Identification of new Trypanosoma cruzi-specific protein synthesis inhibitors by using a developed high-throughput in vitro translation assay

Abstract

There are restricted treatment options for Chagas disease, a global health risk causing 12,000 fatalities annually. This situation is exacerbated by environmental changes and migration, which is altering patterns of disease transmission, and by expansion of the vector range. The disease is caused by the protozoan parasite Trypanosoma cruzi, which possesses a divergent translation apparatus, characterized by enlarged ribosomes and a unique mRNA cap appended to the 5′ end of a 39-nucleotide spliced leader sequence at the 5’ end of all its mRNAs. Given that many antimicrobial agents are protein synthesis inhibitors and that there are significant differences between the T. cruzi protein synthesis machinery and that of its mammalian host, it is likely that T. cruzi-specific protein synthesis inhibitors could be employed for the treatment of Chagas disease. Such inhibitors can be identified by screening chemical libraries in an in vitro translation assay. However, lysates of T. cruzi cannot reinitiate translation of exogenous mRNAs. Here, we demonstrate that T. cruzi extracts derived from a strain deficient for hemin accumulation efficiently translated a capped and polyadenylated reporter mRNA. We optimized assay conditions and adapted them to a 384-well format. We further miniaturized this assay to a 5 μL volume, demonstrating its suitability for ultra-high throughput screening. Further, in a proof-of-principle pilot study, we identified quinazoline compounds that inhibit translation of reporter mRNAs by T. cruzi extracts. These compounds inhibited protein synthesis and proliferation of insect-derived forms and wild-type parasites growing in mammalian cells at a low micromolar range and a selectivity index higher than 200. This first-of-its-kind T. cruzi in vitro translation system will enable high-throughput screening of very large chemical libraries, while first-in-class hit compounds identified in the pilot study can be developed into lead compounds by design and synthesis of focused libraries to develop therapeutics for Chagas disease.

Graphical abstract

Highlights

• •Protein synthesis inhibitors are used to treat microbial infections.
• •We developed an assay to test potential protein synthesis inhibitors for Trypanosoma cruzi.
• •The assay consisted of parasite extracts to translate an artificial RNA for luciferase.
• •We identified some efficient chemicals that inhibit parasite translation.

1. Introduction

Chagas disease is a significant public health concern. Originally endemic to Latin America, it has been globally disseminated due to the emigration of infected individuals and the expanding range of insect hosts (DNDI, 2025; World Health, 2025). Chagas disease accounts for >12,000 fatalities annually, with around 100 million individuals at risk of infection. The disease is caused by the protozoan Trypanosoma cruzi, which is transmitted by blood-sucking insect vectors, through blood transfusion and organ transplants from infected patients, or by congenital contact (Coura, 2015; Shikanai Yasuda, 2022; Zingales and Bartholomeu, 2022). Furthermore, parasite-contaminated food and beverages may result in oral contamination (Shikanai Yasuda, 2022). T. cruzi multiplies in the midgut of the insect host as epimastigotes and transforms into nonproliferating but infective, metacyclic trypomastigotes, which are released in the feces or urine during the blood meal. They reach mucosal membranes of mammalian hosts and enter cells, forming a parasitophorous vacuole, then escape to the cell cytosol, where they proliferate in amastigote form (Martin-Escolano et al., 2022; Zuma, 2022). Amastigotes differentiate into trypomastigotes that are released after lysis of the infected cells, reaching the extracellular environment to disperse through different tissues and cell types. Upon human infection, the parasite causes various symptoms, including headache, body aches, and nausea (de Castro Neto et al., 2021; Maguire et al., 2021). Acute symptoms of the disease typically self-resolve due to a robust immune response. However, a persistent infection remains for decades, and in some cases heart or digestive tissues start to present alterations, most likely due to imbalances in the host's immune system leading to cardiac, gastrointestinal, and neurological complications, which are potentially fatal (dos Santos et al., 2012).

Two drugs used for the treatment of Chagas disease are benznidazole and nifurtimox (Hasslocher-Moreno et al., 2021; Kratz et al., 2022). They are prodrugs, requiring metabolization by the parasite to become active, and are primarily useful during the acute and chronic indeterminate phases of the disease (Wilkinson et al., 2008; Maguire et al., 2021). The treatment is complicated due to poor drug tolerability and side effects, not infrequently leading to treatment interruption (Maguire et al., 2021). During the chronic symptomatic phase, drugs are prescribed according to the patient's medical condition and the disease evolution, but chances of cure are low (Murta et al., 2008; Campos et al., 2017; Diaz-Viraque et al., 2018; Porta et al., 2023). There are currently no effective treatments for chronic T. cruzi infections, especially against persistent parasites (Barrett et al., 2019).

Inhibition of protein synthesis is the mechanism of action of numerous antimicrobial medicines employed in the clinic (Coyle and Society of Infectious Diseases, 2003; Hilal-Dandan and Brunton, 2016). Despite T. cruzi being a eukaryote, its translation machinery shows significant divergence from that found in mammals. T. cruzi genomes are organized in polycistronic clusters that are transcribed as one long pre-mRNA that undergoes trans-splicing to add a peculiar 39-nucleotide sequence named spliced leader (SL) to its 5’ end (Sather and Agabian, 1985; Zwierzynski and Buck, 1991; Liang et al., 2003). The SL contains a distinct 7mGTPpp cap with the next 4 nucleotides also methylated, generating a Cap4 structure, which is not present in mammalian mRNAs (Sather and Agabian, 1985; Zwierzynski and Buck, 1991; Liang et al., 2003). This structure is recognized by different isoforms of the eukaryotic initiation factor 4F (Zinoviev and Shapira, 2012; Gabiatti et al., 2025). In addition, both the small and large ribosomal subunits from T. cruzi are larger than their mammalian counterparts (Gao et al., 2005; Liu et al., 2016). These differences have the potential for being exploited for the development of new antiparasitic drugs for the treatment of Chagas disease.

