Targeting L-Proline Uptake as New Strategy for Anti-chagas Drug Development

L-Proline is an important amino acid for the pathogenic protists belonging to Trypanosoma and Leishmania genera. In Trypanosoma cruzi, the etiological agent of Chagas disease, this amino acid is involved in fundamental biological processes such as ATP production, differentiation of the insect and intracellular stages, the host cell infection and the resistance to a variety of stresses. In this study, we explore the L-Proline uptake as a chemotherapeutic target for T. cruzi. Novel inhibitors have been proposed containing the amino acid with a linker and a variable region able to block the transporter. A series of sixteen 1,2,3-triazolyl-proline derivatives have been prepared for in vitro screening against T. cruzi epimastigotes and proline uptake assays. We successfully obtained inhibitors that interfere with the amino acid internalization, which validated our design targeting the metabolite's transport. The presented structures are one of few examples of amino acid transporter inhibitors. The unprecedent application of this strategy on the development of new chemotherapy against Chagas disease, opens a new horizon on antiparasitic drug development against parasitic diseases and other pathologies.


INTRODUCTION
Chagas disease is one of the most neglected infectious disease. It is endemic in the Americas, with 8-10 million people infected and 25 million people at risk (Nunes et al., 2013). The disease is divided in the acute and the chronic phase. The first have a noticeable parasitemia, the absence of humoral response and is largely asymptomatic. The chronic phase has a non-evident parasitemia and a robust IgG response being asymptomatic in 60-70% of the cases (Pérez-Molina and Molina, 2018). The chemotherapy against Chagas disease relies mainly on two drugs introduced more than 40 years ago: Nifurtimox (Nf) and Benznidazole (Bz) (Urbina, 2010). Both drugs are efficient on the acute phase, but in the chronic phase is controversial (Morillo et al., 2015). In addition, severe side effects due to toxicity and the emergence of resistance calls for urgent development of new drugs (Guedes et al., 2011).
Trypanosoma cruzi is a hemoflagellated parasite that causes Chagas disease. This parasite presents a complex life cycle among two kinds of hosts: mammals and reduviid insects, which transmit the infection. Along its life-cycle at least four stages were clearly identified, epimastigotes (replicative stage) and metacyclic trypomastigotes (infective, non-replicative stage) in the insect and blood stream trypomastigotes and amastigotes (intracellular replicative stage). Also, during its life-cycle the parasite faces different environments and it must adjust its "life-style" including its metabolism to changes in nutrients availability, temperature, together other environmental variables.
Among many other important metabolites, amino acids are particularly relevant for the biology of T. cruzi, besides protein synthesis (Marchese et al., 2018), playing fundamental roles in energy management (Pereira et al., 2000) and nitrogen metabolism Girard et al., 2018). When epimastigotes proliferation arrests, there is a metabolic switch from a carbohydrates to an amino acids based metabolism, with a consequent change in the protein expression profile (Barisón et al., 2017;Avila et al., 2018). In fact, it has been demonstrated that amino acids such as proline (Paes et al., 2013), histidine (Barison et al., 2016), and even alanine (Girard et al., 2018), as well as the proline oxidation product P5C (Mantilla et al., 2015), can fuel electrons to the respiratory chain, powering the mitochondrial ATP synthesis (Sylvester and Krassner, 1976;Martins et al., 2009;Paes et al., 2013;Barison et al., 2016). Some neutral amino acids can also function as osmolytes, serving to counteract volume perturbations following a shift in extracellular osmolarity (Rohloff et al., 2003;Silber et al., 2005;Avila et al., 2018).
Taking the proline uptake as a novel drug target we decided to develop new transporter inhibitors and evaluate their antiproliferative activity against Trypanosoma cruzi. These new compounds were initially evaluated on T. cruzi epimastigotes, validating their action mechanism by proline transport experiments. A comprehensive analysis of the structure-activity relationship allowed a rational pipeline to design selective metabolite transporter inhibitors.

