Introduction

Leishmaniasis is a chronic intracellular infectious disease caused by the protozoan parasite of the genus Leishmania through the bite of infected phlebotomine sand fly [1, 2]. The disease is endemic in large areas of the tropics, subtropics and Mediterranean basin, spanning more than 98 countries [3]. About 350 million people are susceptible to the disease [4], which is classified as one of the most important neglected tropical diseases [5]. Leishmania parasites produce a broad spectrum of clinical manifestations in humans and other mammals, ranging from asymptomatic infection to fatal disease [6], including visceral, cutaneous and mucocutaneous symptoms, depending on complex interactions between the parasite and host’s immune response [7].

Leishmaniases are diseases that have been neglected by the pharmaceutical industry. There is no human vaccine; and the drugs available are limited because of the high cost, toxicity, side effects and increasing signs of resistance [4, 8,9,10]. The development of new, inexpensive and safe drugs for the treatment of leishmaniasis remains a great challenge for researchers worldwide [11, 12].

Parallel, silver nanoparticles belong to the category of strong antimicrobial agents [13] and act as a growth promoter [14] as well as an immune booster [15] at limited doses, and hence, are used for treating wounds and burns [16]. Studies show that internalized Ag nanoparticles in human cells can be translocated to target organelles to elicit several biological effects, including alterations in cellular morphology, oxidative stress, DNA damage, inflammation, genotoxicity and mitochondrial alterations dysfunction, with subsequent apoptotic or necrotic cell death [17, 18]. It was also reported that silver nanoparticles used as an antibacterial agent appeared to provide protection against the bacterial challenge [19].

Beside the interest in silver nanoparticles, the potential medical applications of silver complexes have become an active and exciting research field in biological and bioinorganic chemistry [20]. Antibacterial activity of silver complexes depends on their stability and solubility in aqueous solution, lipophilicity and rate of release of the Ag+ ions [20]; therefore, a careful choice of appropriate ligands is of crucial importance. Among the available ligands to provide new active silver-containing compounds, 1,10-phenanthroline (phen) and its derivatives are a good choice because they have the ability to act on a wide variety of biological processes [21]. In particular, silver complexes incorporating phenanthroline-type ligands have been found to be extremely active in vitro against several pathogens, including Leishmania. For instance, silver compounds of the type [AgL2]NO3 (L = polypyridyl ligands) also were shown to be biologically active against Leishmania mexicana, through interaction with DNA [22]. In our previous report, we have described the synthesis of ternary compounds of the type [{Ag(phen)(μ-tu)}2]X2 (X = NO3; CF3SO3), where tu = thiourea), which were demonstrated to be effective in vitro against promastigote and intracellular amastigotes forms of Leishmania (L.) amazonensis and less toxic to murine peritoneal macrophages [23]. The discovery of the antileishmanial activity of such silver-containing compounds with low toxicity has motivated us to investigate their potential uses in antileishmanial chemotherapy. Thus, we have initiated a research program aiming to identify those parts of the molecular structure that are important to the antileishmanial activity of these types of silver compounds. In this work, we hypothesized that the replacement of thiourea by imidazolidine-2-thione (imzt) would result in more active silver compounds, since imidazolidine-2-thiones are well-known to exhibit a wide range of biological activity [24].

In the current work, the compounds [Ag(phen)(imzt)]NO3(1), [Ag(phen)(imzt)]CF3SO3(2), [Ag(phen)2](BF4)·H2O (3) {phen = 1,10-phenanthroline; imzt = imidazolidine-2-thione} and [Ag2(imzt)6](NO3)2(4) have been synthesized and characterized by elemental analysis, infrared (IR), 1H and 13C NMR spectroscopies, MS/ESI and conductivity measurements. These silver-based compounds were tested against promastigote and amastigote forms of L. (L.) amazonensis and L. (L.) infantum chagasi, and their cytotoxicity against murine peritoneal macrophages in vitro was determined. In addition, the most promising compounds have been selected for further in vivo evaluation and their effects on Leishmania have denoted a possible mechanism of action.

Materials and methods

The compounds

The silver salts (Aldrich), 1,10-phenanthroline monohydrate (Aldrich) and imidazolidine-2-thione (Sigma) were employed without further purification. Acetonitrile (Merck) and chloroform (Tedia) of analytical purity were used as solvents. Deuterated dimethylsulphoxide (Sigma-Aldrich) solution was used for dissolution of the compounds of silver for NMR studies.

C, H, and N analyses were performed on a Leco Instruments LTDA—TruSpec CHNS (Federal University of Alfenas—UNIFAL-MG). Silver crucibles were used with the reference compounds EDTA, cysteine and acetaldehyde, according to the proposed carbon content. Conductivities of the complexes in DMSO solutions (c = 1.0 × 10−3 M) were determined on a Digimed-DM-31 conductometer. Electrospray mass spectrometric analyses were performed on a LCP Fleet—Thermo Scientific electrospray, operating in positive and negative-ion modes (sheath gas flow N2: 8 a.u.; capillary voltage: in positive ion mode 20 V; ion transfer capillary temperature: 250 °C). Sample solutions (0.1 mg cm−3 in CH3OH) were directly injected into the ESI source by use of a syringe pump at a flow rate of 20 mL min−1. Infrared spectra were recorded as KBr pellets on a Spectrum 2000 Perkin Elmer spectrophotometer in the spectral range 4000–400 cm−1 with resolution of 2 cm−1. 1H and 13C NMR spectra were obtained as dmso-d6 solutions, on a Varian INOVA 500 spectrometer.

