemotherapy. Previous studies have demonstrated the inhibition of triosephosphate isomerase from T. cruzi, but cellular enzyme inhibition has yet to be established. This study demonstrates that rabeprazole inhibits both cell viability and triosephosphate isomerase activity in T. cruzi epimastigotes. Our results show that rabeprazole has an IC₅₀ of 0.4 µM, which is 14.5 times more effective than benznidazole. Additionally, we observed increased levels of methyl-glyoxal and advanced glycation end products after the inhibition of cellular triosephosphate isomerase by rabeprazole. Finally, we demonstrate that the inactivation mechanisms of rabeprazole on triosephosphate isomerase of T. cruzi can be achieved through the derivatization of three of its four cysteine residues. These results indicate that rabeprazole is a promising candidate against American trypanosomiasis.

Keywords:

• Chagas disease
• glycolysis
• proton pump inhibitors
• methylglyoxal
• molecular docking

Introduction

Chagas disease (CD) was first reported by Carlos Chagas over a century ago when he discovered the first paediatric patient infected with Trypanosoma cruzi in Brazil. In 2005, the WHO recognised CD as a neglected tropical disease located mainly in Latin America. However, in the last decade, the migration of infected people with T. cruzi has facilitated the disease’s efficient spread, reaching countries such as Australia and some regions of Europe and Africa, as well as countries in the Eastern Mediterranean and Western Pacific¹. To prioritise the care and prevention of CD, the WHO instituted April 14 as World Chagas Disease Day in 2020².

Worldwide, CD causes approximately 12,000 deaths per year, with 6 to 7 million people infected, mainly in Latin America, and with a risk of 70 million people contracting it¹. T. cruzi which parasitises more than 100 species of mammals, including humans, is the causative agent of CD. This organism has consistently developed in the tropical and subtropical regions of the American continent, leading to CD also known as American trypanosomiasis. It is primarily transmitted by the bite of haematophagous triatomine insects of the Reduviidae family. It is known that infection by T. cruzi is a complex zoonosis involving more than 100 species of hemipteran insects as well as more than 100 species of mammals³. Other transmission mechanisms have been reported, such as the oral route (food and water contaminated with triatomine excretions), blood transfusions, congenital transmission (mother to child), organ transplantation, laboratory accidents, and even sharing needles among drug users^4,5. Non-vector-associated transmission mechanisms have facilitated the spread of CD with congenital transmission rates ranging from 2 to 12% in endemic countries⁶.

Currently, Benznidazol (Bzn) and nifurtimox with over 40 years of use, are the only relatively effective treatments for acute CD. However, their efficacy against the chronic stage of the disease is uncertain in some cases, and their therapeutic regimens can extend for several months with common toxic side effects⁷. One of the most reported adverse effects for Bzn are dermatological alterations and gastrointestinal complications⁸ with a higher incidence in females and young adults. A study involving patients with established Chagas cardiomyopathy suggested that Bzn treatment may be less effective in those with chronic CD⁹. Regarding nifurtimox, a CDC report between 2001 and 2021 indicates that the patients’ most common adverse effects were nausea, anorexia, weight loss, headache, and abdominal pain, with some reporting severe adverse events as depression, peripheral neuropathy, paresthaesia, dizziness/vertigo¹⁰. Moreover, treatment-refractory strains of parasites are frequently observed^11–13. The absence of approved and efficient drugs to combat CD has contributed to the disease’s increasing incidence. Consequently, research efforts have increased over the past decades to identify new anti-chagasic compounds and therapeutic targets.

Studies focussed on drug discovery for the treatment of CD follow two main lines of research: phenotypic screening of compound libraries and target-based drug discovery¹⁴. High throughput screening assays have been conducted on thousands of natural and synthetic compounds, and a few have advanced to clinical trials listed in the Clinical Trials database (https://clinicaltrials.gov)¹⁵. Some therapeutic targets studied include sterol-14-α demethylase (CYP51)¹⁶, the 80 kDa enzyme prolyl oligopeptidase (Tc80)¹⁷, cysteine peptidase cruzipain^18,19, the proteasome²⁰, trypanothione reductase²¹, superoxide dismutase (Fe-SOD) and triosephosphate isomerase (TIM)²². TIM catalyses the interconversion of DHAP and GAP, which is a vital mechanism for the complete catabolism of glucose and the generation of net energy in the form of ATP and metabolites such as NADH. TIM is located in the glycosome, which is a distinctive organelle from kinetoplastid protozoans such as Leishmania spp, T. brucei, and T. cruzi²³. Glycosomes are essential for parasite survival in the bloodstream since glycolysis is the only source of ATP at this stage²⁴.

Numerous compounds, including thiazoles and benzothiazoles, thiadiazoles, oxathiazoles, thioles, tricyclic compounds, benzimidazoles, phenazines, and thiadazines, have been tested against recombinant TIM from T. cruzi (TcTIM)²⁵, in approaches for enzyme inhibition in vitro. The main mechanisms reported so far are based on homodimer affectation through destabilisation of the interface^26–28; or adjacent regions surrounding Cys118²⁹. Likewise, the perturbation of the interaction network between loop 3 and the adjacent subunit has also been studied as a mechanism of action of some inhibitors. Nevertheless, to date, none of these studies have shown that inhibition of this glycolytic enzyme is a clear target within the parasite.

Drug repurposing has recently become a well-intentioned strategy that contributes to reducing time, costs, and risks in drug discovery for the treatment of various diseases³⁰. Some researchers have taken on the task of developing novel antiparasitic compounds with currently approved drugs. For example, the non-steroidal anti-inflammatory drug nimesulide³¹ and chloroquine³² have been proposed as alternatives for the treatment of CD. Other molecules with high potential in drug repurposing are proton pump inhibitors (PPIs), which are drugs widely used to relieve symptoms of acid reflux and gastroesophageal reflux disease (GERD). Omeprazole is the most representative and commonly used, other PPIs as pantoprazole, lansoprazole, esomeprazole, and rabeprazole have been used with success. In fact, these drugs have been proposed as anticancer and anti-protozoal agents using in vitro tests and in animal models, with very promising results in both^33–36. For example, rabeprazole (Rbz) has been proposed as an antiparasitic agent against Giardia lamblia and Entamoeba histolytica^36,37. This drug is a substituted 2-pyridyl methylsulfinyl benzimidazole, Rbz was the third drug to emerge from the PPI family, it is prescribed short-term to treat GERD symptoms in both adults and children who are at least 1 year old. Similar to other PPIs, Rbz binds to some cysteine residues and inactivates the H+/K+-ATPase that is a proton pump expressed on the surface of the gastric parietal cells, causing the flow of protons through the pump to be permanently blocked, thereby resulting in a decrease in gastric acid.

To date, at least one of the PPIs (omeprazole) has been shown to have some activity against trypanosomatids (T. cruzi, Leishmania donovani, and T. brucei)^38,39. In both works, it was demonstrated successfully that such a drug inhibits the growth of parasites without the identification of a specific target. However, until date, there are no studies demonstrating the effectiveness of Rbz on T. cruzi.

In this work, we identified the glycolytic enzyme TIM as a target of Rbz. Our results show that TIM is an excellent target for the Rbz, which causes a drastic decrease in enzymatic activity leading to parasite death. Notable consequences of the ablation of TIM activity are increases in the levels of the highly toxic MGO and AGEs. We also show that three of the four cysteine residues in the amino acid sequence of recombinant TcTIM are derivatized by Rbz and cause the inactivation of the enzyme.

In conclusion, our findings indicate that TcTIM is a promising target for drug development against American trypanosomiasis, and Rbz can be suggested as a safe, effective alternative with fewer side effects and low cost.

