Synthesis, in silico Study and Antileishmanial Evaluation of New Selenides Derived from 7-Chloro-quinoline and N-Phenylacetamides

Departamento de Química, Universidade Federal da Paraíba, 58051-900 João Pessoa-PB, Brazil Centro de Educação e Saúde, Universidade Federal de Campina Grande, 58175-000 Cuité-PB, Brazil Departamento de Biologia Molecular e Celular, Universidade Federal da Paraíba, 58051-900 João Pessoa-PB, Brazil Programa de Pós-Graduação em Produtos Naturais e Sintéticos Bioativos, Universidade Federal da Paraíba, 58051-900 João Pessoa-PB, Brazil


Introduction
Leishmaniasis is considered to be a neglected disease, caused by the intracellular invasion of protozoans belonging to the Leishmania genus, which are transmitted by sandfly bites. Leishmaniasis is considered to be an endemic disease in 98 countries, with more than 350 million people at risk. 1-3 The development of new drugs represents a major challenge for medicinal chemistry, with the aim of identifying more efficient and less toxic drugs than those that are currently used (pentavalent antimonials and amphotericin B). Recent research has suggested that organic selenium-based compounds may play a fundamental role in combating this parasitosis, in addition to other parasitic diseases, such as malaria and yellow fever. [4][5][6][7][8] These studies postulate that selenic compounds may represent starting points for the design of new drugs to treat Leishmania infections.
The medications that are currently available for the treatment of visceral leishmaniasis (VL) and cutaneous leishmaniasis (CL) include antimonial drugs (SbV), amphotericin B (AMB), and miltefosine (MIL), which all display high levels of toxicity and/or require long-lasting treatment regimens. 9 In addition, except for AMB, these drugs must be administered parenterally. 10 In the last decade, resistance to SbV has increased, primarily due to low rates of compliance with the treatment schedule. In areas where parasites have become resistant to SbV, AMB is commonly used because it is more active; however, AMB treatment requires a month of hospitalization to monitor kidney function. MIL is highly active for the treatment of VL and has good tolerance, but its potential teratogenicity remains a risk. In addition, these drugs are costly. Therefore, research has sought to identify new treatments for leishmaniasis that are safer, cheaper, and easier to administer than the currently available drugs.
Recently, enzymes involved in the prevention of oxidative damage have been reported as interesting targets for new antileishmanial drugs. 11 Substances with antienzymatic activities, such as organoselenic compounds, have been considered of particular importance. 12 Some anti-cancer drugs, with anti-proliferative activities, have also been shown to be effective for combating protozoa, and several selenium compounds have also been found to display potent anticancer activities. Thus, active compounds against leishmaniasis may exist among current antitumor drugs. 13 Our research group has performed several studies 7,8,14 that have been successful in identifying new candidates for antileishmanial drug development, with the assistance of computational tools to advance the screening process. This study aimed to perform a virtual screen (using both ligand-based and structure-based methodologies) on two series of selenide-derived compounds (one series containing derivatives of 7-amino-quinoline and another series containing derivatives of N-phenylacetamides chlorides). Based on the screening results, the compounds that were considered most likely to be active were synthesized. After their structures were determined using conventional methods, the synthesized compounds were subjected to in vitro tests against Leishmania amazonensis cells, to verify their antileishmanial potential.