General Remarks
Chemical reagents were purchased from commercial suppliers and used without further purification, unless otherwise noted. Dry, deoxygenated diethyl ether (Et 2 O), tetrahydrofuran (THF), and dichloromethane (DCM) were obtained bypassing commercially available pre-dried, oxygen-free solvents through activated alumina columns. DMF was distilled from BaO. Reactions were monitored by thin-layer chromatography (TLC) performed on 0.2 mm Merck silica gel aluminum plates (60F-254) and visualized using ultraviolet light (254 nm) and by potassium permanganate and heat as developing reagents. All reactions were performed under an atmosphere of nitrogen using oven-dried glassware and standard syringe/septa techniques. Column chromatography was performed with silica gel 60 (230-400 mesh). Yields were calculated for material judged homogeneous by thin layer chromatography (TLC) and nuclear magnetic resonance ( 1 H NMR).
1 H and 13 C NMR spectra were acquired on a Bruker Avance II 300 MHz (75.13 MHz) using CDCl 3 as solvent. Chemical shifts (δ) were reported in ppm downfield from tetramethylsilane and coupling constants are in hertz (Hz). NMR spectra were obtained at 298 K unless otherwise stated and samples run as a dilute solution of the stated solvent. All NMR spectra were referenced to the residual undeuterated solvent as an internal reference. The following abbreviations were used to explain the multiplicities: s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet, br = broad. Assignment of proton resonances was confirmed by correlated spectroscopy. Electrospray ionization high-resolution mass spectra (ESI-HRMS) were recorded on a Bruker MicroTOF II. IR spectra were obtained using an FT-IR Shimadzu spectrometer and only partial spectral data are listed. Melting points were measured on an Electrothermal 9100 apparatus and are uncorrected.

Experimental Procedures and Spectroscopic Data
Synthesis of N-Propargyl Methyl Prolinate (2) To a solution of methyl prolinate (200 mg, 1.22 mmol) in 10 mL of Et 2 O (anh) , NEt 3 (439 mg, 4.3 mmol) and 80 % propargyl bromide in toluene (263 µL, 2.45 mmol) were added in this order and the reaction mixture was stirred at room temperature for 12 h. The solvent was evaporated, and the crude product was purified by column chromatography in silica gel with increasing ethyl acetate/hexane gradient to yield the expected product as a yellow oil (129 mg, 72 % General Procedure for the Cu(I) Mediated 1,3-Dipolar Cycloaddition N-Propargyl methyl prolinate (1 eq) and the azide (1.1 eq) were suspended in 10 mL/eq of tBuOH:H 2 O (1:1) and then 1M CuSO 4 solution (0.05 eq) and finally 1M sodium ascorbate solution (0.2 eq) and the mixture stirred overnight at room temperature. Brine was added, and the solution was extracted with dichloromethane. Combined organic extracts were dried over sodium sulfate and evaporated. Products were purified by column chromatography in silica gel with increasing hexanes/ethyl acetate/methanol gradients.

BIOLOGY Reagents
All reagents were purchased from Sigma-Aldrich (St. Louis, MO, USA). Culture medium and fetal calf serum (FCS) were purchased from Cultilab (Campinas, SP, Brazil).

Growth Inhibition Assays
T. cruzi epimastigotes in the exponential growth phase (5.0-6.0 x 10 7 cells mL −1 ) were cultured in fresh LIT medium. The cells were treated with different concentrations of drugs or not treated (negative control). A combination of Rotenone (60 µM) and Antimycin (0.5 µM) was used as a positive control for inhibition as previously described. (Magdaleno et al., 2009) The cells (2.5 x 10 6 mL −1 ) were transferred to 96well culture plates and incubated at 28 • C. Cell proliferation was quantified by reading the optical density (OD) at 620 nm for 8 days. The OD was converted to cell density values (cells per mL) using a linear regression equation previously obtained under the same conditions. The concentration of compounds that inhibited 50% of parasite proliferation (IC 50 ) was determined during the exponential growth phase (4 days) by fitting the data to a typical sigmoidal dose-response curve using OriginPro8. The compounds were evaluated in quadruplicate in each experiment. The results shown here correspond to three independent experiments. As a cell growth inhibition control, growth curves in which 200 mM rotenone and 0.5 mM antimycin were added to the culture medium were run in parallel for all experiments.

L-Proline Transport Inhibition Assay
Parasites in exponential growth phase were washed three times by centrifugation and resuspended in phosphate buffered saline (PBS), pH 7.4. Cells were counted in a Neubauer chamber, adjusted with the same buffer to a final density of 200 × 10 6 cells/mL and distributed in aliquots of 100 µL (containing 20 × 10 6 cells each). Transport assays were initiated by the addition to the assay tubes of 100 µL of the desired dilution of Lproline in PBS in the presence of 0.5 mCi of L-[3H] proline. Unless otherwise specified, V0 was measured at 28 • C for 30 s by incorporation of 0.75 and 3 mM L-proline traced with 1 µCi of U-14 C-L-Pro (Perkin Elmer). In all cases, proline transport was stopped by addition of 800 mL of cold 50 mM proline in PBS (pH 7.4), and rapid washing by centrifugation at 10,000x g for 2 min. Background values in each experiment were measured by simultaneous addition of radiolabeled L-proline and stop solution as previously described (Silber et al., 2002).