Identifying inhibitors of T. cruzi protein synthesis by screening large chemical libraries using in vitro translation assay systems would be a straightforward strategy to identify new hits for lead optimization campaigns. However, we found that regular lysates of T. cruzi could only synthesize proteins at low levels. Based on the accumulation of hemin in T. cruzi vacuoles, we hypothesized that excess hemin or molecules derived from or modified by hemin during lysate preparation inhibited translation re-initiation by T. cruzi extract.

Herein we demonstrate that extracts derived from a T. cruzi strain deficient for hemin accumulation efficiently translate capped and polyadenylated reporter mRNAs. We optimized the assay conditions and adapted it to a 384-well format. In a proof of principle screen, we identified quinazoline compounds that inhibit translation of reporter mRNAs by T. cruzi extracts. These compounds inhibit protein synthesis in live wild-type T. cruzi. Furthermore, they inhibited the growth of epimastigotes, which are the corresponding forms of the parasite that multiply in the insect vector and, more importantly, they inhibited the proliferation of T. cruzi amastigotes in infected mammalian cells with a high (>200) selectivity index. Finally, we miniaturized this assay to a 5 μl volume, demonstrating its suitability for ultra-high throughput screening in a 1536-well format with appropriate automation.

2. Materials and methods

2.1. Compounds

The compounds used in the assays were obtained from a Broad Institute library previously tested for inhibition of Trypanosoma cruzi proliferation [PubChem AID 2294 (Buckner et al., 1996; Bettiol et al., 2009)], purchased from Molport based on similarity searches, or from the authors’ in-house collections at Harvard Medical School (Boston, MA) and from the Drugs for Neglected Diseases initiative (DNDi). All compounds were resuspended in 100% DMSO to a stock concentration of 10 mM, and aliquots were stored at −80°C to avoid freeze-thaw cycles. The list of compounds is presented in the Supplementary Table S1. The initial screen was performed at a final concentration of 40 μM, unless otherwise indicated. Analogs with >80% similarity to the most promising compounds were purchased from Molport (see Table S1).

2.2. Parasite extracts

Wild-type T. cruzi epimastigotes (WT, DM28c strain), genetically modified parasites of the DM28c strain (TCH obtained as described in the supplementary information, Fig. S1), and Sylvio X10 strains were maintained at 28°C in liver infusion tryptose medium (LIT) [75 mM NaCl, 5.4 mM KCl, 62 mM Na₂HPO₄, 0.2% glucose, 0.5% tryptose, 0.5% liver infusion, 10 μg/mL hemin], containing 10% fetal bovine serum (FBS), 59 μg/mL penicillin, and 133 μg/mL streptomycin, adjusted to pH 7.2 (Camargo, 1964). The procyclic form of Trypanosoma brucei 29-13 was maintained in SDM-79 media as described in (Machado et al., 2018). The TCH strain contained the genes of T7 RNA polymerase and Cas9 protein of Streptococcus pyogenes inserted in the pTREX vector (Beneke et al., 2017). It corresponded to a parasite clone selected in the presence of 250 μg/mL of hygromycin, aiming to replace the Vps23 gene of the endosomal sorting complexes required for transport machinery. The clone was produced by transfection in an Amaxa nucleofactor (program X14) with two sgRNA DNA templates and a donor DNA that contained the hygromycin gene flanked by 5′ and 3’ corresponding to the Vps23 gene (see description in the supplementary section).

The extracts for in vitro translation were prepared based on the protocol developed by Kovtun and collaborators, described for Leishmania tarentolae (Kovtun et al., 2011). Briefly, T. cruzi cultures in the log growth phase (2 to 3 × 10⁷/mL) in LIT-10% FBS medium were collected by centrifugation at 2500 × g for 20 min at 4°C, then resuspended and washed twice by the same centrifugation conditions in sucrose elution buffer (SEB, composed of 45 mM HEPES-KOH pH 7.6, 250 mM sucrose, 100 mM KOAc, and 3 mM Mg(OAc)₂, all from Sigma-Aldrich) at 4°C. The final pellet was then resuspended in SEB to a concentration of 1 × 10¹⁰ parasites/mL. These parasite suspensions were lysed by nitrogen cavitation under a pressure of 70 bar of nitrogen for 120 min at 4°C. The resulting lysates were centrifuged at 10,000 × g at 4°C for 20 min, and the recovered supernatant was centrifuged at 30,000 × g at 4°C for another 20 min. Each 3 mL of the supernatant was applied in a 1 × 8 cm column of Sephadex G-25 (Cytiva) pre-equilibrated with SEB without sucrose (EB). The first 3 mL were discarded, and the next 4 mL eluted with EB were retained. The extracts were quick-frozen in dry-ice ethanol in 500 μL aliquots and stored at −80°C. An aliquot was used for determining protein concentration using the Pierce BCA Protein Assay Kit (Thermo Fisher Scientific, USA), following the manufacturer's instructions. These extracts are stable for over 2 years.