Results and Discussion
In silico study

Ligand-based approach
Our research group has previously and successfully used a virtual screening (VS) methodology for the selection of molecules with potential leishmanicidal activity. 7,8 The methodology uses a combination of ligand-based (LB) and structure-based (SB) tools, which are based on the ligand and receptor structures, respectively. The selenide banks 4a-4n and 7a-7n (Scheme 1) were screened, using our VS model, which compares the structures of the new molecules against databases containing 722 molecules with known activity against promastigote forms of L. amazonensis and 818 molecules with known activity against promastigote forms of L. donovani. For each molecule, 128 molecular descriptors were calculated using the Volsurf program (v 1.0.7), 15 which were then transferred to the Konstanz Information Miner (KNIME) program (v. 3.4.0) 16 to build a random forest (RF) prediction model. 7,8 Then, the probabilities of activity against L. amazonensis and L. donovani were calculated for each compound in 4a-4n and 7a-7n, and the results are shown in Table 1. Compounds 4f, 4g, 4h, 4l, 4m, 7j, and 7n were predicted as being active (A), indicating a probability (p) of activity above 50% against at least one Leishmania species. These molecules were selected for synthesis, and the synthesized molecules were sent for in vitro testing against L. amazonensis, which was the species available in our laboratory and is one of the most prevalent species in Brazil. 17 Docking study which are available at the Protein Data Bank (PDB), with the entire selenide bank, because the isoforms of these enzymes are expressed by Leishmania species and are not in humans, so molecules selective for these targets can be selective for these pathogens and have a lower probability of causing side effects. The combined approach uses the data from the LB study and the data from the SB study in the same formula, refining the probabilities that a molecule is active against a specific target. 7,8 The data for the enzymes used, the interaction energies, and the results of the combined approach are described in Tables S1 and S2 (Supplementary Information section), and Table 2, respectively. Table S2 (Supplementary Information section) shows that compounds 4g and 7n had the lowest interaction  energies for the enzymes used in our docking study. The most favorable interactions were identified between these two compounds and the NMT and OASS enzymes. These results corroborated the results of the in vitro study, in which compounds 7n and 4g were found to have the lowest half-maximal minimum inhibitory concentrations (IC 50 ) for these enzymes (Table 1), suggesting that these compounds may be able to exert leishmanicidal activity by inhibiting these enzymes; however, additional studies remain necessary to clearly elucidate the mechanism of action. Prediction of absorption, distribution, metabolism, excretion, and toxicology (ADMET) properties Pharmacokinetic parameters can be estimated to predict which substances are likely to present properties of pharmacological interest. The use of these parameters to predict biological properties that can interfere with absorption, distribution, metabolism, excretion and toxicology (ADMET) have been applied in recent years. 18,19 Determining these properties can provide information regarding the permeability and the concentrations of certain compounds for therapeutic targets and the consequent elimination from the organism. 18,19 ADMET parameters can be verified by in silico studies, based on the calculations of physicochemical properties, such as lipophilicity (clogP), solubility in water (logS), and molecular weight (MW).
In this study, an in silico analysis of the selenide derivatives was performed to verify the Lipinski's parameters, using the software OSIRIS Property Explorer 20 and SwissADME. 21 In addition, other parameters such as rotating bonds (Rb), the topological polar surface area (TPSA), absorption percentage (ABS), and drug score, were also included in the study, as these represent important parameters that are considered when designing and evaluating novel drug candidates. The absorption percentage was calculated using the equation ABS (%) = 109 -(0.345 × TPSA), as described by Zhao et al. 22 The values calculated in this study are shown in Table 3.
The in silico results, described in Table 3, showed that the selected compounds, which were identified as most likely to be active in the combined approach, followed Lipinski's rules which included: molar mass ≤ 500 g mol -1 , log P ≤ 5, number of hydrogen acceptors ≤ 10 (determined as a function of the number of N or O atoms in the molecule), and number of hydrogen bonding donors ≤ 5 (represented as a function of the number of NH or OH groups in the molecule). Only compound 4l registered a subtle violation, for the MW parameter, which is permitted by the rule. Predictions of oral absorption percentages were also calculated from the TPSAs, and the results ranged from 57.30% (4g) to 98.96% (7j), indicating that the derivatives present good permeability across the plasma membrane of cells. When analyzing the values of the drug score (DS), which combines clogP, logS, MW, and toxicity risks for a given species, variations from 0.12 (4l) to 0.41 (4f) were observed, indicating that these compounds have the potential for new drug development. The DS is calculated as a numerical value ranging from 0.0 to 1.0. 18,19 Chemistry The synthesis of the target molecules was performed according to a procedure adapted from Lira et al. 23 and Souza et al.; 14 thus, seven unheard-of selenides (4f, 4g,  4h, 4l, 4m, 7j, and 7n) were obtained, with good yields, derived from 7-amino-quinoline and N-phenylacetamides chlorides, as described in Scheme 1.
The first stage was the preparation of nucleophilic selenium species (2). The synthesis process used elemental selenium, in its solid, amorphous form (Se 0 , black powder), which is the most abundant form of selenium in nature. Selenium was reduced through a reaction with sodium boron hydride (NaBH 4 ), which produces an intense release of hydrogen gas and the reactive species of interest (NaHSe), in the presence of ethanol. 24 The reaction was considered complete after the total consumption of selenium. The second stage was performed as a "one-pot" reaction. Dimethylformamide (DMF) was added to the reaction medium, followed by the respective phenylacetamides chlorides (3f, 3g, 3h, 3l, and 3m), under stirring conditions and at room temperature, which resulted in the formation of the respective bis-selenides (4f, 4g, 4h, 4l, and 4m), with yields between 70-80%.
The bis-selenide structures were confirmed using infrared (IR), 1 H, 13 C nuclear magnetic resonance (NMR) spectroscopic techniques and high resolution mass spectromety (HRMS). Analyzing the NMR spectra of the 4f compound, taking into account its 3f slapper, a signal shift was observed, referring to the H-6 hydrogens that were previously observed at 4.90 ppm and later at 3.38 ppm, due to the replacement of the chlorine element by selenium. A similar trend was observed for the entire series of compounds. The other signals undergo small modifications due to the new interactions, including a singlet at 10.02 ppm, referring to hydrogen linked to nitrogen, a set of doublets at 6.88 and 7.49 ppm, referring to aromatic hydrogens. As the sub-substitution of this compound and the methoxy group, a singlet referring to CH 3 hydrogens is presented at 3.71 ppm.
The same shielding effect of selenium was observed on the 13 C NMR spectra, where the C-6 carbon moves from 43.0 ppm (3f) to the most shielded region, at  (7j and 7n), the same NaHSe production procedure was performed, followed by the addition of 3,7-dichloro-quinoline, which is a reaction that has not been described in the literature, giving rise to intermediate 6. The reaction proceeded with the addition of N-phenylacetamides chlorides, generating the compounds 7j and 7n, with yields of 64.5 and 71.8%, respectively.
The selenide structures were confirmed using 1 H, 13  In the high field region, we have the signal referring to carbonyl carbon that is present in the lower field region at 169.3 ppm (C-12), the other signals in the low field region are identified as C-11 at 39.0 ppm and C-7 at 37.6 ppm. These measurements were taken relative to the percussion of the final molecule, and these molecules have already been elucidated in the literature. The molecular mass of the compounds, 4f, 4g, 4h, 4l, 4m, 7j and 7n, were confirmed by spectra copying masses of high resolution.