Competition Assay
For the competition assays, the transport experiments were performed as described above using the L-Pro concentration corresponding to the K m (0.31 mM). Those conditions were chosen considering an inhibitory activity by structurally related compounds should be competitive, and so, should be evidenced by a change in the V max at L-Pro concentrations close to the K m . Non-competitive inhibitors, if any, should diminish V max , which is measured at 10 x K m L-Pro (3.1 mM) (Silber et al., 2002).

Statistical Analysis
One-way ANOVA followed by the Tukey test was used for statistical analysis. The T-test was used to analyze differences between groups. P<0.05 was considered statistically significant.

Synthesis of L-Proline Transport Inhibitors
A proper uptake blocker needs to be recognized by the transporter, but not being able to go through, by adding a bulky substituent or a membrane interacting portion that prevents its transport (Figure 1).
This model will require a compound holding a recognition motif, a linker and a membrane anchor or a voluminous substituent to block the L-proline incorporation. The recognition portion will have L-proline to specifically interact with the  transporter, the membrane anchor should be a non-polar group and both parts will be connected by a 1,2,3-triazole (Figure 2).
To validate the proposed model, we proceeded to make a small library following the synthetic strategy shown on Scheme 1. The synthesis started from commercial L-proline, preparing the key intermediate in two steps, including an esterification to produce the methyl prolinate 1 followed by an N-alkylation with propargyl bromide. Once the required key propargyl methyl prolinate 2 intermediate was prepared, a pool of different azides covering a variety of steric moieties including, aryl, alkyl and isoprenyl substituents was selected to explore their capacity to interact with the membrane. Azides were prepared by direct substitution of bromide with sodium azide on DMF except for geranyl-, farnesyl-, and phytylazides, which were synthesized from geraniol, farnesol and phytol, respectively, using diphenylphosphorylazide (DPPA) following Thompson's procedure (Thompson et al., 1993). Phytylazide was found to be a mixture of three chemical entities in equilibrium: tertiary azide, E and Z isomers of the primary azide, following the same behavior previously observed on geranyl-, and farnesylazide (Porta et al., 2014). Allylic azides can be obtained as a mixture, because they exist as equilibrating mixtures of regioisomers due to SCHEME 1 | Synthesis of proline transport inhibitors. the [3,3] sigmatropic rearrangement (Winstein rearrangement) (Gagneuz et al., 1969). That was the case of geranyl, farnesyl and phytylazide, a mixture of primary:tertiary (8:1), being the primary as 1.6:1 (E:Z) ratio (Porta et al., 2017b).
Having prepared the pool of azides, we continued with the synthesis of a focused library of inhibitors through click chemistry. Reactions were conducted in a parallel solution synthesis setup under copper (II) sulfate catalytic conditions in water:t-BuOH (1:1), using sodium ascorbate as a reductant (Rostovtsev et al., 2002;Labadie et al., 2011;Porta et al., 2017a). In general, reactions needed an excess of azides for completion and a reaction time was 18 h. All the products have 1,4-substitution on the 1,2,3,-triazol as expected, based on the original description of this methodology and our previous work (Porta et al., 2014(Porta et al., , 2017a. Reactions with aliphatic and benzylic azides produced a single product, with yields that are slightly better for the last ones. The reaction with the mixture of geranyl azides generated 3i, which was identified as an inseparable mixture of E and Z isomers ( 1 H NMR, 1.5:1), in accordance with our previous results (Porta et al., 2017a). When farnesyl azide was used, a mixture of regioisomers was also obtained with the same ratio, but in this case they were separable (Figure 3). When phytylazide was used the same regioisomers were isolated after the reaction with N-propargyl methyl prolinate 1, presenting a higher E:Z ratio (1.8:1). Extensive purification work allowed only the isolation of the E-isomer of a pure compound (3p-E, Figure 3), leaving the remaining product 3q as a mixture E:Z.
Final products 3a-3q, presented in Table 1, were obtained with an average 67 % yield after purification and were completely characterized by 1D and 2D NMR and ESI-HRMS.
In order to evaluate the biological activity of the prepared collection we decided to initially determine the activity on T. cruzi epimastigotes. Then, to validate the L-proline transporter as the molecular target, the intracellular concentration of the amino acid was measured in competition assays with compounds that shown the best antiparasitic activity. Finally, the cytotoxicity of the selected candidates was evaluated in African green monkey kidney epithelial (VERO) cells to estimate the selectivity toward the parasite.