2.3. RNA preparation

A synthetic DNA sequence comprising the SP6 promoter, 39 base pairs corresponding to the T. cruzi spliced leader sequence, followed by 34 randomly assigned base pairs, and the Renilla luciferase open reading frame, along with 306 base pairs of the 3′ untranslated region (UTR) of T. cruzi β-tubulin, was inserted into the KpnI and NotI sites of the pBSK + plasmid (Stratagene). The pBSK plasmid was amplified in Escherichia coli DH5α and purified using the illustra™ plasmidPrep Midi Flow Kit (GE Healthcare, USA). For RNA synthesis, plasmids were linearized with NotI (Thermo Fisher Scientific) and repurified with the PureLink Quick PCR Purification Kit (Invitrogen). Transcription reactions were carried out with the HiScribe® SP6 RNA Synthesis Kit (New England Biolabs, USA) in a final volume of 20 μL with 1 μg of template DNA following the manufacturer's instructions. The generated RNAs were recovered and quantified using the NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, USA). 10 μg of RNA was capped using the Vaccinia Capping System (New England Biolabs, USA), followed by the addition of a polyadenine tail at the 3′ end by the E. coli Poly(A) Polymerase (New England Biolabs), according to the manufacturer's recommendations. The modified RNA was analyzed by TapeStation (Agilent Technologies) and stored in aliquots containing 80% ethanol at −80°C.

2.4. Translation assay

Aliquots of frozen lysates (500 μL) were thawed and supplemented with 200 μL of (5X) translation buffer adapted from (Kovtun et al., 2011) [6 mM ATP, 12 mM Mg(OAc)₂, 200 mM creatine phosphate, 128 U/mL creatine phosphokinase from rabbit muscle, 0.7 mM of each amino acid trace from TraceCERT Amino Acids Mix Solution from Sigma-Aldrich (Cat. No. 79248), 99 mM HEPES-KOH pH 7.6, 4.98% (v/v) polyethylene glycol 3.350 from Sigma-Aldrich, 0.675 mM GTP, 1.2 mM spermidine, 10 mM dithiothreitol from Thermo-Fisher Scientific, and 5× protease inhibitor cocktail (cOmplete™ EDTA-free) from Roche, Switzerland] (Table I).

Table 1.

– Composition of stock solutions for assembling translation buffer.

Solution	Unsupplemented lysate	Stock	Translation buffer (200 μL)	Supplemented lysate	Final concentration per well
Gel filtered lysate	100 %	n/a	n/a	71.4%	49.9%
Hepes-KOH pH 7.6	45 mM	300 mM	99 mM	60.42 mM	42.3 mM
Mg(OAc)2	3 mM	40 mM	12 mM	5.57 mM	3.9 mM
KOAc	100 mM	n/a	n/a	71.42 mM	49.9 mM
PEG 3.350	n/a	16.6%	4.98%	1.42%	0.99%
AA mix TraceCERT	n/a	2.5 mM each	0.7 mM each	0.2 mM each	0.14 mM each
Spermidine	n/a	12 mM	1.2 mM	342 μM	239.4 μM
Creatine phosphokinase	n/a	1280 U/mL	128 U/mL	36.57 U/mL	25.6 U/mL
Creatine phosphate	n/a	2 M	0.2 M	57 mM	39.9 mM
Protease inhibitor	n/a	n/a	5.25X	1.5X	1X
Asparagine	n/a	2.5 mM	0.7 mM	0.2 mM	0.14 mM
Glutamine	n/a	2.5 mM	0.7 mM	0.2 mM	0.14 mM
Tryptophan	n/a	2.5 mM	0.7 mM	0.2 mM	0.14 mM
ATP	n/a	120 mM	6 mM	1.71 mM	1.19 mM
GTP	n/a	135 mM	675 μM	190 μM	135 μM
RNase inhibitor	n/a	40.000 U/mL	500 U/mL	142.8 U/mL	100 U/mL
dithiotreitol	n/a	1M	10 mM	2.8 mM	2 mM

The reactions were performed twice in duplicates in a 384-well white plate, with a final volume of 20 μL, containing 14 μL of the supplemented lysate and 2 μL of modified RNA in each well. Reactions without RNA or 10 μM of hygromycin were utilized as negative controls. To test protein synthesis inhibitors, 14 μL of supplemented lysate was mixed with 4 μL of each compound, diluted in assay buffer (final concentration of 40 μM), centrifuged, and pre-incubated at room temperature for 15 min. The reporter RNA was then added, and the reaction mixture was incubated for 2 h at 28°C.

The Renilla Luciferase Assay System (Promega, USA) was employed to measure luciferase expression, according to the manufacturer's instructions. Luminescence detection, measured in relative light units (RLU), was done immediately after the translation reaction upon the addition of 50 μL of reagent per well in the SpectraMax M3 Microplate Reader (Molecular Devices LLC, USA) with a 500 ms integration time. To rule out the possibility that the compounds are direct inhibitors of Renilla luciferase, the initial hit compounds were added to the mixture after the translation reaction, before the addition of the Renilla reagent. The effect of hemin was tested by adding 2 μL of a 0.5 mM stock diluted in 30 mM HEPES-KOH, pH 7.8. to each well in the 8 μL reaction format, and the translation tested as described above. Alternatively, hemin was added before the luciferase reaction.

2.5. Assay miniaturization to low-volume 384-well format

We adapted and optimized the assay to low reaction volumes (8 μL and 5 μL) using the same parasite extracts to test multiple reagents with the best possible sensitivity. We incubated mixtures of 3 μL of the supplemented lysate (2.6 μg protein/μL) and 2 μL of each compound (40 μM final) for 15 min at room temperature and added 3 μL of RNA (9 ng total) to initiate translation. The in vitro translation reactions were incubated for 2 h at 28°C, and the reporter activity was determined by adding 8 μL of Renilla luciferase reagent and reading by luminometer. To further miniaturize assays, we used 2 μL of the supplemented parasite lysate, 1 μL of compound, and 2 μL of mRNA per reaction. Translation was also performed with the rabbit reticulocytes (Retic Lysate IVT kit, Ambion) following the manufacturer's instructions using 0.6 μg per reaction of capped and polyadenylated mRNA employed for the in vitro assays with T. cruzi extracts.