Synthesis

Compound [Ag(phen)(imzt)](NO3) (1)

1,10-Phenanthroline monohydrate (0.37 mmol; 73.3 mg) dissolved in 10 mL of CH3CN was added to a 10 mL CH3CN solution containing AgNO3 (0.37 mmol; 63.5 mg) leading to a yellow suspension. After stirring for 10 min, imidazolidine-2-thione (0.37 mmol; 38.2 mg) dissolved in 10 mL of CH3CN was added. After stirring for 15 min, the product was isolated by simple filtration, washed with cold chloroform and dried under vacuum. Yield 30%. Anal. Calcd. for C15H14AgN5O3S (%): C: 39.84; H: 3.12; N: 15.49. Found (%): C: 39.76; H: 3.10; N: 15.43. ΛM = 30 Ω−1 cm2 mol−1. FT-IR (KBr): 3161 (νNH), 2895 (νasCH2), 1533 (νCN + δNH), 1512 (νring), 1320 (νsNO2), 1094 (δCH), 823 (δNO2), 722 (γCH), 494 cm−1 (νCS). 1H NMR (ppm; t, triplet; d, doublet; s, singlet): 9.09 (dd, 4.4 Hz, 1.5 Hz, [2H], H1/H1′), 9.05 (s, 2H, NH), 8.71 (dd, 8.3 Hz, 1.4 Hz, [2H], H3/H3′), 8.14 (s, [2H], H4/H4′), 7.97 (m, [2H]. H2/H2′), 3.72 (s, [4H], –CH2–). 13C {1H} NMR (ppm): 178.56 (C=S), 125.39–151.75 (Car), 45.31 (–CH2–).

Compound [Ag(phen)(imzt)](CF3SO3) (2)

1,10-Phenanthroline monohydrate (0.39 mmol; 77.3 mg) dissolved in 10 mL of CH3CN was added to a 10 mL CH3CN solution containing AgCF3SO3 (0.45 mmol; 116.5 mg) affording a white suspension. After stirring for 10 min, imidazolidine-2-thione (0.41 mmol; 42 mg) dissolved in 10 mL of CH3CN was added. After stirring for 15 min, the product was isolated by simple filtration, washed with cold methanol and dried under vacuum. Yield 76%. Anal. Calcd. for C16H14AgF3N4O3S2 (%): C: 35.63; H: 2.62; N: 10.39. Found (%): C: 35.7; H: 2.6; N: 10.2. ΛM = 45.3 Ω−1 cm2 mol−1. FT-IR (KBr): 3258 (νNH), 2895 (νasCH2), 1533 (νCN + δNH), 1513 (νring), 1246 (νSO3), 1224 (νCF3), 1094 (δCH), 1027 (νSO3), 723 (γCH), 498 cm−1 (νCS). 1H NMR (ppm): 9.14 (dd, 4.4 Hz, 1.6 Hz, [2H], H1/H1′), 9.00 (s, 2H, NH), 8.76 (dd, 8.3 Hz, 1.6 Hz, [2H], H3/H3′), 8.18 (s, [2H], H4/H4′), 8.02 (m, [2H]. H2/H2′), 3.72 (s, [4H], –CH2–). 13C {1H} NMR (ppm): 179.05 (C=S), 125.44–151.85 (Car), 45.28 (–CH2–).

Compound [Ag(phen)2](BF4) H2O(3)

1,10-Phenanthroline monohydrate (0.78 mmol; 140.6 mg) dissolved in 10 mL of CH3CN was added to a 10 mL CH3CN solution containing AgBF4 (0.39 mmol; 76 mg) leading to a yellow suspension. After stirring for 15 min, the product was isolated by simple filtration, washed with cold methanol and dried under vacuum. Yield 70%. Anal. Calcd. for C24H18AgBF4N4O (%): C: 50.3; H: 2.9; N: 10.1. Found (%): C: 52.1; H: 3.0; N: 10.2. ΛM = 36.1 Ω−1 cm2 mol−1. FT-IR (KBr): 3393 (νOH), 1510 (νring), 1246 (νSO3), 1224 (νCF3), 1056 (νBF), 1027 (νSO3), 725 (γCH), 522 (νBF). 1H NMR (ppm): 9.14 (dd, 4.5 Hz, 1.5 Hz, [4H], H1/H1′), 8.76 (dd, 8.1 Hz, 1.4 Hz, [4H], H3/H3′), 8.20 (s, [4H], H4/H4′), 8.00 (dd, 8.1 Hz, 1.6 Hz, [4H]. H2/H2′).

Compound [Ag2(imzt)6](NO3)2 (4)

Complex [Ag2(imzt)6](NO3)2(4) was synthesized as previously described [25]. Briefly, a hot solution of AgNO3 (170 mg, 1 mmol) in water (2.5 mL) was added to a hot solution of imidazolidine-2-thione (315 mg, 3 mmol) in water (2.5 mL). Colourless crystals of the product were isolated upon cooling.

Biological assays

Determination of the proliferation curve

Promastigotes (106 mL−1) were added to 25 cm2 flasks containing LIT medium and were kept at 26 °C. On a daily basis, an aliquot was removed, and the number of promastigotes was determined using a Neubauer chamber [26, 27]. The experiments were carried out in triplicate and were repeated independently three times.