Material and methods

During all the experimental procedures, most of the reagents were acquired from Sigma–Aldrich (St. Louis, MO, USA), except for the following. From Roche (Penzberg, Upper Bavaria, Germany) was acquired Glycerol-3-phosphate dehydrogenase (α-GDH) and reduced nicotinamide adenine dinucleotide (NADH); from AMRESCO LLC (Cochran Road Solon, OH, USA) was acquired Luria Bertani (LB) medium and isopropyl-β-D-thiogalactopyranoside (IPTG); from Thermo Scientific was acquired bicinchoninic acid; from Bio-Rad (Hercules, California, USA) was acquired IMAC resin and from Amersham Biosciences (Amersham, UK) was acquired Sephadex G-25 Fine Resin. Finally, from Merck-Millipore Corporation (Billerica, Massachusetts, USA) was acquired Amicon Ultra 30 kDa filters. The PPIs were purchased from Sigma with the purity of 99.9%, Benznidazole (Lot 130302), 99% purity was obtained from LAFEPE; Pharmaceutical Laboratory of Pernambuco State, Recife, Brazil.

Cultures of epimastigotes from Trypanosoma cruzi, strain CL brenner

Epimastigotes from T. cruzi strain CL Brener, TINF/BR/1963/CL-Brener (T. cruzi II) (CL-Brener) (International Symposium to Commemorate the 90th Anniversary of the Discovery of Chagas Disease 1999) were grown in LIT medium supplemented with 10% new-born calf serum at 28 °C. For inhibition curves, 10 ml cultures were seeded with 1 × 10⁶ epimastigotes and incubated at 28 °C until reaching 10 × 10⁶ epimastigotes per mL⁴⁰.

Determination of Rbz effect on viability and activity of cellular TIM from T. cruzi epimastigotes

1 × 10⁶ parasites were taken from epimastigote cultures in the exponential phase, and incubated at 28 °C with varying concentrations of Rbz (100–2000 μM) for 4 h. After incubation, an aliquot was withdrawn to determine cell viability using MTT assay⁴¹, and cultures were centrifuged at 500 × g for 15 min at 10 °C. The pellet with the parasites was resuspended in PBS to obtain 20,000 parasites/μL and the suspension was lysed by five freeze-thaw cycles. The activity of cellular TIM was determined from the protein extracts.

Determination of effect on viability and activity of cellular TIM by Rbz using different numbers of epimastigotes from T. cruzi

From epimastigote cultures in the exponential phase, a different number of parasites (from 1 × 10⁵ to 2 × 10⁶) were taken and incubated with 250 μM of Rbz for 24 h or with 1 mM of Rbz for 4 h at 28 °C. An aliquot was withdrawn after incubation to determine cell viability, using MTT assay, and cultures were centrifuged at 500 × g for 15 min at 10 °C. A pellet containing parasites was resuspended in PBS to obtain 20,000 parasites/μL, and the suspension was lysed by five freeze-thaw cycles. Total protein was determined from the protein extracts using the Pierce BCA protein assay (Thermo Scientific, Whaltman, MA, USA), and the activity of cellular TIM was also determined.

Enzyme activity assays

The enzymatic activities of cellular TIM from epimastigotes and recombinant TcTIMs were assayed as described by Gómez-Puyou et al.⁴² as follows. The conversion of glyceraldehyde 3-phosphate (GAP) to dihydroxyacetone phosphate (DHAP) was monitored spectrophotometrically by the oxidation of the cofactor NADH at 340 nm using glycerol-3-phosphate dehydrogenase (α-GDH) as the coupling enzyme. For enzymatic assays, was used a reaction mixture that consisted of 1 mM GAP, 0.2 mM NADH, and 0.9 units/mL of α-GDH in TE buffer. The assays were initiated by adding to the reaction mixture, an aliquot (60 μg/mL) of total protein extracted from epimastigote cultures treated with Rbz or 5 ng/mL of recombinant TcTIM under a constant temperature of 25 °C.

Determination of IC₅₀

Fifty thousand epimastigotes from T. cruzi were suspended in 1 ml of LIT medium and incubated at 28 °C for 48 h with increasing concentrations (0, 0.01, 0.025, 0.05, 0.075, 0.1, 0.2, 0.4, 0.8, 1, 1.25 and 1.5 μM) of Rbz or (0, 0.05, 0.1, 0.25, 0.5, 0.75, 1.5, 3, 6, 12, 25, 30, 40 and 50 μM) of the reference drug Bzn. At the end of the incubation time, aliquots of 50 µL per condition were withdrawn, and the cells were read three times. The cells were reseeded and adjusted to 1 ml with LIT medium, and they were incubated for another 48 h under the mentioned conditions. At the end of the assays, cell counts were performed using a Neubauer chamber after mixing them with trypan blue. The half-maximal inhibitory concentration (IC₅₀) was calculated by means of Probit analysis (SPSS package).

Quantification of MGO and AGEs

Intracellular free MGO was measured spectrophotometrically by using DNPH according to the method of Gilbert and Brandt⁴³ with a few modifications⁴⁴. Briefly, both quantifications, of MGO and AGEs, were determined using 1 × 10⁶ epimastigotes which were exposed to different concentrations of Rbz, from 100 to 2000 μM, for 4 h at 28 °C. After incubation time, cultures were centrifuged at 500 × g for 15 min at 10 °C, and the pellet containing parasites was resuspended in PBS to obtain 20,000 parasites/μL. The suspensions were lysed by five freeze-thaw cycles, after which 0.45 M perchloric acid was added to each sample, chilled on ice for 10 min, and centrifuged at 3000 × g at 4 °C for 10 min. The supernatant was collected and stored at −70 °C for further measurements. Before determining the concentration of MGO in the samples, we made a standard curve of MGO. Stock solutions of 20 mM DNPH in HCl-ethanol (12:88) and 0.1 mM MGO in distilled water were prepared. Increasing concentrations of MGO (from 0 to 10 μM) were incubated with 0.2 mM DNPH at 42 °C for 45 min. These samples were cooled for 5 min at room temperature, and the absorbance of MGO-bis-2,4-dinitrophenylhydrazone was recorded at 432 nm on a microplate spectrophotometer (Epoch, BioTek, Winooski, VT, USA). Subsequently, the cell supernatants from epimastigotes treated with Rbz were taken and used to quantify MGO levels with DNPH. The MGO concentrations from the cells and the standard were estimated using the extinction coefficient ε = 33,600 M⁻¹cm⁻¹ for MGO-bis-2,4-dinitrophenyl-hydrazone and the standard curve. The assays were performed in quadruplicate and the results are expressed as [μM] MGO/1 × 10⁶ cells.

AGEs were determined by using an AGE ELISA kit (MyBioSource, San Diego, CA, USA) following the manufacturer’s instructions. The protein concentration in the samples was adjusted to 1 mg/mL, and the samples were diluted 1:100 and loaded onto ELISA plates to determine the AGE concentration. Avidin-peroxidase conjugates were added to ELISA wells, and 3,3′,5,5′-tetramethylbenzidine (TMB) was used as the substrate for colouring (after the reactant was thoroughly washed out with PBS). A standard curve was made with the AGE standard included in the kit. Standard concentrations were 0, 3.12, 6.25, 12.5, 25, 50, 100, and 200 ng/mL. The absorbance at 450 nm was measured within the first 10 min using an Epoch microplate spectrophotometer (Epoch, BioTek, Winooski, VT, USA). The results are the mean of four independent experiments and are expressed as μg of AGEs/mL.