2.6. Puromycylation assay and western blotting

T. cruzi epimastigotes (WT, DM28c strain) maintained at 28°C in LIT medium containing 10% FBS at 1 × 10⁷ parasites/mL were centrifuged at 2000 × g for 5 min, washed, and resuspended in LIT-10% FBS to the 1 × 10⁷ parasites/mL concentration. The parasites were then pre-incubated at 28°C for 1 h after adding 1 μL of each compound pre-diluted in DMSO for each 1 mL parasite suspension to obtain a concentration of 40 μM. With this concentration of DMSO, we observed no changes in T. cruzi motility and viability measured using a Guava Muse cell analyzer. The controls included incubation with DMSO and 5 μM cycloheximide. Thereafter, 10 μg/mL puromycin (Thermo Fisher Scientific) was added to all tubes, which were maintained at 28°C for 15 min, followed by centrifugation at 2000 × g for 5 min, washing with PBS, and lysis with 50 μL of RIPA buffer (50 mM Tris, 150 mM NaCl, 0.5% sodium deoxycholate, 1% NP-40, 0.1% SDS pH 8.0). 10 μL of lysates were used to measure the protein content using the Pierce BCA Protein Assay Kit (Thermo Fisher Scientific, USA). The remaining solution was mixed with SDS-PAGE sample buffer (4% SDS, 20% glycerol, 10% 2-mercaptoethanol, 0.004% bromophenol blue, 120 mM Tris HCl, pH 6.8), heated at 95°C for 5 min, and 20 μg protein was loaded into a 10% SDS-PAGE. After electrophoresis, the proteins were transferred to nitrocellulose membranes using a semidry Bio-Rad transblot at 20 V for 40 min. After staining in 0.5% Ponceau S in 1% v/v acetic acid, the membranes were soaked in 5% BSA in 10 mM Tris-HCl pH 7.4 (TBS). After blocking, the membrane was washed three times with TBS with 0.05% Tween-20 (TBS-Tween) and incubated with anti-puromycin antibody (mouse ab, PMY-2A4-s, Developmental Studies Hybridoma Bank) diluted 1 to 100 in 5% BSA-TBS containing 0.02% NaN₃ overnight at 4°C. Following 3 × 10-min washes in TBS-T, bound antibodies were detected by incubation with IRDye680 goat anti-mouse IgG diluted 1 to 10,000 in TBS-T, washed again, and visualized using the Odyssey Fc Imager (LI-COR Biosciences - USA). As a loading control, we used a rabbit anti-aldolase antibody (Barbosa Leite et al., 2020), diluted 1 to 2,000.

2.7. Parasite inhibition

Epimastigotes were maintained at 28°C in LIT medium with 10% FBS at 1 × 10⁷ parasites/mL. The parasites were diluted to 3.75 × 10⁶ parasites/mL, and 40 μL (150,000 parasites) were plated in triplicate on 96-well black/clear bottom plates containing 40 μL of testing compounds 39G and 39H in decreasing concentrations from 50 μM to 0.39 μM, previously diluted to achieve a final concentration of 0.5% DMSO. We employed benznidazole at 200 μM concentration and 0.5% DMSO as positive and negative controls, respectively. The reaction blank was the LIT medium. Parasite growth was indirectly measured using the SYBR Green assay modified from (Faria et al., 2015). The plates were kept at 28°C, and after 0 to 72 h, 20 μL of lysis solution (30 mM Tris pH 7.5, 7.5 mM EDTA, 0.012% saponin, 0.12% Triton X-100, and SYBR Green I 1:2000 from Thermo-Fisher Scientific) was added to each well and incubated protected from light at 28°C for 2 h under agitation, followed by fluorescence reading in the SpectraMax M3 Microplate Reader (Molecular Devices LLC, USA) with excitation at 485 nm and emission at 530 nm. The assay was performed in biological duplicates.

U2OS cells were used as hosts to assess inhibitory effects of hit compounds on intracellular parasite growth. The cells were grown at 37°C and 5% CO₂ in 75 cm² flasks with 80% confluence in Dulbecco's Modified Eagle's Medium (DMEM) with 10% FBS, 50 μg/mL penicillin, and 133 μg/mL streptomycin. After three subcultures, the cells were trypsinized and plated on black/clear bottom 384-well plates at 500 cells per well. After 24 h, cells were infected with 5000 T. cruzi trypomastigotes (Sylvio X10/1 clone). Following 24 h incubation, the infected cells were treated with decreasing concentrations of compounds 39G and 39H. The final DMSO content was 0.5% in all cases. Benznidazole was used as the treatment control, while 0.5% DMSO was used as the untreated control. After 48 h of drug treatment, the cells were fixed with 12% paraformaldehyde to a final concentration of 4%. Plates were then incubated for 1 h at room temperature before being washed with PBS 1× and stained with DAPI (4′,6-diamidino-2-phenylindole) and DRAQ 5 (ThermoFisher Scientific) for further analysis, as described in (Moraes et al., 2014). Images were acquired employing the Operetta CLS High-Content Analysis System, featuring a mask for amastigotes, nuclei, and DNA. Following acquisition, the data was normalized and plotted. The mean and standard deviation are expressed as a ratio of cells to amastigotes. EC₅₀ and CC₅₀ were computed using nonlinear fit [inhibitor] vs. response (three parameters) analysis in GraphPad Prism v10.