Antileishmanial activity against promastigotes

Promastigotes of L. (L.) amazonensis (strain MHOM/BR/71973/M2269) and L. (L.) infantum chagasi (strain MHOM/BR/1972/BH46) were grown on 24-well plates in LIT medium, supplemented with 10.0% (v/v) heat-inactivated foetal bovine serum and 1.0% penicillin (10,000 UI mL−1)/streptomycin (10.0 mg mL−1) (Sigma, USA). Cells were harvested in the log phase, suspended in fresh medium, counted in Neubauer chambers and adjusted to a concentration of 1 × 106 cells mL−1, using 24-wells plates. The compounds (1), (2), (3), (4) and Imzt were added to promastigote cultures (1 × 106 cells mL−1) in the range of 0.10–40.00 µg mL−1, solubilized in dimethyl sulfoxide (DMSO) (0.6%, v/v in all wells) and incubated at 25 °C. After 72 h of incubation, the surviving parasites were counted in a Neubauer’s chamber and compared with controls, with just DMSO in concentration of 0.6% v/v, for the determination of 50.0% inhibitory growth concentration (IC50). All tests were performed in triplicate at three different times and amphotericin B (Sigma) was used as the reference drug.

Antileishmanial activity against amastigotes

Murine peritoneal macrophages were maintained in RPMI 1640 medium (Sigma, USA) supplemented with 10.0% heat-inactivated foetal bovine serum at 37 °C in 5.0% CO2 incubator. These cells were obtained from the peritoneal lavage of Swiss mice, properly approved by the Research Ethics Commission of the Universidade Federal de Alfenas (project number 08/2016) and were performed according to the Guide for the Care and Use of Laboratory Animals [11]. Cells were cultured in 24-well plates chamber on the glass slides of 13 mm (Nunc, USA) in a 8 × 105 cells density per well and infected with late log-phase promastigotes at a ratio of 10:1 (parasite/macrophage) and incubated at 37 °C in 5.0% CO2 incubator for 24 h. Non-phagocytosed promastigotes were removed by washing, and compounds (1), (2), (3), (4) and Imzt solubilized in DMSO (from 0.10 to 40.00 µg mL−1) were administered at the concentration of 0.6% v/v. After 72 h, chamber slides were fixed in absolute methanol, stained with 10.0% Giemsa and examined on an Optical Light Microscope in oil immersion. The percentage of infected cells per well was calculated taking account at least 200 macrophages. The ratio of inhibition (IC50 value) was calculated in comparative to the control only with DMSO. All assays were performed in triplicate at three different times using amphotericin B (Sigma) as the reference drugs.

Cytotoxicity evaluation

Murine peritoneal macrophages were obtained from the peritoneal lavage of Swiss mice, properly approved by the Research Ethics Commission of the Universidade Federal de Alfenas (project number 08/2016) and were performed according to the Guide for the Care and Use of Laboratory Animals [11]. A suspension of 8 × 105 murine peritoneal macrophages, in RPMI 1640 medium, supplemented with 10.0% heat-inactivated foetal bovine serum and 1.0% penicillin (10,000 UI mL−1)/streptomycin (10 mg mL−1) were added to each well in 96-well plates. The plates were incubated in a 5.0% CO2 air mixture at 37 °C to allow for adhesion of the cells. After 24 h, non-adherent cells were removed by washing with the RPMI 1640 medium. Then, several concentrations of compounds and reference drugs, ranging from 3.91 to 500.00 µg mL−1 in DMSO at the final concentration of 0.6% (v/v), were added to the wells containing the cells and the plates were incubated for more 48 h. Non-adherent cells were removed by washing with the RPMI 1640 medium. Afterwards, the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazoliumbromide (MTT) was solubilized in PBS (5.0 µg mL−1) as solvent, and 10 µL was added to RPMI 1640 medium in a final volume of 200.0 µL per well and incubated for 4 h [28]. Then, the medium was removed and 100.0 µL of DMSO was added to each well and homogenized for 15 min. Next, the absorbance of each individual well was calculated at 570 nm according to the following formula (OD represents optical density) using Eq. (1):

$${\text{Inhibition}} = \left( {\frac{{{\text{OD}}_{\text{control}} - {\text{OD}}_{\text{compounds}} }}{{{\text{OD}}_{\text{control}} \times 100}}} \right).$$
(1)

Each experiment was performed in triplicate on three different occasions, and the percentage of viable cells was calculated taking into account the cell culture control (medium + cells + DMSO 0.6% v/v). The 50.0% cytotoxicity concentrations (CC50) were determined and the selectivity index (SI) was established by the ratio between the values of CC50 and IC50 for amastigote forms.

Statistical analysis

The antileishmanial activities of the compounds were expressed as the concentration that inhibits the growth of 50.0% of the protozoan form (promastigote or amastigote). Statistical analysis was performed using non-linear regression to obtain the IC50 and CC50 values (cytotoxic concentration to 50.0% of the macrophages), followed by Tukey test and variance analysis. Differences of IC50 and/or CC50 between standard drug and compounds in question were significant when the p value was less than 0.05.

In vivo assay

Parasites and animals

L. (L.) infantum chagasi (strain MHOM/BR/1972/BH46) promastigotes were maintained in M-199 medium supplemented with 10% calf serum and 0.25% hemin, at 24 °C. Amastigotes were maintained by passaging in female golden hamsters (Mesocricetus auratus). Animals were kept in sterile absorbent material boxes, with food and water ad libitum, and in a natural light/dark cycle. All experimental procedures involving animals were approved by the Research Ethics Commission of the Universidade Federal de Alfenas (project number 08/2016) and were performed according to the Guide for the Care and Use of Laboratory Animals. For experimental infections, the spleens of infected animals were removed and macerated using a tissue grinder and the number of amastigotes was determined as described previously.