In vitro Activity Against T. cruzi Epimastigotes
The compounds collection was assayed against T. cruzi epimastigotes (CL strain clone 14) (Brener and Chiari, 1965) at a maximum concentration of 100 µM. Eight compounds of the total list did not affect the parasite growth at that concentration (3a, 3b, 3c, 3d, 3e, 3f, 3h, 3m), being considered inactive. (Table 1) The naphtyl derivative 3g and the eicosanyl analog 3l have an IC 50 of 100 µM. A second group of active members of the collection includes the remaining prenyl derivatives 3n, 3o,  Table 1) The remaining group contains the aliphatic derivatives 3i, 3j and 3k. Those compounds have IC 50 s below 40 µM, being the most active members the collection. The decyl and the oleyl derivatives 3i and 3k have similar activity (IC 50 of 38.97 and 35.06 µM, respectively), while the cetyl analog 3j has a IC 50 of 24.07 µM, considerably slower than the other two (Table 1).

In vitro Cytotoxicity Assay on Vero Cells
Vero cells are well established model to test cytotoxicity in vitro because it is an aneuploid and a continuous cell linage. Initially the library was screened at a fix concentration of 4.75 µg/mL and none of the analogs showed cytotoxic activity. The most active analogs of the series, compounds 3i, 3j, 3k, and 3n, were submitted to a further analysis to determine their IC 50 . Compound 3j was not soluble at 50 µM which made not possible to calculate its IC 50 . Analogs 3i, 3k, and 3n were soluble in a concentration range allowing to perform the assay, showing IC 50 of 43 µM, 17 µM and 14 µM, respectively.

Proline Transport Assay
In order to obtain a deeper insight into the molecular mechanism of the most active compounds, we performed a proline transport competition assay. The compounds with an IC 50 lower than 50 µM (3i, 3j, 3k, and 3n) were selected to perform the proline uptake assay aiming to determine the transport inhibition (Magdaleno et al., 2009). The analogs were assayed on T. cruzi epimastigotes incubated with proline at the transporter Km concentration (0.31 mM), and the analogs were assayed at concentration 10-fold higher (3.1 mM). Surprisingly, we observed that the compounds with higher activity on T. cruzi epimastigotes (lower IC 50 values) showed inhibition but they were not strong enough when compared to analogs with lower antiparasitic activity (higher IC 50 values). As can be seen on Figure 4, compounds 3i (T. cruzi epimastigotes IC 50 = 38 µM) and 3n (T. cruzi epimastigotes IC 50 = 49 µM) showed a proline transport inhibition higher than 75% being more active than the analog 3j that only produced an inhibition of 20%. The oleyl derivate (3k), with an unsaturated fatty tail, has a similar IC 50 on T. cruzi epimastigotes compared to the cetyl analog (3j) showing no inhibition in terms of proline uptake ( Table 2).

DISCUSSION
The selective inhibition of transporters has been proposed as a valuable target to develop new medications against different pathologies including neurological disorders (Qosa et al., 2016) and parasitic diseases (Sayé et al., 2019). The specific inhibition of neuronal glycine, alanine, serine, and cysteine transporters have been studied as molecular targets for new treatment of schizophrenia (Pinard et al., 2010;Schneider et al., 2012; Carland et al., 2014). A high-affinity transporter for proline has been identified, providing an important evidence for proline as a neurotransmitter (Hauptmann et al., 1983). A high throughput screening campaign for high affinity proline transporter inhibitors resulted in the identification of the selective inhibitor LP-403812 (Yu et al., 2009). Summarizing, the participation of metabolites uptake in critical processes in health and diseases has been well demonstrated. A systematic design pipeline for molecules targeting molecular transporters has not been properly explored. The discovery of most of transporters inhibitors happened by chance or through large HTS campaigns. Burns et al. (2009) pioneered a work in this sense proposing a polyamine-fatty acid conjugate as a polyamine transporter inhibitor. In their rational, the polyamine portion is recognized by the transporter and the fatty acid interacts with the membrane, blocking the polyamine entrance. The newly designed inhibitors (L-Lyz(C18-Acylspermine) combined with DMFO display selective antitumoral activity. All the inhibitors mentioned before contained an amino acid portion, or an amino acid mimic, that is recognized by the transporter as a common feature (Figure 5).
With those precedents in mind, we proposed/presented here a general model for aminoacid transporter inhibitors. Our model  included an aminoacid portion for recognition, linked to a variable region with different lipophilic and bulk substituents. We identified the 1,2,3-triazole as a proper linker for our transport inhibitors. The neutral nature of this heterocycle has properly suited the requirements for bioconjugation, protein labeling and immobilization (Gauchet et al., 2006;Mckay and Finn, 2014) and for combining different pharmacophores to make hybrid compounds or chimeras. The Cu(I) azide-alkyne reaction, the quintessence of click chemistry, had a strong impact in many research areas including Medicinal Chemistry (Tron et al., 2008;Agalave et al., 2011). The easy reaction conditions has made this methodology very useful to rapidly prepare libraries of compounds, including antiparasitic drug candidates (Carvalho et al., 2010;Hamann et al., 2014;Porta et al., 2017a).
Using 1,2,3-triazole as linker, we prepared products introducing substituents with proper membrane anchor properties. Fatty acids and isoprenyl chains are selectively introduced in proteins by post translational modifications and served as a mediator for membrane association increasing their molecular hydrophobicity (Hannoush and Sun, 2010). Looking to produce similar anchoring properties, isoprenyl and linear long alkyl chain were selected as some of the 1,2,3triazole substituents, introduced as azide on the heterocycle. Additionally, azides with bulky substituent were also used looking to obstruct the transporter. A library of 17 compounds was finally prepared including two alkyl ester, five alicyclic and aryl, five alkyl and five prenyl derivatives.
The proline uptake systems have similar biochemical characteristics in epimastigotes and the mammalian T. cruzi stages (Silber et al., 2002;Tonelli et al., 2004). It has been demonstrated that these stages are sensitive to proline availability (Tonelli et al., 2004) and uptake (Magdaleno et al., 2009). With those precedents in mind, we decided to evaluate the activity on epimastigotes. The results of the inhibition on T. cruzi epimastigotes shown that half of the compounds prepared did not show activity at the maximum concentration tested of 100 µM. Between the inactive analogs, 3a and 3b holds the shortest substituents of the library and have esters on the side-chain. Being the most polar substituent of the series and susceptible to hydrolysis, seems to be a possible explanation for the lack of activity. Analogs 3c, 3d, 3e, and 3f, contain non-polar cyclic substituents either arylic or alicyclic, were also inactive. Interestingly, the naphtyl derivative 3g, the bulkier member of the collection, has an IC 50 of 100 µM. Together those results shown that polycyclic bulky substituents could be required to improve the activity. Nevertheless, the fact that the aromatic analogs were inactive discouraged the idea that π-stacking interactions are involved on the binding to the molecular target.
As was mentioned before, fatty acids and isoprenyl chains contribute as mediators for protein membrane anchoring, therefore, ten analogs were included with those substituents. The hypothesis that this kind of substituents contribute to improve the activity, seems to be validated because only the analogs holding the shorter substituents, 3h (R=octyl) and 3m (R=geranyl) were inactive. A detailed look at the aliphatic derivatives' activities did not show a direct correlation between the IC 50 and the chain length. The activity increases from octyl 3i to decyl 3h derivative (IC 50 >100 and 38.27 mM, respectively), being the cetyl analog 3j, the most active (IC 50 24.54 mM). Then, the activity decreases to 35.06 mM for the oleyl derivative 3k and 100 mM for the eicosanyl analog 3l. A similar behavior is observed for the prenyl derivatives, but in this case the difference is less pronounced. Moving from the geranyl derivative 3h to the farnesyl analogs the activity increase to 48.32 mM for the E-isomer 3n that is slightly more active than the Z-isomer 3o (59.60 mM). Then, as happened with the alkylated analogs, the activity decreases for the longer member of the family, the phytyl derivatives 3p (E-isomer) and 3q (E/Z mixture), with IC 50 s 69.75 and 48.27 mM, respectively.
A tendency can be visualized when the log IC 50 was plotted against the carbon chain length and the aliphatic and the prenylated are separately correlated (Figure 6). One interesting outcome of this chart is that both curves have their minimum around 14 carbon atoms, that appears to be the optimal chain length for the activity.
These comparisons of the compounds have shown that most of them are considerable more active than L-thiazolidine-4carboxylic acid (T4C), the only reported antichagasic proline derivative, with an IC 50 of 890 µM on T. cruzi epimastigotes (Magdaleno et al., 2009). Analogs 3j, 3i, 3k, 3n were 36, 23, 22, and 18 times more active than T4C. That marked difference on the activity highlight the importance of the proline ring decoration on the antiparasitic activity.
The cytotoxicity of most active compounds (analogs 3i, 3k, and 3n) were tested in cultured monkey kidney Vero cells to estimate the selectivity toward the parasite. The IC 50 were 43 µM, 17 µM and 14 µM, respectively, being an adequate concentration window for future studies in intracellular stages. The selectivity index, (calculated as IC 50 Vero cells / IC 50 T. cruzi epimastigotes) were 1.13, 0.43 and 0.25 for compounds 3i, 3k, and 3n, respectively. This numbers shown a similar susceptibility to the mentioned compounds. The toxicity displayed may be linked to the proline transport inhibition, but that hypothesis must be validated. These results are not promising at this point to propose that compounds of our collection could be good candidates for anti-chagasic drug development. Nevertheless, the applied strategy settles the basis to design inhibitors against T. cruzi biological targets in both, the insect, and the mammalian stages.
Finally, to validate the L-proline transporter as the molecular target, the intracellular concentration of the amino acid was measured in competition assays with compounds that shown the best antiparasitic activity. Analogs 3i, 3k, and 3n were assayed. Interestingly, the proline uptake inhibition of those compounds did not follow the antiparasitic activity. The first hypothesis to explain that behavior was based on the presence of unsaturations on fatty tail. The difference between 3i and 3k relay on the unsaturation on the oleyl chain of the last. The structure of 3k shows a twisted conformation produced by the C9-double bond that also restricts the rotation around the neighbor bonds. Furthermore, the tail length of 3i and 3k differ in 6 carbon and the most active is the decyl analog (3i) in terms of transport inhibition, matching the behavior displayed in Figure 6. The E-farnesyl derivative 3n has a trimethyl substituted side-chain that is twelve carbon long, with three double bonds. Interestingly, this analog produced a complete inhibition of the proline uptake but is the less active of this series against T. cruzi.
( Table 2) The fact that this analog was a superior inhibitor of the transporter could be the result of a better binding. Also, it could be attributed to the markedly different conformation of the analog due to the restricted rotation of the isoprenyl chain. Those restrictions should contribute to block the transporter once the proline region is recognized (Sayé et al., 2017). The inhibition of proline uptake by analogs 3i and 3n, resulted considerably more active than that inhibition by T4C (Magdaleno et al., 2009). Those differences were clearly related with the N1-alylated-1,2,3triazolyl chain introduced on the proline.

CONCLUSIONS
In the present study, a strategy to design amino acids transport inhibitors was proposed. The uptake blocker is composed by a recognition motif, a linker and a bulky substituent or a membrane interacting portion. A set of seventeen 1,5-subtituted-1,2,3-triazole derivatives of methyl prolinate were prepared to validate the design. They were initially assayed against T. cruzi epimastigotes showing comparable potency than the control drug benznidazole. The antiparasitic activity profile of the series allowed us to establish a well-defined structural-activity relationship were the nature of the side-chain play a critical role. In order to validate the design, the inhibition of the proline uptake was studied with the analogs 3i, 3j, 3k, and 3n that displayed the best antiparasitic activity. The analogs with 3i (R=decyl) and 3n (R=E-farnesyl) produced a markedly reduction of the internalized proline. Those studies are strong evidence to validate our design of the transporter inhibitor that also linked the antiparasitic activity with the proline uptake.
The proline uptake has been explored as target for Chagas disease by drug repurposing. Using that approach, crystal violet has been identified as an interesting candidate (Sayé et al., 2020).
Our approach is the first report of new compounds that have been design, prepared and validated as proline uptake blockers with antiparasitic activity. Unfortunately, the most active products displayed low selectivity toward the parasite, not being good candidates for future development as antichagasic drugs. Nevertheless, a complete study on the other T. cruzi life cycle stages could confirm or discard that hypothesis. The validated design of aminoacid transport inhibitor should open new applications on the study of physiological role of amino acid transporters in the central nervous system.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

AUTHOR CONTRIBUTIONS
LF and EP-Z were responsible for the synthesis, purification and structural characterization of all the products. LF and MB performed the antiproliferative assays against Trypanosoma cruzi, carried out the proline transport inhibition assays, and interpreted the results. LF and LP performed the cytotoxicity assays on Vero cells. GL wrote most of the manuscript and supervised the study. GL, AS, and JC were the project leaders organizing and guiding experiments. All authors contributed to refining the manuscript and approved the final manuscript.