3. Results

3.1. Establishment of optimal conditions for in vitro translation by T. cruzi extracts

We initially tested the capacity of wild-type extracts made from T. cruzi epimastigote cultures prepared as described for L. tarentolae to translate an mRNA containing the spliced leader and 5′UTR sequence fused to the Renilla luciferase open reading frame and a 3′UTR of the T. cruzi β-tubulin, without success. It was intriguing that by using Trypanosoma brucei procyclic extracts made in parallel, we could generate active luciferase (data not shown). A characteristic of T. cruzi epimastigote cultures is the accumulation of hemin within the parasite as a growth requirement, which results in brownish fractions during extract preparations. Recently we obtained a mutant parasite (TCH), which accumulated much less hemin due to its incapacity to adequately perform endocytosis (manuscript in preparation). Supposing that the presence of hemin might interfere with the extract preparation, we decided to test the capacity of an extract prepared from this mutant to perform translation. In fact, the resulting parasite extract translated capped and polyadenylated reporter mRNA significantly more efficiently than the wild-type parasite extract (Fig. 1A). Both extracts contained similar amounts of protein (Fig. 1B), but the wild-type extracts presented a yellow color, probably resulting from a high concentration of hemin (and/or its metabolites). Nevertheless, the addition of hemin up to 30 μM did not affect translation using the TCH extract, ruling out a direct inhibitory effect. Most likely, hemin disturbed the components required for translation during the extract processing.

Fig. 1.

A cell extract from a modified Trypanosoma cruzi line is able to promote in vitro translation of Renilla. (A) Renilla luciferase activity obtained by 3 h at 28°C incubation of a supplemented extract from a modified (TCH) and original DM28c strain (WT) in the absence (−) or presence of 3 ng/μL capped and polyadenylated Renilla luciferase containing the SL sequence (+), followed by the addition of Renilla luciferase reagent (Promega) and Relative Light Units (RLU) detection. The values are means and SD of triplicates analyzed by one-way ANOVA with Šídák's multiple comparisons test (p < 0.0001). (B) Total protein content used in TCH and WT extracts, quantified by Pierce BCA Protein Assay. The values correspond to triplicates analyzed by one-way ANOVA and p < 0.02 calculated with Dunnet's pos-test.

To establish ideal conditions for the assay, we then tested different RNA concentrations, incubation times, TCH extract concentrations, and various mRNA modifications. The translation of unmodified in vitro transcribed RNA was almost non-existent. Capping the mRNA, particularly capping and polyadenylation, dramatically increased the translation (Fig. 2A), demonstrating the importance of those modifications to the parasite's translation machinery. The luminescence signal was directly proportional to the concentration of fully modified RNA up to 3 ng/μL (Fig. 2B) with a linear signal up to 3 h reaction time (Fig. 2C). Translation was also linear by up to 3 μg/mL of the supplemented parasite extract (Fig. 2D). A decline in translation was observed at higher protein concentrations. These optimal conditions were then used in the subsequent experiments.

Fig. 2.

Effect of mRNA modifications, concentration, time of incubation, and extract concentration on Renilla luciferase translation. (A) Renilla luciferase activity using 20 ng of SP6-transcribed RNA (mRNA), before or after capping (Cap) or polyadenylation (PolyA) modifications after 3 h with supplemented TCH extract, followed by Relative Light Units (RLU) detection. (B) RLU of translation reactions with different concentrations of capped and polyadenylated mRNA incubated with supplemented TCH extract. (C) RLU of translation reactions of 20 ng of capped and polyadenylated mRNA incubated for the indicated times with supplemented TCH extract. (D) RLU produced after translation using different extract concentrations with 20 ng of capped and polyadenylated mRNA for 3 h. The results are means and standards (SEM) of means of triplicate experiments, each made in triplicate samples. Asterisks indicate significant differences (p < 0.0001) using a one-way ANOVA with Tukey's multiple comparisons test.

We next validated this assay by demonstrating that known general protein synthesis inhibitors reduced translation when tested at 10 μM (Fig. 3A). Paromomycin, a protein synthesis inhibitor used to treat visceral leishmaniasis (Coser et al., 2020; Pokharel et al., 2021), did not inhibit translation in T. cruzi extract (Fig. 3B) and did not affect parasite growth (Fig. S2). In contrast, hygromycin fully inhibited above 50 nM and cycloheximide above 0.5 μM (Fig. 3C). Furthermore, hemin inhibited the translation reaction above 1.5 μM when added with the translation extract (Fig. S3). Above 20 μM, it also affected the luciferase luminescence reaction. This could partially explain the reduced translation using wild-type extracts of T. cruzi.

Fig. 3.

Translation of Renilla luciferase is inhibited by known protein synthesis inhibitors. (A) Capped and polyadenylated Renilla mRNA (20 ng) was incubated with the supplemented TCH extract in the presence of 10 μM of the indicated compounds for 3 h, followed by luminescence detection. The values show the percentage of inhibition relative to the RLU produced in samples containing DMSO. (B) Capped and polyadenylated Renilla mRNA (9 ng) was incubated with the supplemented TCH extract (2.6 μg/μL) in the presence of 40 μM of paromomycin and benznidazole for 2 h in a final volume of 8 μL, followed by luminescence detection. Each point represents a mean of triplicate measurements (n = 3) in two different experiments. (C) Translation reactions performed with different concentrations of hygromycin B and cycloheximide. All values are means and SEM of triplicates.