In vivo testing of experimental compounds against Leishmania

Female golden hamsters that had been recently weaned (120 g) were infected intraperitoneally with 1 × 107 amastigotes of L. (L.) infantum chagasi (MHOM/BR/1972/BH46) [29] and maintained in sterile absorbent material boxes, with water and food ad libitum. In the chronic phase (around 50 days of infection), animals were divided into four groups (n = 5/group), and subjected for 10 consecutive days to one of the following treatments: 0.5% of carboxymethyl cellulose (vehicle) suspension, administered orally (untreated, or UTG group); 50 mg kg−1 day−1 of Glucantime® (GLU), by intraperitoneal injection (GLU group); 50 mg kg−1 day−1 of imzt and 25 mg kg−1 day−1 of compound (4), administered orally as suspensions in 0.5% of carboxymethyl cellulose. After 10 days of treatment, animals were sacrificed in a CO2 chamber, and a sample of the spleen and the liver (approximately 50 mg) was removed, weighed and used for total RNA extraction to subsequently undergo transformation into cDNA, as previously described [29].

DNA extraction

Standard curves of parasite DNA for use in quantitative real-time PCR (qPCR) experiments were produced as described previously [30]. In the log curve phase, the parasites were harvested, washed twice by centrifugation in phosphate-buffered saline (pH 7.2) at 1000 g for 10 min. A stock solution of 1 × 108 promastigotes of each strain was estimated by microscopic count and was used to prepare the standard curves. The parasite pellets were used to DNA extraction and determined the DNA concentration. Next, eight different concentrations of promastigotes were performed to obtain the points of the standard curve ranging from 1 × 107 to 1 promastigotes. DNA was extracted from samples using the QIAamp DNA extraction Mini Kit (Qiagen), according to the manufacturer’s instructions.

RNA extraction and cDNA synthesis

Fragments of liver and spleen (~ 50 mg; weighed using sterile and disposable surgical material) removed from treated animals were placed in sterile microfuge tubes and frozen immediately at − 80 °C in storage buffer (RNAlater). RNA extraction was performed 24 h after fragment removal, using the RNeasy Mini Kit (Qiagen), according to the manufacturer’s instructions. RNA samples were frozen immediately after extraction. For reverse-transcription into cDNA, 1 μL of dNTPs mix (10 mM) and 1 μL of random primers (3 μg μL−1) were added to 11 μL of RNA sample, and samples were incubated in a thermal cycler for approximately 5 min, at 65 °C. Then, tubes were placed on ice for 20 s, and 2 μL of DTT (100 mM) and 4 μL 5 × buffer (Tris–HCl 250 mM, pH 8.3, containing 375 mM KCl, 15 mM MgCl2) were added, and samples were incubated again in the thermal cycler for 20 s, at 37 °C. Finally, 1 μL (200 U μL−1) M-MLV RT enzyme was added and samples were incubated for 50 min, for cDNA synthesis. The purity of the cDNA sample was confirmed by measuring the absorbance at 260/280 in a NanoDrop ND2000, and sample integrity was verified by agarose gel electrophoresis and PCR. Samples were frozen at − 20 °C for subsequent use in qPCR.

Parasite load estimation by LINJ31 quantitative PCR (qPCR)

Quantitative real time PCR (qPCR) was performed using the TaqMan® probe 5′CCT CCT TGG ACT TTG C3′ (double-labelled with FAM at the 5′-end and a non-fluorescent quencher at the 3′-end), and the primers LINJ31F (5′CCG CGT GCC TGT CG3′) and LINJ31R (5′CCC ACA CAA GGA GCG ACT3′), which amplify L. (L.) infantum chagasi hypothetical protein [31]. Reactions were performed in a 7500 Real Time PCR System (Applied Biosystems), and reaction mixtures contained 3 μL of DNA or cDNA samples (or control samples, see below), 10 μL of 2X TaqMan Universal PCR Master Mix, 1 μL of a mixture of forward (LINJ31F) and reverse (LINJ31R) primers (at a concentration of 18 μM), and 5 μM of the labelled TaqMan®probe, in a final volume of 20 μL. For negative and positive controls, water or DNA extract from L. (L.) infantum chagasi (MHOM/BR/1972/BH46) were added, respectively. The following PCR conditions were used: one step of 50 °C for 2 min, followed by one step of 95 °C for 10 min, and 40 cycles of 95 °C for 15 s and 60 °C for 1 min. The number of parasites per gram of spleen or liver tissue was calculated based on the linear regression data from the standard curve performed with promastigote DNA [29]. Statistical analysis was performed by Student’s t test with Mann–Whitney (unpaired, two tailed) for the significance test (p < 0.05) [30].

Histopathological evaluation of non-infected hamsters

For histological analysis, fragments from liver and kidney collected from non-infected hamsters were fixed in 10% buffered formalin and processed through conventional histological techniques. Sections of 5 μm thick were stained with hematoxylin and eosin (HE) and analysed by a pathologist blinded to the experimental groups. Particular attention was paid to microscopic findings in liver such as steatosis, hydropic degeneration, necrosis, microabscesses, fibrosis and disorganization of hepatic cords; glomerular and tubular degeneration as well necrosis and fibrosis in kidney [32]. All sections were analysed under an optical light microscope (Zeizz Axio Scope A1) equipped with a camera (Axio Cam CC3, Zeizz).