Biological study
After using the selenide screening model, the compounds that were identified as likely to be active, with a probability above 50% for at least one of the species, were selected for in vitro tests against the promastigote forms of L. amazonensis (IFLA/BR/1967/PH8), which are species available in our laboratory . Compounds 4f, 4g,  4h, 4l, 4m, 7j, and 7n were selected, and all were found to be effectively active against L. amazonensis cells, with IC 50 values ranging from 5.67-221.6 µg mL -1 (Table 1). In particular, compounds 7n and 4g demonstrated IC 50 values of 5.67 and 10.81 µg mL -1 , respectively. After the good results against L. amazonensis, cytotoxicity tests were also performed against peripheral blood mononuclear cells (PBMCs), to determine whether the compounds were selective for parasitic cells. All compounds were found to be selective, with selectivity index (SI) values ranging from 70.5-1.8. Again, compounds 7n and 4g were also the most selective, in addition to being the most active, showing very promising potential as new drugs.

Conclusions
Our in silico study used a combined approach, based on the structures of both the ligand and the receptor, which identified the molecules 4f, 4g, 4h, 4l, 4m, 7j, and 7n as being potentially active against at least one of the species L. donovani and L. amazonensis. These molecules were also examined using Lipinski's filters and did not violate any of the rules, suggesting that they have potential as new drugs. These seven compounds, which are new molecules, were synthesized, and their structures were confirmed by IR, 1 H, and 13 C NMR. Subsequently, they were subjected to cytotoxicity tests against L. amazonensis cells and PBMCs, to calculate SI values. All compounds were shown to be active and selective, especially 7n and 4g, which presented IC 50 values of 5.67 and 10.81 µg mL -1 , respectively, against L. amazonensis and SI values of 70.5 and 37.0, when cytotoxicity against L. amazonensis was compared with toxicity against PBMCs. For the two most active molecules, their interaction energies were analyzed in detail, through molecular docking studies, with four Leishmania enzymes. The studied molecules showed better anchorage profiles with NMT and OASS, suggesting that these molecules may exert their leishmanicidal effects by inhibiting these enzymes; however, further studies are necessary to unequivocally elucidate the mechanisms of action for these molecules.