3.2. Proof-of-concept pilot screen

We reasoned that protein synthesis inhibitors should be enriched among the compounds that inhibit protozoan proliferation. We therefore assembled a small library of such agents rather than testing a larger random library as a proof-of-concept demonstration. We compiled a 128-member library composed of Broad Institute and DNDi compounds (Table S1) and screened it at a final concentration of 40 μM (see materials and methods). The Z′ scores ranged from 0.68 to 0.92, and a signal-to-noise ratio (SNR) exceeded 200. Of these, eleven compounds reproducibly reduced the Renilla luciferase signal in the in vitro translation assay, ranging from 20% (3, 39, 124, and 125) to nearly 100% inhibition (108) (Fig. 4A). To determine whether the compounds were bona fide inhibitors of translation rather than inhibitors of the enzymatic activity of Renilla luciferase (assay interference), we incubated the extract with these 11 compounds either before or after the 2-h translation reaction and measured the Renilla luciferase activity. We reasoned that if a compound inhibits translation but not enzyme activity of Renilla luciferase, it will display inhibitory activity only if added to the translation reaction before the addition of mRNA. In contrast, a compound that inhibits the enzymatic activity of Renilla luciferase would display inhibitory activity regardless of whether it is added to the reaction before or after the start of the in vitro translation reaction. A comparison of luciferase activity of compound addition before and after the in vitro translation reaction indicates that compounds 39 and 97 (quinazolines) and also 3 are most likely inhibitors of translation (Fig. 4B), while compounds 48, 77, and 108 are direct inhibitors of the Renilla luciferase enzyme. Compounds 25 and 43 exhibited a noticeable impact on translation when administered prior to its initiation; nevertheless, they also seemed to inhibit the enzymatic activity of Renilla luciferase when introduced post-translation, although the results were not conclusive. All results are presented in Table S2.

Fig. 4.

In vitro translation inhibition by compounds known to affect parasites. (A) Capped and polyadenylated Renilla mRNA (20 ng) was incubated with the supplemented TCH extract in the presence of 40 μM of the compounds named in Supplementary Table S1 for 3 h, followed by luminescence detection. The values show the triplicate means of the percentage of inhibition of three independent experiments (n = 3) relative to the Relative Light Units (RLU) produced in samples containing DMSO. The significant inhibition (p < 0.05 using two-way ANOVA and Tucker test) is shown in green symbols. The inhibition by hygromycin at 10 μM is shown in red symbols. (B) Luminescence values obtained with the same reaction as in (A) with the compounds added before the translation reaction (green bars) or after the translation reactions relative to the values of DMSO (blue bar). The values are means and SEM of three independent experiments (n = 3) performed in technical triplicates. Compounds that were added after translation and which produced less than 75% luminescence were considered Renilla inhibitors.

3.3. Miniaturization of assays to low-volume and concentration-response studies of hits and their analogs

To demonstrate the utility of the miniaturized assay for testing and considering the structural similarities of compounds 39 and 97 (quinazolines), which showed specific translation inhibition, we decided to test a small commercial library of analogs of compounds 39/97 (Fig. 5A). The quinazoline analogs 39A, 39B, 39D, 39G, 39H, 39I, and 39K decreased translation by over 50%, whereas 39C, 39F, and 39L demonstrated around 50% inhibition, and 39M exhibited less than 40% inhibition (Fig. 5B and Table S3). We selected three quinazolines (39, 39D, and 39L) to conduct a dose-response curve, utilizing doses ranging from 2.5 μM to 40 μM. Compound 39 exhibited a 65% inhibition at 30 μM, whereas 39D and 39L only achieved this inhibition at 40 μM (Fig. 5C and Table S3). In the miniaturized 8 μL tests, the medium Z′ score was 0.77 and SNR > 200. Unfortunately, we could not retest compound 97 in these dose-response studies because this compound was no longer available. These data indicate that structural modifications influence the efficacy of the quinazolines in inhibiting T. cruzi protein synthesis.

Fig. 5.

Effect of compound analogs on in vitro translation. (A) Analogous compounds. (B) Capped and polyadenylated Renilla mRNA (9 ng) was incubated with the supplemented TCH extract (2.6 μg/μL) in the presence of 40 μM of the compounds shown in (A) for 2 h in a final volume of 8 μL, followed by luminescence detection. Each point represents a mean of triplicate measurements. The horizontal bar indicates the mean, and the vertical bar the SEM with p < 0.0001 (red), p < 0.001 (orange), p < 0.01 (green), and non-significant (black) as determined by two-way ANOVA with Dunnett multiple comparisons. (C) Inhibition of Renilla activity by the indicated compound concentrations as in (A). Each point represents the means and SEM of triplicate measurements in three independent experiments. Asterisks indicate significant inhibition (p < 0.05) as determined by a two-way ANOVA test with Tukey comparisons. (D) Luminescence assays were made with 2 μL of parasite extract, 2 μL of mRNA (22 ng), and 1 μL of compounds to 40 μM.

The volume of 5 μl was chosen for further miniaturization because it is the most often used volume in 1536-well ultra-high throughput screening assays. Gratifyingly, the assay performed as well in 5 μl as it did in the 8 μl format, as we demonstrated a clear inhibition of in vitro translation by compound 39 (Fig. 5D). The Z′ score was 0.9 and SNR > 200. We confirmed that the analogs weren't Renilla luciferase inhibitors by adding the compounds after the translation reaction (Fig. S4). We also verified that the select compound (39G) did not inhibit the protein synthesis using rabbit reticulocytes (Fig. S5).