Biological effects on Leishmania

Antiproteolytic activity

Inhibitory activity against rCPB 2.8 The procedure was performed using 1 mL of 100 mmol L−1 sodium acetate buffer (pH 5.5, containing 5 mmol L−1 EDTA, 100 mmol L−1 NaCl, 20% glycerol and 0.01% triton X-100), to which 3 mmol L−1 dithiothreitol (DTT) and enzyme were added. The mixture was left to pre-activate for 10 min, then the fluorogenic substrate Z-FR MCA (7.4 μmol L−1) was added. The enzymatic activity was monitored by substrate hydrolysis at 360 nm excitation and 480 nm emission wavelengths in a spectrofluorimeter. The reaction solution was continuously stirred and the temperature maintained at 37 °C in thermostatically-controlled water bath. The values of the enzymatic activity were calculated by linear regression and expressed as UAF min−1 (arbitrary units of fluorescence per minute). The assays were repeated with two different concentrations (10 and 100 μmol L−1) of each compound, enabling the activity values in UAF min−1 to be determined for each one. The values obtained in the absence of the compound were assumed to represent 100% activity, and the activity values in the presence of the compounds were calculated as a proportion of these values. Enzyme inhibition was expressed as the compound concentration causing a 50% decrease in enzyme activity (IC50 value) [11]. IC50 value was calculated by non-linear regression, based on dose–response curves using (1), (2), (3), (4) and Imzt compound at different concentrations, and the data were analysed with Grafit 5.0 software20 using Eq. (2):

$$y = \left( {\frac{\text{Range}}{{1 + \frac{x}{{{\text{IC}}50}}}}} \right).$$
(2)

In Eq. (2), y is the enzyme activity, x is the inhibitor concentration, and s is a slope factor. The equation assumes that y reduces with increasing x. As reference inhibitors were used E-64 for rCPB2.8 [11].

Determination of hydrogen peroxide release

Promastigotes (108 cells mL−1) of L. (L.) amazonensis at lag (2 days) and log (4 days) phases of the proliferation curve treated or not (control) with compound (4) were incubated in reaction buffer 1X from Amplex Red® kit in the presence of 5 mM succinate, 40 μM digitonin, 0.1 U mL−1 horseradish peroxidase and 25 μM Amplex Red (Molecular Probes®). The fluorescence was monitored at the excitation and emission wavelengths of 571 nm and 585 nm, respectively, at 96-well plate using a Varian Cary Eclipse Fluorescence Spectrophotometer. The quantitative correlation between the fluorescence and the H2O2 released by the cells was determined as previously described [26]. The experiments were carried out in triplicate and were repeated independently three times.

Results and Discussion

Chemistry

The new complexes [Ag(phen)2(imzt)]X {X− = NO3−(1), CF3SO3−(2)} were prepared in acetonitrile by reacting the appropriate AgX salt, 1,10-phenanthroline (phen) and imidazolidine-2-thione (imzt) in a 1:1:1 molar ratio, respectively. Aiming to investigate which ligands were important for activity, the bis-chelated complex [Ag(phen)2]BF4·H2O (3) and the binuclear compound [Ag2(imzt)6](NO3)2(4) [25]. The complexes were air and light stable solids. Elemental analysis data revealed that (1) and (2) have molar ratios Ag+: phen: imzt: X− = 1:1:1:1 (X− = NO3− and CF3SO3−) whereas (3) displayed a 1:2:1 stoichiometry for Ag+: phen: BF4−. The molar conductivity values of (1–3) in DMSO were found between 30 and 46 Ω−1 cm2 mol−1, in agreement with their 1:1 electrolytic behaviour [33]. These findings suggested that these complexes have mononuclear structure in solution, as observed in similar compounds reported by us [23]. The ESI/MS spectra of complexes (1) and (2) showed the expected signals of the mononuclear [Ag(phen)(imzt)]+ ion at m/z 389, together with the presence of the peak at m/z 287 assigned to the [Ag(phen)]+ ion. The ESI/MS spectrum of (3) showed the typical signals of the fragments [Ag(phen)2]+ and [Ag(phen)(OH2)]BF4+ at m/z 467 and 391, respectively. In these cases, the observed isotope patterns were in agreement with those calculated for the formulated ions. Compound [Ag2(imzt)6](NO3)2(4) has been successfully synthesized as described by literature [25].

Spectroscopic evidence provided relevant information on the structures of compounds (1) and (2). The shift to lower frequency of the νC=S band from 516 cm−1 (free imzt) to 494 (1), 498 cm−1(2) suggested the coordination of imzt via S atom [34]. In addition, the displacement to higher frequency of the νCN absorption from 1522 cm−1 (free imzt) to ca. 1533 cm−1 in complexes (1–2) also supported the S-coordination of imzt. The chelating coordination mode of phen is supported by the appearance of its typical bands in the IR spectra of (1–3) at ca. 1590, 1426, and 726 cm−1. The presence of uncoordinated NO3−, CF3SO3− and BF4− ions in (1), (2) and (3), respectively, was also inferred on the basis of IR evidence, confirming their cationic character.

The 1H NMR spectra of (1–2) in DMSO-d6 showed only one set of signals for coordinated phen and imzt, which suggests the presence of a single species in solution. The coordinated phen has equivalent halves due to the appearance of the four signals assigned to the H2/H2′, H3/H3′, H4/H4′and H5/H5′ nuclei. For example, the NMR spectrum of complex (2) displayed two double doublets at 9.14 and 8.76 ppm, which are shifted 0.03 and 0.26 to downfield, respectively, as compared to the free ligand. One multiplet at 8.02 ppm and one singlet at 8.18 ppm experienced a downfield shift of 0.23 and 0.18 ppm, respectively. The integration curves of the 1H NMR signals of complexes (1) and (2) indicated a 1:1 molar ratio of phen: imzt, which is in agreement with the analytical results. The coordination of neutral imzt is supported by the presence of broad NH resonance at ca. 8.9 ppm in (1) and (2), being shifted 1.0 ppm to downfield related to free ligand. Upon coordination, the –CH2– singlet changed from 3.49 (free imzt) to 3.72 ppm in (1) and (2). Such displacements are attributed to the delocalization of π-electrons between nitrogen atoms as a result the strengthening of the CN bond together with the weakening of the CS bond on complex formation [34]. For (3), only the set of signals for coordinated phen was observed. The 13C{1H} NMR spectra of (1) and (2) confirmed the S-coordination mode of imzt. The C=S resonance was displaced from 183.4 ppm (free imzt) to 179 ppm in these complexes, as a result of the formation of Ag–S bond together with the downfield shift of 1.2 ppm of the –CH2– signal.

Taking into account that the results from elemental analysis, conductivity measurements, mass spectrometry and NMR spectroscopy, compounds (1) and (2) may exist in solution as mononuclear species, such as [Ag(phen)(imzt)]+ or [Ag(phen)(imzt)(DMSO)]+. For (3), the existence of the complex ion [Ag(phen)2]+ in solution was observed. Infrared, 1H and 13C NMR spectra and MS/ESI results for the silver (I) compounds (1–3) are available in the Supplementary Material.

Biology

The growth kinetics of the parasites were studied before assessing the in vitro activity of the test compounds to assess extension of the log-phase to ensure that drug efficacy assays were conducted during that period.

In vitro tests are important in biomonitoring studies in the search for active principles associated with the detected antileishmanial activity [35]. Thus, compounds with better antileishmanial activities and selectivity index (SI), ratio between CC50 (macrophages) e IC50 (amastigotes) were selected and these data suggest that alternative treatments may be effective in controlling the disease.

Therefore, all synthesized compounds were initially evaluated in vitro for their antileishmanial activities against promastigote and amastigote forms of L. (L.) amazonensis and L. (L.) infantum chagasi. To identify the selectivity of binding of compounds to the parasites with little or no effect on host cells, the cytotoxicity assay of the compounds in murine peritoneal macrophages was performed to determine CC50. Thus, SI was also determined (Table 1).

Table 1 Antileishmanial activity in vitro against promastigote and amastigote forms of L. (L.) amazonensis and L. (L.) infantum chagasi, cytotoxicity, selectivity index values, of compounds (1), (2), (3), (4) and Imzt compared to Amphotericin B
Full size table

The results showed a dose-dependent inhibition of growth after the incubation time (data not shown). The inhibitory effect of the compounds against the promastigote and amastigote forms showed some different responses. Although stage-specific activity has been reported previously for antileishmanial drugs, this variation in the sensitivity is difficult to interpret as it could be due to differences in the rate of division; or exposure of intracellular and extracellular stages to drugs; or biochemical targets; or drug metabolism [36]. In this work, two Leishmania species were assayed and the results reflect differences in the sensitivity of these parasites to the compounds tested. This fact is not surprising and previous in vitro studies have shown differences in sensitivity of Leishmania species in different classes of drugs [37, 38].

The process of differentiation of Leishmania forms throughout their life cycle has been extensively studied [39]. Significant changes in mRNA expression and translation rates [40] allow the parasites to adjust to the environmental differences according to the phase of the cycle by changing biochemical processes. For example, the transition could be accompanied by an increase in the enzyme activity rates involved with gluconeogenesis, making this path essential for the amastigote form [40]. In addition, Mondal et al., 2014 demonstrated differences regarding the mitochondrial bioenergetics between the amastigote and promastigote forms of L. donovani [41]. Biochemical changes such as those involved with post-translational modifications [42] occur at all stages of differentiation, and comparative analyses from transcriptomic and proteomic data throughout genomic time clearly indicate that the differentiation process is highly regulated and coordinated revised in [43]. In this sense, the involvement of different signaling pathways in differentiation is evident, as well as those prevalent in the amastigote and promastigote forms revised in [43]. All these data confirm that it is not appropriate to extrapolate activities obtained for one form of the parasite to the other. The modulation of cell-mediated response and the cellular and biochemical pathways of amastigotes differ considerably from those of promastigotes, suggesting that the chemotherapeutic potential of antileishmanial drugs depend on their action against amastigotes [11]. Given that the amastigote is the parasitic form responsible for disease, it should constitute the chemotherapeutic target in studies of new antileishmanial agents [44].

Among the compounds tested, compound (4) was the most potent against amastigote forms of L. (L.) amazonensis and L. (L.) infantum chagasi, with IC50 of 1.88 μM and 8.05 μM, respectively; it was also the compound that stood out against the others against promastigote forms. According to these results, the highest concentration tested showed a lower infection rate (data not shown), and the results were comparable to the reference drug in the treatment (Amphotericin B), that is its IC50 showed that its activity was comparable to that of the amphotericin B control drugs (IC50 = 1.30 μM, 4.00 μM, respectively), in most tests. In addition, this compound was found to be the least toxic to human macrophages, with a CC50 of 94.06 μM and more selective folds (SI = 50.00, 11.7, respectively) than the standard drug amphotericin B (CC50 = 27.05 μM, SI = 20.8, 6.7, respectively). Considering SI, expressed as the ratio of cytotoxicity (CC50) and antileishmanial potency (IC50), compound (4) exhibited the best biological profile. Although an SI greater than 10 is necessary to ensure the safety of a drug [45], the results for compound (4) are very important for the development of new anti-leishmania drugs. Analysing the cytotoxicity results, the different concentrations of the compounds used did not generate significant cytotoxic effects, in comparison with the control group (cells without treatment), showing that they were not toxic to murine peritoneal macrophages.

Research demonstrated that Ag nanoparticles can be generated by biological synthesis, where they showed a higher antileishmanial effect and lower toxicity in host cells [46, 47], resulting in inhibition of parasite proliferation. Cytotoxicity was not observed in the treated macrophages, and a reduction in the growth of promastigote and amastigote forms was observed when they were treated with compound (4). Current therapy against leishmaniasis based on Amphotericin B and Glucantime®, has a variety of side effects, in addition to high toxicity [48]. Thus, it is extremely important to search for new compounds that are more effective in the treatment of leishmaniasis and with low toxicity. As described above, we obtained excellent results for compounds containing silver in their structure, with potent antileishmanial activities in vitro, in addition to presenting low toxicities in macrophages, compared to the standard drug.

In this sense, we evaluated the antiparasitic effect of compounds (4) and Imzt on established infection hamsters (60 days post-inoculation with L. (L.) infantum chagasi) through quantification of live amastigotes of L. (L.) infantum chagasi in the tissues of the liver and spleen. Therefore, a quantitative detection of the marker LINJ31 was used [29] (Fig. 1).

Fig. 1
figure 1

Quantification of parasite burden in the spleen and liver from hamsters infected with Leishmania (L.) infantum chagasi, as quantified by PCR for the detection of the LINJ31marker. The numbers of parasites (amastigotes) per gram of tissue were estimated based on a qPCR standard curve using promastigotes. UTG: untreated group (vehicle-treated), GLU: animals treated with Glucantime® (50 mg/kg/day); Imzt: compound imidazolidine-2-thione (50 mg/Kg/day); compound (4) (25 mg/kg/day). n = 5 animals/group. Data are represented as group mean and individual values for each animal. *Statistically similar (p < 0.05)

Full size image

The results showed that the compound (4) reduced the number of amastigotes in the liver and spleen when compared to the untreated group (UTG). In addition, when comparing the effect of compound (4) and Glucantime® on the spleen, no significant difference was observed. In making a correlation between the compounds in question, we can conclude that the insertion of silver and nitrate into the chemical structure improved its activity in both, in vitro and in vivo assays. According to the literature, studies have shown that the addition of the nitrate group in metallic compounds may contribute to increase the antileishmanial activities in vitro in different evolutionary forms of L. (L.) amazonensis [49, 50].

Additionally, different routes of administration can lead to different pharmacological responses, due to variations in the absorption processes, since the absorbed dose is different from that administered [29]. The chemical nature of compounds can lead to different routes of absorption; as an example, compounds in solution can be absorbed faster than those in suspension [51]. This difference in the route of administration may have contributed to the increased efficacy of Glucantime®, also in the liver.

The liver from hamsters daily treated with 25 mg kg−1 day−1 of compound (4) exhibited some histological alterations in areas adjacent to centrilobular vein including disorganization of the hepatic cords, hydropic degeneration and necrotic cells with karyolytic or picnotic nuclei (Fig. 2a, b). Histological examination of kidney after compound (4) treatment revealed degenerative changes in some parenchymatous elements in comparison with untreated control. Atrophy of renal corpuscle and, degeneration and atrophy of lining epithelium in some renal tubules were occasionally noticed in some areas (Fig. 2c, d). However, no sign of significant histological changes such as microabscesses, fibrosis or marked loss of hepatic and/or renal tissue that would indicate a major toxicity were noticed after compound (4)-treatment.

Fig. 2
figure 2

Histological examination of non-infected hamsters liver and kidney after compound (4) treatment. Photomicrographs of livers after intraperitoneal administration of compound (4) at a untreated and b 25 mg/kg/day, respectively. Liver of hamsters daily injected with 25 mg/kg/day of compound (4) shows disorganization of hepatic cords at proximity to centrilobular vein, hydropic degeneration (arrows), and necrotic cells (circle). Sections of untreated (c) and compound (4)-treated (d) hamsters kidneys, respectively. Kidney of compound (4)-daily treated hamsters shows atrophy of renal corpuscle (green arrow), and degeneration and atrophy of some renal tubules (star). Original magnification × 200 (scale bar: 20 μm)

Full size image

Biological effects on Leishmania

Since the effect of compound (4) on different forms of Leishmania has been observed, this fact led us to investigate which changes could be detected in the parasites incubated with the compound in question. Besides the fact that some compounds derived from imidazole exhibited activity of scavengers from reactive oxygen species [52]; and activity in proteases [53]. In this sense, since the compound (4) used in our experiments is an imidazole derivative, the evaluation of H2O2 production could be interesting, as well as the protease activity to be evaluated.

All compounds and the reference protease inhibitor E64 (trans-epoxysuccinyl-l-leucylamido(4-guanidino)butane) were next assayed to determine their inhibitory potential against cysteine proteases rCPB2.8 (Table 2).