In silico study
The VS methodology (LB and SB approaches) used the same databases for L. donovani and L. amazonensis, in addition to the same target enzymes, that were used in previous studies performed by our group, to screen selenides with active potential. 7,8 Chemistry All reagents and solvents were purchased from commercial sources (Sigma-Aldrich, São Paulo, Brazil) and used without further purification. The progress of the reactions was monitored by thin-layer chromatography (TLC) on silica gel plates. The purification of the compounds was performed by recrystallization in ethanol and confirmed by determining the melting ranges of each compound on an MQAPF-3 brand hotplate. The IR spectra were obtained on a Shimadzu model IR Prestige-21 FTIR spectrometer, using attenuated total reflection. 1 H and 13 C NMR spectra were obtained with a Bruker Avance Ultrashield TM (200 MHz for 1 H and 50 MHz for 13 C). Deuterated chloroform (CDCl 3 ) and deuterated dimethyl sulfoxide (DMSO) were used as solvents, and tetramethylsilane (TMS) was used as the internal standard. Chemical shifts (d) were measured in parts per million (ppm), and the coupling constants (J) were measured in hertz (Hz). High resolution mass (HRMS) analysis was obtained with time-of-flight (TOF) mass spectrometer (microTOFII-Bruker) instrument using electron impact ionization (EI) and reported as m/z (relative intensity) for the molecular ion [M] and reporting the molecular ion [M + H] + .
General procedure for the preparation of N-phenylacetamide chlorides In a 50 mL flask, 20 mmol of p-substituted arylamine and 24 mmol of triethylamine (Et 3 N) were added to 20 mL dichloromethane, under constant stirring, at a temperature of 0 °C, which was controlled by a crushed ice bath. After a brief period, 24 mmol of 2-chloro-acetyl/3-chloropropyl chloride was added slowly. Then, the ice bath was removed, and the reaction remained under constant stirring and at room temperature for 24 h. At the end of the reaction, the solvent was removed, under reduced pressure. The resulting precipitate was washed with distilled water, vacuum filtered, and recrystallized in a suitable ethanol/water solution. This procedure yielded a crystalline form.  An amount equal to 20 mL of the solid mixture, consisting of 2.5 mmol elemental selenium and 5.0 mmol NaBH 4 , was placed in a flask. The mixture was stirred continuously until the mixture became homogeneous at room temperature. Next, 5.0 mmol ethanol was added, dropwise, followed by the dropwise addition of 5 mL DMF. During the course of this reaction, the color of the mixture abruptly changed to a dark color, and the reaction was left to stand for 5 min. Then, 5.0 mmol 2-chloro-N-phenylacetamide was dissolved in 5 mL DMF and added to the reaction mixture, which remained at room temperature under constant stirring. For the purification process, 20 mL distilled water was added, which promoted the formation of a precipitate, which was filtered and recrystallized with an appropriate solvent. This procedure yielded a crystalline form (white crystals with a fibrous appearance).

Biological activity
Obtaining PBMC cells For cell isolation, a discontinuous Percoll gradient technique was employed. 7,25,26 Peripheral blood mononuclear cells (PBMCs) were obtained from human peripheral blood and stored in vacuum tubes containing an anticoagulant (ethylenediaminetetraacetic acid (EDTA)). Whole-blood samples were first diluted in phosphate-buffered saline (PBS), at a ratio of 3 mL blood in 5 mL PBS (3:5). Commercial Percoll was diluted in PBS to obtain 40, 50, and 70% solutions, and the pH was adjusted to 7.4. Subsequently, 3 mL of each concentration of Percoll (from the highest to the lowest density) was deposited in 15 mL polypropylene plastic tubes, resulting in a 9 mL total gradient. Aliquots containing 2 mL blood were deposited on top of the discontinuous gradient and centrifuged for 30 min, at 2,000 rpm at room temperature. After centrifugation, the PBMC cloud was recovered and transferred to a new tube containing 11 mL PBS and centrifuged again, for 20 min at 1000 rpm. The pellet was resuspended in Roswell Park Memorial Institute (RPMI) 1640 medium, containing 20% fetal bovine serum and 1% penicillin/streptomycin and amphotericin. Then, cell counting and viability determination were performed.