3.4. Inhibition of protein synthesis in living parasites by selected compounds

To conclusively demonstrate that our confirmed inhibitors of in vitro translation indeed inhibit protein synthesis in living parasites, we conducted a puromycylation assay in which puromycin was incorporated into the elongating protein chains (David et al., 2012; Aviner, 2020). We treated exponentially proliferating T. cruzi WT Dm28c epimastigotes with selected quinazoline analogs for 1 h prior to the addition of trace amounts of puromycin. As shown in the two panels of Fig. 6A, compounds 39, 43, 39D, 39G, and 39H at 40 μM inhibited puromycin incorporation into elongating protein chains, indicating these compounds do indeed inhibit protein synthesis in living wild-type parasites. A quantitative analysis conducted in three distinct experiments indicates that these agents decrease translation considerably (Fig. 6B). The results demonstrate that our in vitro translation assay is capable of identifying compounds that block T. cruzi protein synthesis in living parasites.

Fig. 6.

Effect of selected compounds on the puromycylation in intact T. cruzi epimastigotes. (A) Epimastigotes at 1 × 10⁷/mL were incubated in LIT medium containing 10% FBS for 1 h at 28°C with the indicated compounds at 40 μM. Puromycin was added to each sample to 10 μg/mL, and after 15 min at 28°C, the parasites were collected by centrifugation, washed once with PBS, and lysed with RIPA buffer to estimate the amount of protein per assay. The samples were then mixed with SDS-PAGE sample buffer and used for Western blots probed with mouse anti-puromycin and rabbit anti-aldolase as loading control and detected using anti-IgG IRDye 680 and 800, respectively. Markers (M) were added to the left side of each gel, and the respective size was indicated in kDa. (B) Quantitative analysis of puromycin incorporation relative to the incorporation in the sample containing only DMSO in the region between 150 and 20 kDa, from 3 to 4 independent experiments, except for compound 39G, tested in only two experiments. The values are means and SEM, with p < 0.001 (*), estimated by two-way ANOVA with Dunnett's multicomparisons tests relative to the sample with DMSO.

3.5. Quinazolines inhibit T. cruzi proliferation

To further determine if quinazoline analogs could reduce parasite growth, we first assessed the effect of compounds 39G and 39H on epimastigote proliferation. The EC₅₀ for compound 39G is 2.8 ± 0.2 μM, while compound 39H has an EC₅₀ of 4.8 ± 2.0 μM (Fig. 7A). The growth curves reveal that 39G achieved maximal inhibition of epimastigote growth at ∼15 μM, whereas > 50 μM of 39H was required to fully block parasite proliferation.

Fig. 7.

Impact of quinazolines on T. cruzi proliferation. Epimastigotes (DM28c) were incubated in LIT medium supplemented with 10% FBS and decreasing amounts of the specified compounds. Following a 72-h period, SYBR Green lysis buffer was added, and the resulting fluorescence signal was normalized based on the initial fluorescence. The results are means and SD from two independent experiments, and the EC₅₀ was determined using GraphPad Prism. U2OS cells (B) and (C) were infected with trypomastigotes and subsequently treated with compounds 39G and 39H at concentrations ranging from 100 μM to 0.7 nM, respectively. The values represent the mean and standard deviations of the ratio of detected amastigotes to cell counts per well, with 100% corresponding to wells without compounds and 0% to the count of non-infected cells. EC₅₀ and CC₅₀ were determined using nonlinear fitting of [inhibitor] against response (three parameters) analysis in GraphPad Prism. The selectivity index was determined using the EC₅₀ and CC₅₀ averages for each specified chemical.

Subsequently, we assessed the effect of these compounds on U2OS cells infected with T. cruzi trypomastigotes of the Sylvio strain to evaluate both cytotoxicity and efficacy in inhibiting the proliferation of intracellular amastigotes. 39G exhibited antiparasitic activity with an EC₅₀ value of 0.08 ± 0.07 μM, while cytotoxicity was characterized by a CC₅₀ of 33 ± 20 μM (Fig. 7B). Compound 39H also showed inhibitory effects on amastigotes (EC₅₀ of 0.4 ± 0.2 μM) with comparatively low cellular toxicity (CC₅₀ 84 ± 40 μM, Fig. 7C). These data demonstrate very high selectivity indexes (∼415 for 39G and ∼200 for 39H). Comparable values were noted in an analogous experiment for 39H and slightly high values for 39G, utilizing Vero cells and trypomastigotes from the DM28c parasite strain (Fig. S6).

4. Discussion

We established an ultra-high-throughput in vitro translation assay to identify novel T. cruzi protein synthesis inhibitors. We utilized extracts from a genetically modified parasite exhibiting decreased hemin incorporation relative to wild-type strains. The fact that it was possible to detect significant amounts of luciferase translation in the mutant was likely due to improved fractionation of the active translation components.

We also confirmed that efficient in vitro translation requires capping at the 5′ end and polyadenylation at the 3′ end as well as the presence of a spliced leader. However, the cap incorporated into the in vitro transcribed RNA by the vaccinia enzyme was Cap0, characterized by the sole presence of 7-methyl guanosine, equivalent to the majority of mammalian mRNA, while trypanosomes possess a cap with supplementary methylations (Cap4) (Perry et al., 1987; Mair et al., 2000). Therefore, we suggest that either the extract carried active methylating enzymes, thus further modifying the reporter mRNA before translation, or that the entire Cap4 was not necessary for in vitro translation.