Table 2 Quantitative in vitro inhibitory effect of synthetic compounds on cysteine protease rCPB2.8 activity of L. (L) mexicana, compared to E64
Full size table

The L (L.) mexicana assays are performed for not yet obtaining cysteine proteases from other Leishmania species. Therefore, by homology experiments in L. (L.) mexicana are accepted and used to infer the potential/activity against the cysteine proteases [11]. As can be seen, not all compounds presented as potent cysteine protease inhibitors B. The compound (1) was the most potent inhibitor compound against rCPB2.8 with IC50 value of 2.37 μM, but no less active than the E64 standard. Proteases are also key enzymes in the metabolism of proteins, or biologically active peptides [54]. These enzymes are also involved in several adaptation mechanisms for the survival of parasites: regulation of immune response, invasion and damage tissue, differentiation and dissemination of parasite and uptake of essential nutrients to parasite [54, 55]. In our experiments it was possible to observe that, although the compounds tested not present potent activity against the cysteine protease (CPB 2.8), when compared to the E64 standard, it was possible to speculate that this would not be the main mechanism of action presented by these compounds.

Thus, another possibility for the mechanism of action could be related to concentration of reactive oxygen species in the parasite. Therefore, we evaluated the production of hydrogen peroxide in Leishmania promastigotes treated with compound (4).

Leishmania, like other trypanosomatids, possesses a unique mitochondrion with peculiar characteristics, such as the presence of enzymes of exclusive metabolic pathways and a specific arrangement of mitochondrial DNA. During the life cycle of trypanosomatids, their unique mitochondria undergo profound changes, reflecting the adaptation to different environments, such as hypoxia and oxidative stress [55]. In parallel, the mitochondria are involved in the processes of differentiation, proliferation, calcium homeostasis, redox balance, and programmed cell death [56], beside being the main source of reactive oxygen species, such as hydrogen peroxide [57], but not the only one; since Machado et al. (2018) reported that Leishmania species are seen to make ROS through NADPH oxidase enzymes [58]. In this sense, the maintenance of mitochondrial integrity is fundamental for the survival of the parasite, therefore this organelle is a potential target in the development of drugs. The mechanism of action of some drugs, or even of potential drug candidates, involves pathways related to mitochondrial bioenergetics and generation of ROS.

For the survival of Leishmania, as well as other trypanosomatids, maintaining a reduced intracellular environment is necessary. This maintenance is done by the activity of antioxidant enzymes, such as peroxidases [59, 60]. In this sense, the increase in the production of H2O2 signals an increase in the synthesis of these antioxidant enzymes (cytosolic and mitochondrial enzymes) as previously reported [26], which guarantees a reduced intracellular environment and consequently the survival of the parasite [61], thereby avoiding oxidative stress. On the other hand, it was observed that the proliferation of parasites, as an example, Trypanasoma cruzi, also a trypanosomatid such as Leishmania; has a relation in terms of signaling with the concentration of H2O2 [62]. Low concentrations of H2O2 stimulate cell proliferation in T. cruzi [62], in mammalian [63] and yeast cells [64].

In recent years, the understanding of the importance of ROS as a regulator of biological and physiological response has increased [65,66,67], highlighting superoxide and hydrogen peroxide (H2O2) as co-products of mitochondrial respiration. Regarding H2O2, studies have shown that low levels of this reactive species are implicated in a reduction in growth and cell proliferation [68, 69], while high levels may lead to oxidative stress, leading to cell impairment. Researchers observed in T. cruzi, that in low concentrations of H2O2, cell proliferation was stimulated and parasites increased their resistance to sublethal doses of H2O2 (100 µM) if previously treated with a non-toxic H2O2 concentration (20 µM) [69].

Thus, to evaluate whether incubation with compound (4) could or could not lead to increased or decreased H2O2 production, we determined the H2O2 release in promastigotes L. (L.) amazonensis at lag and log phases of the proliferation curve treated with or without compound (4). In this experiment, using Amplex® red as a probe, under our experimental conditions, it was possible to notice that the level of peroxide production was maintained between the lag and log phases; however, after treatment with compound (4) there was a reduction of 25.44 and 49.13% in the peroxide rates when compared to the lag and log phases, respectively (Fig. 3). This result suggests this compound has a higher effect of reducing peroxide production in the log phase.

Fig. 3
figure 3

Determination of H2O2 release in promastigotes L. (L.) amazonensis at lag and log phases of the proliferation curve treated or not with compound (4). *Present statistical difference

Full size image

Additionally, Finzi et al. (2004) observed that the treatment of parasites with concentrations of 20 uM H2O2 signaled the increase in the number of parasites. In this sense, as observed in our results, the treatment with compound (4) led to a reduction in H2O2 production, and as previously observed [62]. The reduction of this reactive oxygen species reduces the signaling for cellular proliferation, which can be observed by the reduction in the number of parasites after treatment with the compound (4).

Thus, our results demonstrated the production of silver compounds and their greater effectiveness against the parasite Leishmania, highlighting a promising approach to improve drug efficacy in treatment of the disease. In addition, this study also demonstrated the reduction of toxicity, which is a major problem of the drugs currently used in treatment. Consistent with these findings and based on the results obtained in our study; the silver and nitrate-containing compound is promising for the control of leishmaniasis and may represent an important option in the development of new drugs for the treatment of the disease.

Conclusion

In short, all compounds were evaluated for their antileishmanial and cytotoxicity properties. The most notable result was observed for compound (4) as the most active. In a comparative analysis of all the results, we verified that the modifications made in the molecule improved its activity in vitro and in vivo. Therefore, we conclude that probably the presence of a nitrate ion and the absence of the 1,10-phenanthroline ligand in the structure of the compound (4), together with its higher content of Ag+ ions were important for the modulation of the antileishmanial activity. This compound may represent a promising model for the development of a new class of antileishmanial agents and deserves further investigation of its effects on Leishmania to obtain possible mechanisms of action.