Western Blot analysis of cellular TIM from epimastigotes treated with Rbz

Epimastigotes (1 × 10⁶) from T. cruzi were incubated with different concentrations of Rbz (100–2000 μM) for 4 h at 28 °C. After incubation, the cells were lysed by five freeze-thaw cycles, and the total protein was determined by the Pierce BCA protein assay (Thermo Scientific, Whaltman, MA, USA). 100 µg/lane were loaded on a 16% SDS-PAGE gel. After, proteins were transferred to a polyvinylidene difluoride (PVDF) membrane (0.8 mA/cm², 2 h) in 25 mM Tris buffer containing 192 mM glycine and 20% methanol. The membrane was blocked for 1 h with Tris-buffered saline with 0.1% Tween-20 (TBS-T) supplemented with 5% bovine serum albumin (BSA), washed one time with TBS-T, and incubated overnight at 4 °C with anti-TcTIM diluted 1:1000 in TBS-T containing 1% BSA. The membrane was then washed three times with the same buffer, and an anti-rabbit IgG secondary antibody (diluted 1:3000) conjugated with horseradish peroxidase (HRP) (Anti-rabbit IgG, HRP-linked Antibody #7074, Cell Signalling Technology, Danvers, Massachusetts, USA) was used to reveal the immunoblot bands by chemiluminescence (Clarity Western ECL substrate, Bio-Rad) following the supplier’s instructions. Blot image acquisition was performed using a Chemi-Doc XRS + system (Bio-Rad Laboratories, Inc., USA). Additionally, β-actin was used as a reference and identified with antibody against β-actin (β-actin (C4): sc-47778, Santa Cruz, Biotechnology, Dallas, Texas, USA) diluted 1:1000 in TBS-T, 1% BSA, and visualised using an anti-mouse IgG secondary antibody conjugated with HRP (Anti-mouse IgG, HRP-linked Antibody #7076, Cell Signalling Technology, Danvers, Massachusetts, USA), diluted 1:3000 and revealing the immunoblot bands by chemiluminescence.

Production of anti-TcTIM polyclonal antibodies

Two female rabbits (Oryctolagus cuniculus) (2 months old) were immunised with TcTIM. Immunisation was performed according to NOM-062-Z00-1999. Eight intramuscular injections were administered with 100 µg of TcTIM at intervals of two weeks between each inoculation. The initial immunisation dose was performed with Freund’s incomplete adjuvant (FIA), and the remaining doses were administered without adjuvant. To test for antibody production, a blood sample was collected from the rabbit’s ear artery a week after each inoculation every 15 d. After completing the immunisation schedule, a cardiac puncture was performed on each rabbit after previous treatment with 44 mg of sodium pentobarbital/kg of body weight.

Construction of TcTIM Cys-mutants by site-directed mutagenesis

The DNA sequence U53867 for TcTIM at the NCBI database was used to create single mutants. All mutants except the C15A mutant were constructed on a pET-HisTEVP-modified plasmid⁴⁵ using the QuickChange protocol (Agilent Technologies, CA). The C15A mutant was constructed using the plasmid pET-3a containing the WT TcTIM sequence as a template and introducing the mutation with the polymerase chain reaction (PCR) using Vent DNA Polymerase (New England Biolabs, MA, USA). The external nucleotides were the T7 Promoter and the T7 Terminator (Novagen). The mutation was introduced as follows: 25 cycles for 1 min at 94 °C, 1 min at 55 °C, and 1 min at 72 °C and the extension incubation for 10 min at 72 °C. The PCR product was cloned into the pET-3a expression vector (Novagen) as an NdeI-BamHI fragment.

The mutagenic oligonucleotides used to produce TcTIM Cys-mutants are shown in Supplementary Table S6. The plasmids containing the different mutants were transformed into the Escherichia coli BL21 Codon Plus (DE3)-RIL strain (Agilent Technologies, CA).

Expression and purification of recombinant enzymes

Recombinant enzymes from T. cruzi WT (TcTIM WT), and single Cys TcTIM mutants with alanine (Ala) substitutions (C15A, C40A, C118A, and C127A) were overexpressed and purified. Additionally, recombinant TIM from Homo sapiens (HsTIM) was purified as previously reported⁴⁶. DNAs that code for the recombinant TIMs, except TcTIM C15A mutant, were cloned into the pET-HisTEVP plasmid and expressed in either Escherichia coli BL21(DE3)-pLysS (TcTIM WT and TcTIM Cys mutants) or in E. coli BL21 Codon Plus (DE3)-RIL strain (HsTIM). The gene encoding TcTIM C15A mutant was cloned into the pET-3a expression vector. Transforming bacteria with plasmids for TcTIM WT, TcTIM Cys mutants, and HsTIM were grown separately in LB medium supplemented with 0.1 mg/mL of ampicillin and 0.050 mg/mL of chloramphenicol for Codon plus strain, and incubated at 37 °C, until the cultures reached A_600nm = 0.8, thereafter were induced with 0.4-mM IPTG and incubated overnight at 30 °C with shaking at 180 rpm. After the bacteria were harvested by centrifugation (4010 × g, 15 min) and suspended in 40 ml of lysis buffer (pH 8.0) containing 50 mM Tris, 50 mM NaCl, and 0.2 mM PMSF. The bacterial suspension was disrupted using sonication cycles and clarified by centrifugation at 7690 × g for 1 h. at 4 °C. All the recombinant enzymes except TcTIM C15A were purified by IMAC as previously reported^45,46, using a Profinity Ni²⁺ charged resin that was previously equilibrated with lysis buffer. The soluble protein fraction was mixed with the equilibrated charged resin and incubated at 4 °C with shaking for 30 min. The sample was washed with the same buffer to remove undesired proteins. The TIMs were eluted with lysis buffer containing 200 mM imidazole adjusted to pH 8.0 and concentrated to a minimal volume using Amicon ultrafiltration units. The His-tag sequence was removed from the TIMs by incubation with TEVp7M⁴⁷ at a molar ratio of 30:1 (TIM/TEVp) at room temperature for 17 h, in buffer containing 50 mM Tris pH 8.0, 0.5 mM ethylene diamine tetraacetic acid, and 1 mM dithiothreitol (DTT). On the other hand, TcTIM C15A was purified by ion exchange and hydrophobic interaction chromatography. The supernatant was applied to a fast-flow SP-Sepharose column previously equilibrated with five-column volumes of lysis buffer. The protein was eluted with a 0–500 mM NaCl gradient in the same buffer, the eluted protein was precipitated with ammonium sulphate to a final concentration of 2.2 M and was applied to a hydrophobic Toyopearl column which had been previously equilibrated with 100 mM TEA, 1 mM EDTA, pH 7.4 and 2.2 M ammonium sulphate. The TcTIM C15A mutant was eluted with a linear gradient of ammonium sulphate of 2.2 M–0 M. All recombinant TIMs were loaded in sodium dodecyl sulphate-polyacrylamide gel electrophoresis (16% SDS–PAGE) and revealed with colloidal Coomassie Brilliant Blue. Protein concentration was spectrophotometrically measured at 280 nm using the molar extinction coefficient, 36,440 M⁻¹cm⁻¹ for WT TcTIM and the TcTIM Cys mutants, and 33,460 M⁻¹cm⁻¹ for HsTIM. Finally, the enzymes were precipitated with 75% ammonium sulphate and stored at 4 °C until use.

Inactivation of TcTIM WT, TcTIM Cys mutants, and HsTIM by Rbz

Assays were performed to investigate the inactivation of TcTIM WT, TcTIM Cys mutants, and HsTIM with Rbz. Enzyme samples at 0.2 mg/mL (7.2 nM) in TE buffer were incubated with increasing concentrations of Rbz (ranging from 25 to 150 µM) for 2 h at 37 °C. After the incubation period, aliquots were withdrawn and assayed for enzyme activity with a standard reaction mixture as previously reported⁴². The results were expressed as the percentage of enzyme activity, taking as 100% the activity of each enzyme incubated without Rbz. HsTIM was also assayed using the above-mentioned conditions.

Secondary structure determination and thermal assays

To estimate the secondary structure of the enzyme, we employed circular dichroism as follows. The TcTIM WT at 2 mg/mL was incubated with or without Rbz at the maximal inactivation concentration (150 μM) for 2 h at 37 °C, and enzyme activity was measured. To remove Rbz excess centricon tubes (30-kDa cut-off, Millipore) were used. Immediately after, TcTIM WT treated or not with Rbz were diluted to 0.2 mg/mL with 25-mM phosphate buffer (pH 7.4), and far-UV circular dichroism spectra (190–240 nm) were performed with a J-810 spectropolarimeter (Jasco, Easton, Maryland, USA) in a 0.1-cm quartz cell at 25 °C. The background buffer spectrum was obtained under the same experimental conditions (without protein) and was subtracted from the obtained spectrum of the protein. Additionally, the same samples were used to estimate the thermal unfolding (Tm) of the enzyme as previously described⁴⁶.

Quantification of derivatized Cys in TcTIM and its mutants

One of the reported mechanisms of Rbz is the Cys derivatization (chemical modification), so we decided to quantify the derivatized Cys in TcTIM WT and TcTIM Cys mutants not treated and treated with Rbz. For the assays, was used 2 mg/mL of fresh protein previously incubated with or without 150 μM Rbz for 2 h at 37 °C. After the excess of the Rbz was removed by centrifugation-filtration procedure as follows, a gel filtration spin column was loaded with Sephadex G-25 Fine resin, subsequently, the protein concentration was recalculated at 280 nm. Modified Cys were estimated spectrophotometrically in the filtered samples with 1 mM DTNB and 5% SDS dissolved in TE buffer. For the calculation of derivatized Cys, the absorbance of the samples was recorded at 412 nm (ε412 nm = 13.6 mM⁻¹cm⁻¹) as reported⁴⁸. Additionally, the derivatized Cys with DTNB were determined for TcTIM WT in real-time as follows, the enzyme (150 μg) was incubated at 25 °C in 1 ml of a buffer containing 100 mM TEA, 10 mM EDTA and 1 mM DTNB pH 7.4 for 25 min. 5 SDS % was added to completely denature the enzyme, and the absorbance at 412 nm was recorded immediately after adding the enzyme. The number of derivatized Cys under native conditions was determined at 11 min.

Molecular docking

For molecular docking studies, the crystallographic structures of TcTIM WT (1tcd) and TcTIM derivatized with DTBA (2oma) were downloaded from the Protein Data Bank (PDB). The 3D atomic coordinates of both structures were prepared by removing all water molecules and heteroatoms with PyMOL version 2.0.7 (Schrödinger Inc, NY, USA). The Chimaera software⁴⁹ was used to minimise the energy of the crystallographic coordinates, and the resultant structures were used for docking calculations. The 3D structure of the Rbz was obtained from the ZINC database (https://zinc.docking.org/) and energy was minimised with Avogadro version 1.2. The structures of both enzymes were prepared by adding hydrogen atoms and Kollman charges (6.00 and 3.999, respectively) with AutoDock Tools (ADT) version 1.5.6⁵⁰. Molecular docking was carried out with the free distribution software Docking Vina version 1.2.0⁵¹ and “Achilles” Blind Docking Server, available at: http://bio-hpc.eu/software/blind-docking-server/ using the default settings. Output files were saved in pdbqt format. After docking, close interactions for binding of the target with the ligand Rbz were analysed and visualised using the PyMol Molecular Graphics System software (version 2.5.0, Schrödinger, LLC, New York, NY, USA).

Results

Rabeprazole is a thiol drug that affects the viability of epimastigotes and inhibits the enzymatic function of cellular TIM

Initial studies in our research group had shown that rabeprazole was the best PPI that affects the activity of triosephosphate from T. cruzi (TcTIM). Furthermore, we observed that the inhibitory effect of this PPI on TcTIM is similar in acidic or neutral conditions (Supplementary Figure S1). Therefore, we chose this drug as a potential candidate to evaluate its effect on the viability of epimastigotes cultures and on the activity of recombinant TcTIM.

In a first attempt to determine whether the drug could induce the death of trypanosomes, we challenged epimastigotes of the CL Brener strain with increasing concentrations of Rbz ranging from zero to 2 mM, using a standard number of parasites as described in the Experimental section.

After incubation for 4 h at 28 °C, we assessed the viability of the epimastigotes. As shown in Figure 1(A), cell viability gradually decreased, as the concentration of Rbz increased, reaching marginal viability of 40%. In parallel, we took aliquots of the protein extracts of the epimastigotes exposed to the drug and determined the enzymatic activity of the cellular TIM. Our results show that at the lowest drug concentration (100 μM), the enzymatic activity of this glycolytic enzyme decreased by nearly 50%, and as the concentration of the reactive thiol was increased, its inhibitory effect intensified, reaching approximately 81% inhibition (Figure 1(B)).

Figure 1. Percentage of viability of T. cruzi epimastigotes and cellular TIM activity. 1 × 10⁶ epimastigotes were incubated for 4 h at 28 °C with increasing concentrations of Rbz. Viability was subsequently determined using MTT (A), and the cellular activity of the glycolytic enzyme TIM, was also measured (B). Results represent the average of four independent biological experiments.

These results demonstrate that Rbz gradually and effectively affects the viability of epimastigotes and this process is directly related to the inactivation of the TIM enzyme in this parasitic organism. To further investigate if the drug’s effect is independent of the number of parasites used, we evaluated cell viability and enzymatic activity using a different number of cells, at a fixed drug concentration (250 µM) and an incubation time of 24 h.

As shown in Figure 2, Rbz reduces viability by approximately 40% regardless of the number of parasites used, while TIM activity decreases drastically as a function of the number of parasites. Specifically, the enzymatic activity was inhibited by 93% when using a condition of 1 × 10⁵ parasites, suggesting that the enzyme is initially sensitised, leading to gradual parasite death.

Figure 2. Cell viability and enzymatic activity of TIM from T. cruzi epimastigotes challenged with Rbz. Different number of T. cruzi epimastigotes were incubated with 250 µM Rbz for 24 h at 28 °C. The plots show the percentage of cell viability determined using MTT (A) and cellular TIM activity of protein extracts from trypanosomes (B). Results represent the average of four independent biological experiments.

All the previous results demonstrate the effectiveness of the thiol drug in killing T. cruzi epimastigotes. Regardless of whether the concentration of the drug is increased, or the number of parasites is decreased, cell viability is gradually compromised, with concomitant and drastic inhibition of cellular TIM activity.

To further demonstrate the effectiveness of this drug in a shorter time, experiments with a different number of parasites incubated with 1 mM of Rbz for 4 h were performed. (Figure 3). Similar to the results shown in Figure 2, we demonstrate here that with only 4 h of incubation of parasites with Rbz, cell viability dropped approximately 40% with 2 × 10⁶ epimastigotes, and, when fewer parasites were used, their viability was maintained at these levels (Figure 3(A)). Notably, the enzymatic activity of the TIM was strongly affected, decreasing drastically from 2 × 10⁶ parasites, falling by 70%. Additionally, as smaller numbers of epimastigotes were used, the inhibition of the enzymatic activity reached 94% with 1 × 10⁵ parasites (Figure 3(B)).

Figure 3. Cell viability and enzymatic activity of cellular TIM from T. cruzi epimastigotes challenged with Rbz. Different number of T. cruzi epimastigotes were incubated with 1 mM Rbz for 4 h at 28 °C. The plots show the percentage of cell viability determined using MTT (A) and the cellular TIM activity of protein extracts from trypanosomes (B). Results represent the average of four independent biological experiments.

These data show that the reactive thiol drug has a great inhibitory effect on the proliferation of T. cruzi parasites.

Therefore, we determined the half-maximal inhibitory concentration (IC₅₀), which is the concentration of Rbz needed to inhibit T. cruzi proliferation by 50%, and compared this value with those obtained for Bzn as a reference drug. According to our results (Supplementary Figure S2), the IC₅₀ was calculated for Bzn of 5.86 µM ± 1.37, whereas the treatment with Rbz resulted in an IC₅₀ of 0.4 µM ± 0.017. These data demonstrate that, under our experimental conditions, Rbz was 14.6 times more effective in eliminating epimastigotes than Bzn.

Inhibition of the cellular TIM activity of T. cruzi epimastigotes generates an accumulation of MGO and AGEs

Given that our results demonstrate that the cellular TIM activity is gradually lost as the concentration of Rbz increases, we quantified the concentration of the highly toxic metabolite MGO. Mounting evidence has shown that this reactive metabolite is produced by the spontaneous degradation of the TIM substrates, GAP and DHAP, due to substrate accumulation because of TIM deficiency and/or loss of its enzymatic activity^52,53, and consequently, significant damage at the cellular level can be generated. Therefore, we decided to measure the levels of MGO in epimastigotes under the treatment regime. The results show that, as the concentration of the thiol drug Rbz increased, the concentration of MGO also gradually increased too (Figure 4(A)). In fact, an almost linear dose-dependent response was observed, with the concentration increasing three-fold higher than the control experiment from the lowest drug concentration (100 µM), until reaching almost a 12-fold increase with 2 mM of Rbz. This confirms the accumulation of MGO as a direct consequence of cellular TIM inactivation.

Figure 4. Quantification of MGO and AGEs in T. cruzi epimastigotes treated with Rbz. 1X10⁶ epimastigotes were incubated with increasing concentrations of Rbz for 4 h at 28 °C. (A) Concentration of MGO/million cells at increasing concentrations of Rbz. (B) Concentration of AGEs in µg/mL per million cells at increasing concentrations of Rbz. Both metabolites were determined from protein extracts of epimastigotes incubated with different concentrations of Rbz. Results represent the average of four independent biological experiments.

On the other hand, since the accumulation of MGO triggers a series of cellular events such as the production of AGEs, we decided to evaluate the presence of such agents in T. cruzi under drug treatment. It is widely documented that AGEs are the product of the reaction between MGO and some aminoacyl residues of proteins, as well as with nucleic acids, among other biomolecules⁵⁴.

Thus, after incubating the parasites with increasing concentrations of Rbz, AGEs were measured as described in Material and Methods. In agreement with the results obtained with MGO, Figure 4(B), shows that, as the concentration of the drug increased, the AGEs level gradually increased too. In fact, a significant increase of 17 times of the AGEs concentration was observed at the lowest concentration of Rbz (100 µM) compared to the control experiment (in the absence of Rbz). As Rbz concentration was increased, a considerable concentration of AGEs was determined, reaching concentrations around 4 µg/mL (Figure 4(B)).

This indicates that an important consequence of enzymatic inhibition of cellular TIM is the accumulation of MGO, which leads to the generation of AGEs that are directly related to cell damage.

On the other hand, it is well known that proteolytic events in T. cruzi play an important role in the adaptation of this organism to different host environments and to the different conditions to which it is exposed during its complex life cycle⁵⁵. This, coupled with the dynamism of the glycosome, which can undergo autophagy under different circumstances and be recycled along with its protein content⁵⁶, makes it probable that the alteration of the dynamism of organelles or proteins may occurs upon exposing the parasites to the drug Rbz, with consequent degradation of those organelles and proteins, which include glycosomal TIM. Therefore, to rule out the possibility that the abolition of TIM activity, as demonstrated under pharmacological treatment, was a consequence of extensive proteolysis of the glycolytic enzyme, we analysed the expression of cellular TIM in protein extracts from Rbz-treated epimastigote cultures by western blot experiments using polyclonal antibodies against recombinant TcTIM.

As shown in Supplementary Figure, S3A, the intensity of the band corresponding to the control group (without pharmacological treatment) had a relative value of 0.348 (TIM/β-actin), while in the other conditions with pharmacological treatment, the band intensity of cellular TIM increased significantly, by approximately 48%, compared to the control condition (0.348 vs 0.551), as demonstrated in the densitometric analysis (Supplementary Figure, S3B).

At all concentrations of the drug that were used with 1 × 10⁶ epimastigotes, a slight increase in the band intensity corresponding to TcTIM was observed (compared to the control group), while the band corresponding to β-actin, a constitutive and stable protein, remained constant at all concentrations. These results show that the intensity of the cellular TIM band did not decrease in each experimental condition. On the contrary, an increase in the intensity of the band was observed with the pharmacological treatment. Thus, when taken together with the inhibition of TIM activity (Figures 1(B), 2(B), and 3(B)), the aforementioned enzymatic inhibition is shown to be directly promoted by Rbz treatment.

Recombinant TcTIM is selectively inactivated by the drug Rbz, and the mechanism of action is through the chemical modification of its cysteine residues

Previous studies have shown that compounds such as MMTS or DTNB, which chemically and selectively modify cysteine residues can inactivate recombinant TcTIM⁴². Given that this enzyme contains 4 cysteine residues (Cys) in its primary sequence, we tested the effect of the thiol drug Rbz on the enzymatic activity of recombinant TcTIM. In parallel and as a measure of comparison, the same assays were performed on recombinant human TIM (HsTIM), since this enzyme also contains 5 Cys in its primary sequence (NCBI Reference Sequence: NP_000356.1) and could potentially be susceptible to inactivation by thiol compounds. After incubating the recombinant enzymes with increasing concentrations of the drug for 2 h at 37 °C, an aliquot was taken to determine the enzymatic activity. The results demonstrate that a concentration of 75 μM Rbz was sufficient to inactivate approximately 95% of TcTIM (as shown in Figure 5). In contrast, HsTIM maintained 100% of its activity at all tested concentrations (also shown in Figure 5). These findings indicate that TcTIM is selectively inhibited by this reactive thiol drug, while the human glycolytic enzyme maintains normal enzymatic activity in the presence of the drug, despite having cysteine residues.

Figure 5. Effect of Rbz on the enzymatic activity of TcTIM and HsTIM. 0.2 mg/mL of the recombinant enzymes were exposed to increasing concentrations of Rbz for 2 h at 37 °C. Subsequently, an aliquot was taken, and the enzyme activity was measured in a coupled assay. Red squares correspond to TcTIM and green circles to HsTIM. Results represent the average of four independent biological experiments.

As a consequence of the functional inhibition, changes in the structure of the recombinant TcTIM may occur. Therefore, the content of the secondary structure and stability to temperature were determined by circular dichroism assays. The recombinant enzyme was exposed to 150 μM Rbz for 2 h at 37 °C. As shown in Figure 6(A), a canonical circular dichroism spectrum of the TIMs was evidenced in the TcTIM control (enzyme incubated in the absence of the drug), with two characteristic minimum values close to 222 and 208 nm. In contrast, the enzyme with pharmacological treatment shows a decrease in the content of α-helices and β-sheets, with a difference close to 50%, indicating a considerable disturbance of the secondary structure. Additionally, the thermostability in a range of 20–90 °C was measured by circular dichroism. As shown in Figure 6(B), the control enzyme (in the absence of a drug) had a Tm value of 57.1 °C, while the Tm value of the enzyme previously treated with Rbz was 41.5 °C, causing a 15 °C decrease in thermostability. This indicates that Rbz produces strong structural alterations in TcTIM.

Figure 6. Circular Dichroism spectra of TcTIM. Protein was prepared for the experimental conditions, in (A) green squares represent the enzyme incubated in absence of Rbz, whereas red circles represent the enzyme incubated in presence of Rbz. The Tm was also determined for these samples. In (B), open squares are the enzyme without Rbz (control) and open circles are the enzyme incubated in presence of Rbz. The red line corresponds to the Boltzmann sigmoid equation adjusted to the experimental data. Results represent the average of four independent biological experiments.

Otherwise, since it has been demonstrated that the main mechanism of action of Rbz is via the chemical modification of Cys, we determined the Cys content of treated and untreated TcTIM using Ellman’s reagent. To do this, we incubated 100 µg of TcTIM (in native conditions) in the presence of 1 mM DTNB and recorded the absorbance at 412 nm. Supplementary Figure, S4 shows that the absorbance increased as time passed, indicating that 2.6 Cys/subunit were quantified after approximately 11 min. Subsequently, the enzyme began to form a precipitate, causing the sample to become opaque. Therefore, after 15 min of incubation, we added an aliquot of the denaturing agent to finish titrating the Cys. These results demonstrate that more than two Cys residues in TcTIM are accessible to small molecules in short time intervals. These findings prompted us to quantify derivatized Cys by Rbz in the enzyme. To do this, we incubated the enzyme at 2 mg/mL in the absence or in the presence of 150 μM Rbz (total inactivation condition), for 2 h at 37 °C. At the end of the assay, to eliminate the excess drug, we ultrafiltered the samples using centricon tubes (10 kDa cut-off) and recalculated the protein concentration spectrophotometrically at 280 nm. We then determined the Cys content using DTNB under denaturing conditions. As shown in Table 1, in the control enzyme (in the absence of the drug) we quantified approximately 4 Cys/subunit, which is consistent with the Cys content of the primary sequence (4 Cys at positions 15, 40, 118, and 127; NCBI Reference Sequence: XP_818253.1). In contrast, when the enzyme was incubated with the reactive thiol, we determined that approximately 3 Cys/subunit are derivatized with Rbz. These findings indicate that 1–3 derivatized Cys are required to inactivate TcTIM.

Table 1. Determination of the Cys content in the TcTIM in the absence and in presence of Rbz.

TcTIM WT	Quantified Cys/subunit	Derivatized Cys/subunit^a
Control	3.98 ± 0.31	–
+ Rbz	1.04 ± 0.23	≈ 3

^aCys derivatized with Rbz, this result corresponds to the subtraction of the control and the Rbz condition.

The preceding results show that Rbz can derivatize several Cys and can differentially contribute to inactivating TcTIM. Therefore, we performed tests on single Cys mutants of TcTIM, substituting each Cys for alanine (Ala), a relatively silent mutation, to evaluate the contribution of each Cys to the inactivation of the glycolytic enzyme. After overexpressing and purifying the enzymes, we exposed them to increasing concentrations of Rbz, incubating for 2 h at 37 °C. At the end of the assays, we took an aliquot and determined the enzymatic activity (Figure 7). The results showed that all the single mutants experienced an inactivation effect. The C127A and C118A mutants showed the highest sensitivity to the effect of Rbz, followed by the WT, the C40A, and C15A mutants (Figure 7). However, almost total inactivation was promoted in all enzymes. Therefore, our results show that the inactivating effect of Rbz on TcTIM involves the combination of several Cys residues, which contribute to the high sensitivity of the enzyme towards the drug.

Figure 7. Effect of Rbz on the enzymatic activity of TcTIM WT and single mutants of Cys. All recombinant enzymes were incubated at 0.2 mg/mL and were exposed to increasing concentrations of the drug Rbz for 2 h at 37 °C. At the end of the incubation time, an aliquot was taken, and enzyme activity was measured by the coupled assay system as reported in the Material and Methods section. Black and orange squares correspond to WT and the C15A mutant, respectively; green and red circles correspond to the C118A and C40A, respectively; and the purple triangles correspond to the C127A mutant. Results represent the average of four independent biological experiments.

These results were reinforced by quantifying the derivatized Cys in the recombinant mutant enzymes in the absence or presence of Rbz (Table 2). As observed, approximately 3 Cys/subunit were quantified in the absence of drug treatment in these mutants. In almost all of the mutants, nearly 2 Cys/subunit were derivatized with Rbz, except for C127A, where nearly 3 derivatized Cys/subunit were quantified. These data demonstrate that Cys 15, 40, and 118 are involved in the inactivation, while Cys 127 does not participate in the inactivation of this enzyme.

Table 2. Determination of the Cys content in TcTIM WT and single mutants of Cys in the absence and presence of Rbz.

TcTIM	Quantified Cys/subunit	Derivatized Cys/subunit^a
C15A	2.91 ± 0.27	–
+ Rbz	0.97 ± 0.21	≈ 2
C40A	3.12 ± 0.13	–
+ Rbz	1.14 ± 0.18	≈ 2
C118A	2.94 ± 0.32	–
+ Rbz	1.02 ± 0.22	≈ 2
C127A	3.12 ± 0.25	–
+ Rbz	0 ± 0.07	≈ 3

^aCys derivatized with Rbz, this result corresponds to the subtraction of the control and the Rbz-treated enzymes.

Molecular docking simulations of Rbz against TcTIM reveal accessible sites for the drug

To identify the probable molecular interactions of Rbz with the Cys of TcTIM, we used the free distribution software Docking Vina⁵¹, and the Achilles Blind Docking Server, https://bio-hpc.ucam.edu/achilles/ for molecular docking. We used the 3D structure of Rbz from the ZINC database and the crystallographic structure of TcTIM (PDB code: 1tcd). In the first approach, using blind molecular docking on the total surface of the protein, we observed that Rbz can interact with several zones on the TcTIM dimer surface (Figure 8(A)). Interestingly, the cavities at the enzyme interface, (the contact area between the two subunits) had the highest docking score (Figure 8(A,E), respectively). Furthermore, molecular docking demonstrated that Rbz was docked in cavities near Cys 15, 40, and 118 of the enzyme structure (Figure 8(B–D), respectively), and good molecular docking scores for the interface and the Cys were obtained, as shown in Figure 8(E).

Figure 8. Molecular docking of Rbz on TcTIM. The crystallographic structure of TcTIM (PDB code: 1tcd) and the 3D coordinates of Rbz were used for molecular docking using Docking Vina and the Blind Docking Server, as mentioned in the body of the text. (A) Blind docking of Rbz on the 3D structure of TcTIM. (B–D), Docking of Rbz at the cavities close to Cys 15, 40, and 118, respectively. (E) The scores obtained for the docking of Rbz at the interface and the cavities close to Cys 15, 40, and 118. The figures were modelled with PyMol Molecular Graphics System software (version 2.5.0, Schrödinger, LLC, New York, NY, USA).

Finally, we conducted molecular docking with the crystallographic structure of the TcTIM (PDB code: 2oma)²². It is important to note that this 3D structure was solved in complex with the thiol-reactive compound DTBA, and it was established that Cys 118 is derivatized with DTBA, causing structural alteration of the interface. Our molecular docking results of Rbz vs the interfacial Cys15 showed better molecular docking between Rbz and cavities close to that Cys, compared to the structure of the non-derivatized TcTIM (Figure 9(A,B)). It is important to highlight that in this 3D structure, the interdigitating loop 3 (aminoacyl residues 69–80) is displaced by 7.5 Å, compared to loop 3 of the non-derivatized structure (PDB code: 1tcd) (Figure 9(C)), demonstrating that, when Cys 118 is derivatized, the interdigitating loop 3 is significantly disturbed.

Figure 9. Cartoon and stick representations of molecular docking in Rbz vs TcTIM. The crystallographic structures of (A) non-derivatized TcTIM (1tcd), and (B) derivatized-TcTIM with DTBA (2oma) were used with the 3D coordinates of the ligand Rbz; green and cyan colours represent each subunit of the dimeric protein. (C) The crystallographic structures of 1tcd (purple) and 2oma (gray) were superimposed obtaining an overall RMSD of 0.366 Å for C-α. Despite significant structural coincidence, loop 3 of 2oma is displaced 7.5 Å with respect to 1tcd, (zoom view in panel (C). The figures were modelled with PyMol Molecular Graphics System software (version 2.5.0, Schrödinger, LLC, New York, NY, USA).

This claim is supported by the fact that, in the 3D structure of the TcTIM complex with DTBA, the B factors, which measure relative vibration in crystallographic structures, of Cys15 and adjacent amino acids in the B subunit are abnormally and considerably higher (Supplementary Figure S5). This condition could lead to high reactivity of the interfacial Cys, increasing the accessibility of thiol-reactive compounds or other small molecules. Collectively, the functional and structural results from the recombinant enzyme, demonstrate that the Cys residues of TcTIM may be derivatized by Rbz, leading to strong structural alterations and the total inactivation of the enzyme.

Discussion

American trypanosomiasis is considered a global public health problem, that can cause severe complications, even leading to death. The available pharmacological treatments, Bzn, and nifurtimox, were developed over 50 years ago and, while they are relatively efficient in treating the chronic stage of the disease, the presence of resistant strains of T. cruzi and the severe adverse effects associated with these commercial drugs emphasises the need for the development of new alternatives for the treatment of CD^12,57,58. Furthermore, recent reports indicate that the global dissemination of this parasite has increased, as evidenced by the rise in CD cases in non-endemic countries such as North America, Europe, and Australia⁵⁹.

Reports of congenital transmission have also been increasing, to the point that it has become the main mechanism for disease spread in regions where vector control programs have been successfully established⁶⁰. Therefore, it is necessary to explore novel and safe alternatives to treatment for CD, taking advantage of the benefits of drug repurposing.

PPIs have been used in several drug repositioning approaches and different results have been obtained with respect to the best drug depending on the study model and the target protein in question. For example, Suzuki et al. demonstrated that the co-administration of the PPIs delays the elimination of plasma methotrexate in breast cancer⁶¹; in another work, using diverse cancer cell lines (human melanoma, and osteosarcoma) it has been demonstrated that lansoprazole was the best PPI that promoted apoptosis and cell death of cancer cells⁶² and recently, Rbz was probed as a suitable anti-cancer drug in breast cancer cells, demonstrating its efficacy to inhibit cell proliferation and to promote apoptosis³⁴. On the other hand, there are some reports of repurposing omeprazole and Rbz as antiparasitic drugs since 2014^36,37,63, and in a recent study lansoprazole and posaconazole were proposed as anti-leishmanial agents⁶⁴.

In a first attempt to evaluate the best PPI as an anti-trypanosomatid drug, we performed assays with all PPIs at increasing concentrations in the recombinant TcTIM, demonstrating that Rbz inactivated efficiently this enzyme, this effect of Rbz was similar at neutral and acidic conditions (Supplementary Figure S1). Our results contrast with other reports that argue that PPIs must be activated by acidic pH to achieve their effectiveness. Some examples of this include the report where it was shown that esomeprazole inhibited the proliferation of melanoma cells in vitro and induced cytotoxicity in a pH-dependent manner, obtaining the most potent anticancer effect at pH 6.0⁶⁵; and the work of Li et al., which demonstrated that the synergistic effect of pantoprazole and vitamin C was pH-dependent due to pantoprazole was more effective at a slightly acidic pH⁶⁶. However, there are reports of the action of these drugs in conditions that commonly show neutral pH. For example, it has been demonstrated that omeprazole inhibits cell proliferation in Barrett’s oesophagus cells at neutral pH conditions⁶⁷; in another report, it has been shown that omeprazole inhibits melanogenesis by blocking ATP7A trafficking in the same pH conditions⁶⁸. Despite the reports that show sulfenamide formation in different pH values are few, Kromer since 1998⁶⁹ demonstrated that PPIs can be activated at different pH values with a differential rate. Therefore, it is reasonable that the PPIs can react with thiols of Cys residues in proteins at neutral pH. Taking into consideration the aforementioned, we evaluated the effect of Rbz at pH 7.4 based on its efficacy in killing T. cruzi epimastigotes. Our study points to the fact that this drug inhibits the activity of cellular triosephosphate isomerase of this parasite, suggesting it as a therapeutic target.

Triosephosphate isomerase has been postulated as an excellent anti-trypanosomatid target for more than two decades, and several studies have demonstrated the inhibition of the recombinant enzyme; however, none of them have shown the effect on the native enzyme inside the parasite. In that sense, we demonstrated cellular TIM inhibition using Rbz in a drug-repurposing approach. PPIs, including Rbz, are drugs widely used for the treatment of acid-related disease, particularly gastroesophageal reflux disease (GERD), with a few adverse effects reported to date⁷⁰. Their mechanism of action depends directly on their chemical structure and involves the formation of covalent disulphide bonds with Cys of the ATPase. This mechanism has been useful in exploring other uses of these drugs and proposing them as antiparasitic drugs^36,37,71,72.

Several research groups have proposed many molecules of different chemical natures with potential effects against T. cruzi⁷³, including enzymes of the glycolytic pathway. However, to date, the use of Rbz, a PPI widely used worldwide, has not been proposed. PPIs have been proven safe and effective for more than two decades, making it feasible to propose them as a treatment for CD.

Our results show that Rbz is very effective in killing T. cruzi epimastigotes from the CL Brener strain, with an IC₅₀ 15 times lower than that calculated for the reference drug Bzn, a result that is consistent with previous reports⁷⁴.

Comparatively, the IC₅₀ of compounds such as those derived from N-acyl hydrazone against epimastigotes were estimated in the micromolar range (35–250 μM)⁷⁵. Even for some other repurposed drugs such as ciprofloxacin, folic acid, and naproxen, higher IC₅₀ values, around 400 μM, have been obtained⁷⁶. Other studies have reported IC₅₀ values similar to those obtained for Rbz. For instance, IC₅₀ values in the sub-micromolar order (0.8 µM) were observed for bisphosphonate derivatives, indicating that the glycolytic enzyme hexokinase was one of the main targets of these compounds in T. cruzi epimastigotes⁷⁷.

In contrast, Bzn has demonstrated excellent IC₅₀ values. However, several reports indicate that exposure of epimastigotes to this drug for prolonged periods can induce resistance, resulting in increased IC₅₀ values⁷⁸. This suggests that prolonged use of Bzn can generate significant resistance in trypanosomes. Therefore, the development of new drugs, such as Rbz, appears to be promising anti-Chagas therapy. This is supported by studies showing that omeprazole, another PPI, is an effective drug against Trypanosoma brucei and Leishmania donovani, other trypanosomatids causing sleeping sickness and visceral leishmaniasis, respectively^39,79.

We have demonstrated that Rbz targets the glycolytic enzyme TIM in T. cruzi epimastigotes, leading to almost complete inhibition of enzymatic activity. Even though this organism has the glycosomal enzyme that degrades DHAP to glycerol 3-phosphate (NADH-dependent glycerol-3-phosphate dehydrogenase)²³ the accumulation of DHAP occurs due to the interruption of TIM activity. This is highly relevant since in conditions of deficiency or deletion of TIM, an excessive accumulation of DHAP occurs^52,80 which causes its spontaneous degradation and forms the highly toxic metabolite methylglyoxal (MGO)⁸¹. At low concentrations, MGO can help cell growth, but at high concentrations, it can be toxic and causes cell death⁸². Although T. cruzi has detoxification mechanisms such as the glyoxalase system (Glyoxalases 1 and 2), we found that the intracellular MGO levels increased up to almost 10 times the control level when treated with Rbz at the maximum concentration (Figure 4). This suggests that the glyoxalase system and other detoxifying enzymes failed to remove this toxic metabolite. In this regard, another study showed that the exogenous addition of MGO to T. cruzi epimastigotes caused cell death in these organisms, with an EC₅₀ of 171 μM⁸³. Interestingly, our study showed that MGO levels of up to 0.37 μM per million epimastigotes were obtained (Figure 4), which is significantly lower than that reported by Greig in 2009. However, these levels were generated intracellularly due to treatment with Rbz with concomitant inactivation of cellular TIM, in contrast to the exogenous MGO in the other study.

MGO has a significant capacity to generate AGEs in proteins, which in turn affects protein function, leading to exhaustive cellular toxicity⁸⁴, therefore compromising cell integrity. We demonstrated an increase in AGEs levels of more than 50 times, compared to the condition without treatment (Figure 4). Unfortunately, to date, no studies have been reported on the production of AGEs in T. cruzi and there are even no studies on this subject with other trypanosomes. However, it was shown that omeprazole succeeds in inhibiting the activity of TIM in the intestinal parasite Giardia lamblia, which concomitantly led to a significative increase in AGE levels, by almost 12 times with respect to the condition without the drug, consequently causing the death of the parasite⁴⁴. Additionally, it has been shown that the accumulation of dihydroxyacetone can become toxic in the bloodstream forms of T. brucei. It was shown that it can have an anti-proliferative effect because of the non-enzymatic glycation named Maillard reaction, which is generated when AGEs are formed. Therefore, the authors suggested that this glycation phenomenon could contribute to the death of T. brucei by autophagy⁸⁵. The foregoing evidence demonstrates that this type of damage to biomolecules (e.g. generation of AGEs) in parasite proteins can induce very negative effects on viability, as we demonstrated in this study. Moreover, toxicity mediated by highly reactive metabolites has also been reported as a mechanism of action of Bzn, one of the drugs used for the treatment of CD. In this context, Bzn must be activated by a type 1 nitroreductase (NTR) that catalyses the reduction of the nitroimidazole motif to hydroxylamine, which is subsequently converted to the highly reactive metabolite dialdehyde glyoxal, capable of forming adducts with proteins, DNA/RNA and small molecules such as glutathione⁷. Therefore, the mechanism of action of Bzn and Rbz share some similarities, with the advantage that no severe adverse effects have been reported for Rbz.

On the other hand, trypanosomes can undergo abrupt changes such as the passage from one host to another, and even very adverse conditions in pharmacological treatments. Besides, it has been shown that their glycosomes can be renewed through a mechanism called pexophagy, which can involve the generation of new organelles together with the protein composition that characterises them⁸⁶. For this reason, we evaluated the protein expression of cellular TIM from Rbz-treated and untreated epimastigotes by western blot. Our data allow us to demonstrate that the enzyme maintained normal protein levels and was even slightly up-regulated when the parasite was exposed to the drug (Supplementary Figure S3). Currently, there is no clear evidence of how the protein content of the glycosome is altered (especially the glycolytic enzymes) under conditions of pharmacological treatment. However, due to the stress generated under such conditions, it is suggestive that changes in autophagy occur. For example, one study reported in 2016, in which T. cruzi was exposed to the compound β-lapachone, led to the inhibition of the glycosomal enzymes glycerol kinase and glyceraldehyde-3-phosphate dehydrogenase, as well as autophagy and subsequent cell death⁸⁷, while in another study, it was demonstrated that in tumour cells treated with the PPI pantoprazole, autophagy was inhibited, which led to decreased cell viability⁸⁸.

Based on our findings, we propose that the expression of glycosomal TIM is not importantly affected and is even slightly up-regulated by treatment with Rbz, while its enzymatic activity is drastically inhibited.

Initial studies on recombinant TcTIM revealed its susceptibility to inhibition by chemical agents that derivatize Cys residues⁴² and subsequent research also confirmed efficient enzymatic inhibition with other types of non-derivatizing compounds. In our work, we demonstrate that the reactive thiol drug, Rbz completely abolishes TcTIM activity, while the human homologous enzyme retains its enzymatic activity under the same experimental conditions, despite having Cys in its primary sequence. Therefore, Rbz has species-specific selectivity, as shown in other studies demonstrating its selective inhibition of different TIMs of parasites, such as G. lamblia³⁷ and Encephalitozoon intestinalis⁷² without affecting the corresponding homologous human enzyme.

In addition to functional impairment, Rbz caused significant modifications at the structural level of TcTIM, as observed by circular dichroism studies. The main affectation of TcTIM by Rbz consisted in a decrease in the secondary structure signal at 222 nm, and the loss of thermostability reflected in a Tm value 15 degrees lower than the untreated enzyme, indicating that this drug drastically affects the native conformation of TcTIM. Conversely, changes at the structural level in Giardia lamblia TIM (GlTIM) treated with Rbz were more subtle, showing a Tm value 9 °C lower than that obtained for the untreated enzyme³⁷. Our results are according with previous studies that have shown that when using thiol reactive compounds, Cys derivatization, in recombinant TcTIM, caused the loss of catalytic activity and significant structural alterations. In another study, it was demonstrated that the thiol-reactive dithiodianiline altered TcTIM dimer stability by perturbation of the dimeric interface, leading to its dissociation and loss of enzymatic activity²². These findings demonstrate that chemical modification of these aminoacyl residues can lead to important structural alterations that may become irreversible. TcTIM WT contains four Cys per subunit, and our results show that Rbz derivatized three of them, which is directly related to the abolition of the enzymatic activity and the drastic structural effects observed in this work. The results obtained with the single Cys mutants strongly suggest that more than one Cys residue substantially contributes to abolishing the catalytic activity, and that each of them plays an important role in TcTIM inactivation. This contrasts with GlTIM, where Cys 222 is the main residue that contributes to enzyme inactivation³⁷. This result was consistent with the report by Olivares-Illana in 2007, in which they showed similar sensitivity to dithiodianiline of TcTIM WT and three single Cys mutants, excepting the C127A mutant, which was less sensitive to this compound and is similar to our results.

Therefore, it is to be expected that this mutant has greater sensitivity to the compounds, and it is very likely that Cys 127 does not contribute to the inhibition of TcTIM Rbz (Table 2), since the crystallographic structures (PDB codes: 1tcd and 2oma) indicate that is practically buried in the core of the enzyme with no substantial solvent accessibility. Thus, based on our data and those already mentioned, we suggest that Cys 15, 40, and 118 contribute globally to the inhibition of TcTIM by Rbz.

On the other hand, analysing the mechanisms involved in enzyme inhibition at the molecular level can be an important guide to generating better proposals for compounds with pharmacological potential. Therefore, computational simulations of molecular docking can give us an idea of the potential binding sites of the inhibitor and its possible conformers that could be found in the enzyme-inhibitor ligand complex. We identified, by molecular docking and using the crystallographic structure of TcTIM, that Rbz has an affinity for different cavities close to the Cys residues, and significantly, also in the dimeric interface. Crystallography studies of TcTIM soaked with the inhibitor DTBA²² showed that this compound was capable of derivatizing Cys 118, which caused alteration of the interdigitating loop 3 of one of the subunits, leading to a higher vibrational state and permeability of Cys 15 (Figure 9 and Supplementary Figure S5), which notably increases the possibility that this aminoacyl residue is derivatized. In another work involving molecular dynamics, the authors showed that the interface zone is highly susceptible to the coupling of aromatic ligands derived from benzothiazoles⁸⁹, similar to our results. Interestingly, this same situation does not occur in the HsTIM structure. Moreover, the authors mentioned that the inhibitors bind preferentially to the tunnel-shaped cavity that forms at the interface of the two subunits and highlighted the importance of this tunnel for TcTIM inhibition. We suggest that in the series of events promoted by Rbz, derivatization of Cys 118 could take place first by causing such a structural alteration in loop 3 that allows Rbz access to Cys 15 by perturbing the dimer interface, followed by derivatization of Cys 40, leading altogether to a generalised collapse that is reflected in both, abolition of catalysis and structural stability of the enzyme.

Collectively our data demonstrate the detrimental effect of Rbz on the viability of epimastigotes from T. cruzi, and TcTIM was identified as one of the targets susceptible to inhibition by this drug. The PPI Rbz globally induces alterations in TcTIM via derivatization of multiple Cys residues, compromising catalysis and structure. Thus, we suggest Rbz as a possible repurposing drug for the treatment of CD considering the minimal adverse effects reported until now with its use. Further studies must be done for considering its use as a safe option in the treatment of CD.

Author contributions

Conceptualization: SE-F. and I.G.-T.; Formal Analysis: SE-F., G.L.-V, I. G.-T.; Funding Acquisition: S.E-F., G.L.-V., L.A.F.-L.; Investigation: I.G.-T., SE-F. I.D.-D., G.L.-V, N.C., L.A.F.-L.; methodology: SE-F. and I.G.-T.; resources: R.P.-M., I.B., R. H., J.H.-L.; software: SE-F.; writing-original draft preparation: S. E.-F., I.G.-T. writing-review and editing: R. P.-M., S. E.-F., I.G.-T., G.L.-V. All authors have read and agreed to the published version of the manuscript.

Acknowledgements

We thank the technical assistance of Ing. Manuel Ortínez, Q.B.P. Miriam Vázquez Acevedo, and M.V.Z. Héctor Malagón Rivera from Instituto de Fisiología Celular, Universidad Nacional Autónoma de México.

Disclosure statement

The authors report no conflicts of interest.

Additional information

Funding

This work was supported by the Recursos Fiscales para Investigación Program from the Instituto Nacional de Pediatría, S.S. Under Grants [2019/062, 2019/072, and 2020/016].