In this work we optimized the incubation time, RNA concentrations, DMSO concentration, and extract concentrations to obtain an assay with high performance. The assay robustness was evaluated using the Z′-factor. The Z′-scores ranged from 0.68 to 0.92, and the signal-to-noise ratio (SNR) exceeded 200, which is considered excellent and extremely robust. We miniaturized the assay to minimize reagent use, which should facilitate screening of very large compound libraries. The further miniaturization had a Z′-score of 0.9 and SNR > 200, showing that the assay can be miniaturized with no loss of robustness, indicating that this is the best future approach for these tests.

We also validated our assay by demonstrating that translation was completely obliterated by general eukaryotic protein synthesis inhibitors. Interestingly, paromomycin, which preferentially inhibits Leishmania protein synthesis, had no effect on T. cruzi protein synthesis in vitro or in living parasites. Given the similarities between Leishmania and T. cruzi, as both belong to the Trypanosomatidae family, we were surprised that paromomycin did not inhibit T. cruzi translation. Furthermore, we noticed that paromomycin did not impact T. cruzi proliferation (data not shown). These results strongly support our contention that T. cruzi-specific protein synthesis inhibitors can be identified.

In these proof-of-concept studies, we utilized agents previously shown to inhibit T. cruzi proliferation and antiparasitic compounds from DNDi, intending to investigate whether compounds with known activity and unknown MoA could act via protein synthesis inhibition among antimicrobial, antimycotic, antimalarial, and anti-leishmanial agents. Furthermore, these compounds were already shown to have a very low risk of having toxicity to mammalian cells (Ahyong et al., 2016; Hilal-Dandan and Brunton, 2016; O'Sullivan et al., 2018; Dmitriev et al., 2020). After a series of follow-up tests, we determined the initial hit rate to be ∼6% with a confirmed hit rate of ∼50%. Moreover, we found that two structurally similar quinazoline compounds (39 and 97), previously identified in high-throughput screening as intracellular amastigote growth inhibitors of T. cruzi were highly effective as inhibitors of T. cruzi protein synthesis. In follow-up studies, newly purchased 39 (in powder form) was significantly more active than the DMSO aliquot we used in the initial test. Our study additionally showed that derivatives of these compounds also inhibit proliferation of epimastigotes and intracellular forms of the parasite at concentrations that are non-toxic for mammalian cells.

Several compounds from this class have been described in the literature as having T. cruzi activity but with no information about target validation. A 14-quinazoline-2,4,6-triamine derivative and compounds with nitrobenzoyl substituents at the 6-position demonstrated significant anti-trypanosomal activity against both epimastigote and trypomastigote forms while exhibiting low toxicity to human cells (Vazquez et al., 2024). Other studies also found that quinazolinone and quinazoline derivatives might be used against T. cruzi (Bollini et al., 2019). Therefore, our study may suggest that these compounds’ anti-Trypanosoma activities are mediated by inhibition of protein synthesis. In mammalian cells compound 39 was shown to inhibit PAK4, a serine/threonine-protein kinase that regulates cytoskeleton remodeling and acts on tumor cells (Pereira-Veiga et al., 2025). Compound 97 was only described to inhibit Plasmodium falciparum, the agent of malaria, but its target has not been identified (Plouffe et al., 2008). In the same study, 39G and 39H were also recognized as P. falciparum inhibitors. Future studies will investigate whether quinazoline acts directly to inhibit components of translation machinery or indirectly through proteins that play a role in translation initiation, elongation, or termination, both in intact cells and cell extracts.

In summary, we developed a highly efficient ultra-high-throughput assay for screening T. cruzi protein synthesis inhibitors for the development of novel therapeutics for Chagas disease. Our studies suggest that certain quinazoline derivatives may be able to inhibit T. cruzi protein synthesis and parasite proliferation, but additional work to explore structure-activity relationships and modes of action is required. It remains unknow the mechanism by which the identified compounds specifically inhibited protein synthesis in T. cruzi. We are currently addressing this issue focusing on target identification. remains unknow the mechanism by which the identified compounds specifically inhibited protein synthesis in T. cruzi. We are currently addressing this issue focusing on target identification.

CRediT authorship contribution statement

Camila Colombari Mantovani: Data curation, Formal analysis, Investigation, Methodology, Visualization, Writing – original draft, Writing – review & editing. Flavia R. Novais Freitas: Formal analysis, Investigation, Methodology, Visualization. Carolina Borsoi Moraes: Formal analysis, Methodology, Supervision, Writing – review & editing. Guilherme Rodrigo Reis dos Santos: Conceptualization, Investigation, Methodology. Luiza R. Cruz: Data curation, Resources, Validation, Writing – review & editing. Jadel Muller Kratz: Data curation, Formal analysis, Methodology, Resources, Supervision, Writing – review & editing. Bertal Huseyin Aktas: Conceptualization, Data curation, Formal analysis, Funding acquisition, Project administration, Resources, Supervision, Validation, Writing – original draft, Writing – review & editing. Sergio Schenkman: Conceptualization, Data curation, Formal analysis, Funding acquisition, Project administration, Supervision, Writing – original draft, Writing – review & editing.

Availability of reagents

The TCH strain is available upon request following Brazilian regulatory rules for non-commercial use.

Declaration

I declare, in the name of all authors, that we don't have competing interest in the article “Identification of new Trypanosoma cruzi-specific protein synthesis inhibitors by using a developed high-throughput in vitro translation assay”.

Conflict of interest statement

The authors declare no competing financial interest.

Appendix A. Supplementary data

The following are the Supplementary data to this article: