In vitro and in vivo effects of digoxin derivatives on promastigotes and amastigotes of Leishmania infantum

Abstract

Leishmania infantum is responsible for visceral leishmaniasis, the most severe form of the disease. The treatments currently used are pentavalent antimonials and amphotericin B, but they present serious side effects and resistant species that already exist. Therefore, it is necessary to find new compounds to avoid those acute problems. In vitro, this study evaluated the effects of fifteen digoxin derivative compounds on promastigotes and two of these compounds on amastigotes of L. infantum. We also assess the impact of these compounds on mitochondrial metabolism. The results showed that two of the fifteen compounds tested showed a high anti-Leishmania effect. Compounds 3 and 15 showed no toxicity across the cell lines RAW 264.7. Compound 15 inhibited macrophage infection and reduced the number of amastigotes. The two compounds showed a reduction in oxygen consumption by promastigotes. In this way, compounds 3 and 15 appear to be promising anti-leishmanial. Therefore, we checked both compounds in vivo also evaluating the toxicity by measuring ALT, AST, urea, and creatinine. Our results may contribute in the future to the development and application of new anti-Leishmania drugs.

KEYWORDS:

• Leishmania
• L. infantum
• leishmaniasis
• digoxin
• digitalis

1. Introduction

Leishmaniasis is a complex of diseases caused by the protozoan Leishmania genus and transmitted to the vertebrate host through infected female phlebotomine insects. Approximately 1 million new individuals are infected per year, which could lead to 20,000 additional deaths, and currently, some 1 billion people are in areas at risk of infection [1].

Visceral leishmaniasis (VL) is a serious chronic infection and if not treated properly, can be lethal to humans in up to 10% of cases. Its clinical manifestations are characterized by irregular and prolonged fever, anaemia, hepatosplenomegaly and, in some cases, cachexia. Species known as VL agents are Leishmania (L.) donovani and Leishmania (L.) infantum. L. donovani is found only in the Old World and is associated with poor rural populations in northeastern India and with displaced persons from East Africa and China. L. infantum is the responsible agent in the Mediterranean, West and Central Africa, the Middle East, China and the American Continent, including Brazil [2,3].

Most cases of VL occur in suburban and rural poor in five countries: India, Bangladesh, Sudan, Nepal and Brazil, and are a public health problem because they affect developing countries with socioeconomic inequality and low ability to afford the costs of diagnosis and treatment of these diseases [4].

In the last 70 years, the standard treatment for leishmaniasis has been the use of pentavalent antimonials. Other drugs such as pentamidine, amphotericin B (in a non-liposomal formula), miltefosine and paromomycin have been used as second-line drugs [5–7]. However, these drugs have shown several side effects, high toxicity, and the appearance of resistant strains has been reported. Failures observed during the treatment to leishmaniasis is a serious problem that needs close attention. The emergence of strains resistant to pentavalent antimonials is a major global health problem and is more serious in India where antimonial resistance levels exceed 60% [8,9]. Currently, studies on molecular targets involved in the resistance phenomenon have been highlighted. An important target evaluated is calcineurin, which is a phosphatase with a role in apoptosis. Strains of L. infantum were analysed and show that the downregulation of calcineurin could protect the parasites from apoptosis promoted by pentavalent antimonials [10]. In addition to resistance to pentavalent antimonials, resistance to amphotericin B is noteworthy. In the case of L. donovani, the molecular target related to L-asparaginase showed relevance when exposed to amphotericin B treatment. The presence of L-asparaginase improves tolerance to amphotericin B [11]. Many factors are involved in the emergence of resistance to leishmanicidal drugs, we can mention the extrinsic and intrinsic factors to the parasite and mainly the host's immune system. This problem has contributed to the urgent search for new compounds that are more effective and less damaging to patients [12,13].

Digoxin is a glycoside derived from the Digitalis lanata and it is one of the oldest drugs for heart treatments in use up today. It acts by inhibiting Na⁺/K⁺-ATPase activity, preventing intracellular sodium transport to the extracellular medium, which results in the loss of the transmembrane sodium gradient and inhibits Na⁺/Ca²⁺ exchanger activity, which raises intracellular Ca²⁺ levels, inducing cardiac muscle contraction [14,15].

Apart from its role in heart disease treatment, digoxin has been much discussed in the literature due to its high toxicity and its interaction with other drugs [16–18]. A study shows its estrogenic effect, which increased the risk of mortality in breast cancer patients [19]. However, there are also reports in the literature demonstrating its capacity as a potent inhibitor of HIV-1 replication and Chikungunya virus infection [20,21].

Although there are several publications on the use of digitalis glycosides, few works in the literature demonstrated anti-Leishmania effects of cardiac glycosides, describing the immunomodulatory effect of ouabain and digoxin [22–25]. This suggests the need to search and evaluate antiparasitic effects of digitalis glycosides, especially using derivatives modified from digoxin. Accordingly, the present study evaluated effects of digoxin derivatives compounds on promastigotes and amastigotes of Leishmania (L.) infantum to evaluate their possible use as a new antileishmanial drugs.

2. Material and methods

2.1. Synthetic compounds

The compounds were synthesized by Dr. Jose Villar from Universidade Federal de São João Del-Rey, Minas Gerais, Brazil [26,27] (Figure 1). The compounds were maintained at 4°C and stock solutions were dissolved in dimethyl sulfoxide (DMSO).

Figure 1. Chemical structures of digoxin and the respective radicals of their derivatives. The figure represents the radicals of each of the 15 digoxin derivatives inserted in the digoxin structure (highlighted rectangle) and its respective position.

2.2. Parasite cultures

Leishmania (L.) infantum (MHOM/BR/1974/M2682) parasites were kindly provided by Dr. Bartira Rossi-Bergmann (IBCCF, UFRJ – Rio de Janeiro/Brazil). This strain was maintained in vivo in the Syrian hamster (Mesocricetus auratus) model and promastigote forms were obtained after the isolation of amastigotes from spleens. Differentiation into infective promastigote forms was performed in culture medium 199 (Sigma®) supplemented with 20% foetal bovine serum, 5% male urine, 10 mM Hemin (Sigma®) and 10 mM folic acid (Sigma®) and were maintained by weekly subcultures in Schneider’s Insect medium (Sigma® S9895), supplemented with 10% foetal bovine serum (FBS) at 26°C.

2.3. Obtaining infective promastigotes

For differentiation and culture of the parasites, medium 199 (Sigma®) supplemented was used. Briefly, spleens from infected Syrian hamsters were macerated in 1 mL of supplemented medium 199 (Sigma®). Then, serial dilution was performed in sterile 24-well plates. The plates were incubated at 28°C and observed daily until the promastigote forms were fully differentiated. The promastigote forms obtained were transferred to a tissue culture flask 25 cm², containing 10–15 mL of supplemented 199 medium (Sigma®) and 10% Leishmania inoculum. The parasite culture was extended until obtaining the concentration of 1 × 10⁸ promastigotes/mL.

2.4. Mammalian cells culture

RAW 264.7 macrophages were maintained in Dulbecco’s Modified Eagle Medium (DMEM – Sigma® D5523) with 10% foetal bovine serum (FBS) and 1% antibiotic–antimycotic solution (Sigma® A5955) at 37°C in a controlled atmosphere of 5% CO₂. The antibiotic-antimycotic solution was deleted in the anti-intracellular amastigote activity assay.

2.5. Antipromastigote activity

Promastigotes in exponential growth phase were counted in a Neubauer chamber and seeded in Schneider medium supplemented with 10% FBS, at a concentration of 2 × 10⁵ cells per well in 96-well plates. Afterwards, the cells were incubated in the absence and presence of compounds at a range of concentrations from 5 to 100 µM at 26°C for 72 h in culture. Cells in the absence of the compounds but in the presence of 1% DMSO or 0.5 µg/mL of Amphotericin B (Amp B) were used as control. Their viability was verified by colorimetric reduction of 3-[4,5-dimethylthiazole-2-yl]−2,5-diphenyltetrazolium bromide (MTT) (Merck^® M-2003) (5 mg/mL) [28] using a microplate reader at 585 nm (Fluostar Optima, BMG Labtech, Offenburg, Germany). From this assay, the compounds that showed the inhibition of promastigote growth were selected and tested at different concentrations (5, 10, 15, 20, 35, 50, 75 and 100 µM) to determine the IC₅₀. Promastigote viability was measured using the same colorimetric method described above.

2.6. Cell viability assay

The cell monolayer of RAW 264.7 was harvested with a cell scraper and viable cells were counted by the exclusion method with Trypan blue, and 2 × 10⁵ cells were added in 96-well plates containing DMEM. The compounds were serially diluted in DMEM at concentrations of 6.25, 12.5, 25, 50 and 100 µM, at 37°C in a controlled atmosphere of 5% CO₂ for 72 h. Cells in the absence of compounds were used as control. The cell viability was analysed by neutral red (NR) assay with modifications [29]. The same procedure above was performed again, analysing the cytotoxicity of compounds 3 and 15 at high concentrations (50, 100, 150, 200, 250 and 300 µM) for 48 h, to obtain the CC50.

2.7. Haemolytic activity

The haemolytic effect of digoxin derivatives compounds was assessed as previously described [30] with small modifications. Sheep erythrocytes were washed three times and resuspended in Phosphate Buffered Saline (PBS) to a final concentration of 2% (v/v). Then the sheep erythrocyte suspension was incubated in the presence of digoxin derivative compounds (0.5–128 µM) for 1 h at 37°C. Afterwards, the erythrocyte suspension was centrifuged at 3000 rpm for 5 min and cell lysis was determined spectrophotometrically at 540 nm. The blank or the positive control was determined by the absence of compounds and in the presence of DMSO and 1% Triton X-100, respectively. The results were calculated by the percentage of haemolysis compared to the negative and positive controls. The use of sheep blood cells in this study was approved by Ethics Committee for the Use of Animals in research of Universidade Federal do Rio de Janeiro (CEUA-UFRJ: 157/21)

2.8. Anti-intracellular amastigote activity

RAW 264.7 cell line macrophages (4 × 10⁵/well) were placed on glass coverslips in a 24-well plate and incubated with DMEM at 37°C in a controlled atmosphere of 5% CO₂ for 3 h to achieve cell adhesion. Then promastigotes of L. (L.) infantum were incubated with macrophages at parasite: macrophage ratio of 10:1. After 24 h of infection, the plate was washed with phosphate buffered saline (PBS) to remove the non-internalized promastigotes. Infected macrophages were incubated in DMEM with 10% FBS in the absence or presence of different concentrations of compounds at 37°C in a controlled atmosphere of 5% CO₂ for 48 h. Infected macrophages in the presence of 1% DMSO or 0.5 µg/mL of Amp B but in the absence of the compounds were used as control. After that, coverslips were fixed with 3% formaldehyde for 15 min, stained with 10% Giemsa for 45 min, and were adhered onto a slide for analyses by optical microscopy. The percentage of infected macrophages and intracellular amastigotes was determined by counting of 100 cells in each coverslip [31].

2.9. Oxygen consumption

Oxygen consumption rates were measured by the high-resolution respirometry test with modifications [32] using an Oxigraph-2k high-resolution respirometer (Oroboros Instruments, Innsbruck, Austria). The electrode was calibrated between 0% and 100% saturation with atmospheric oxygen at 28°C. Oxygen consumption was measured using promastigotes (10⁶ cells/mL) in the absence or presence of the compound 3 or 15. To evaluate the activity of the mitochondrial respiratory chain, the titration of KCN (1–200 µM) was used. The data were analysed by DatLab software (Oroboros®).

2.10. Transmission electron microscopy (TEM)

To evaluate the ultrastructure of L. infantum, promastigotes treated or not with compounds IC₅₀ values for 72 h, the parasites (10⁶ cells/mL) were washed with serum-free medium and were fixed with 2.5% glutaraldehyde and 4% recently prepared formaldehyde in sodium cacodylate buffer (0.1M, pH 7.2) for 2 h. Cells were washed with sodium cacodylate buffer and post-fixed with 1% osmium tetroxide, 1.6% potassium ferrocyanide and 5 mM calcium chloride in sodium cacodylate buffer for 1 h. Cells were dehydrated with acetone serial concentrations of 30%, 40%, 50%, 70% and 100%. Inclusion was performed with epoxy resin. Ultrafine sections were contrasted with uranyl acetate and lead citrate and observed in a Transmission Electron Microscope FEI Spirit 120Kv [33].

2.11. Selectivity index (SI)

The selectivity index was determined according to the ratio between CC₅₀/IC₅₀ values [35].

2.12. Ethical statement

Syrian hamster experiments were performed with CEUA/CCS/UFRJ/Brazil protocol number 015/20. Animals were maintained with controlled temperature, 12 h light/dark cycles, and given water and feed ad libitum. Animals were euthanized with lethal doses of ketamine (300 mg/kg) and xylazine (30 mg/kg).

2.13. Experimental design in Syrian hamster model

Eight-week-old hamsters were infected by the intracardiac route with 10⁷ promastigote forms of L. (L.) infantum. The animals were anesthetized with ketamine (30 mg/kg) and xylazine (10 mg/kg) to carry out the infection. After 30 days of infection, the animals were treated by the intraperitoneal route (IP) with daily doses for 5 days of 2.5 mg/kg of the best compound or Glucantime® (Sanofi Aventis) [34]. In parallel, uninfected and infected hamsters were treated only with the same volume of PBS buffer. On day 1, after complete treatment, all the hamsters were euthanized with lethal doses of ketamine and xylazine. The sera of all animals were collected to measure the levels of urea, creatinine, aspartate aminotransferase (AST) and alanine aminotransferase (ALT). Finally, the parasite load was evaluated through the limiting dilution assay. The weight of spleens and livers assessed the hepatosplenomegaly since it is a crucial feature of visceral leishmaniasis.

2.14. Limitant dilution assay

Spleens from infected hamsters were aseptically removed after euthanasia. Then the spleens were macerated in 1 mL of the supplemented 199 medium. This suspension was added to the first well of the plate (24-wells) and serial dilution was performed to the end of the plate. The plates were incubated at 26°C for 12–14 days. Cultures were monitored daily by light microscopy. The titre dilution was given to the last well containing Leishmania [36].

2.15. In vivo toxicity

To evaluate the toxicity caused by the treatments, the renal and hepatic functions of the animals were monitored by measuring the levels of urea, creatinine, ALT and AST in the sera. All assays were performed at Laborlife Análises Clínicas, Rio de Janeiro, Brazil.

2.16. Statistical analysis

Statistical analyses were carried out through the SigmaPlot 12.0 program (Systat software, Inc.) using variance (ANOVA) and Tukey test where p < 0.05 value was considered significant. Besides in vivo tests, the existence of differences between the variables was determined by non-parametrical Kruskall Wallis and Mann Whitney tests (GraphPad Prism 8 program).

3. Results

3.1. Antipromastigote activity

Digoxin derivatives were tested at a final concentration of 100 µM against parasite growth for 72 h. Among the fifteen derivative compounds tested, four (3, 11, 13 and 15) inhibited more than 75% of parasite growth when compared to control with digoxin (Figure 2A). After screening, the four selected digoxin derivatives compounds were evaluated in different concentrations to observe the dose response curve in L. infantum promastigotes (Figure 2B and C), which was used to calculate the IC₅₀ (Table 1).

Figure 2. Effect of digoxin derivatives compounds in the growth of promastigotes of Leishmania infantum. (A) The compounds were tested at final concentration of 100 µM; (B and C) the compounds 3, 11, 13 and 15 were tested at different concentrations (5, 10, 15, 20, 35, 50 and 100 µM), at 26°C for 72 h. Data represent mean ± standard error of three independent experiments done in triplicates. In (A) the black bar represents control with promastigotes in the absence of compounds; the wave bar represents promastigotes treated with 100 µM digoxin; the striped bar represents promastigotes in the presence of 1% of DMSO; and grey bars represent promastigotes incubated in the presence of digoxin derivatives compounds. In (B and C), the black circles in the inset represent compounds 3 and 13 and the white circles represent compounds 11 and 15, respectively. * Indicates a significant difference (p < 0.05).

Table 1. Activity against Leishmania infantum promastigotes*.

Compounds	IC50 (µM)
3	46
11	67
13	70
15	42

*Data obtained from plots (fitting curve) of Figure 2(B and C).

Dose–response curves and IC₅₀ values showed that compounds 3 and 15 have a higher anti-leishmanial effect in comparison with other compounds, exhibiting IC₅₀ values below 50 µM, leading to the inhibition of growth at very low levels (around 20%). Because that they were selected for subsequent trials.

3.2. Cell viability assay

To continue studying the effects of compounds on the intracellular amastigote forms, a cytotoxicity assay was performed in macrophages from RAW 264.7 lineage incubated with different concentrations of compounds 3 and 15 for 72 h. The results in Figure 3 show that compounds were not toxic for macrophages, even at high concentrations (100 µM).

Figure 3. Cytotoxicity of compounds 3 and 15 on RAW 264.7 macrophages. The black bar represents the control with macrophages in the presence of 1% of DMSO; dark grey bar represents the macrophages incubated with compound 3 and, light grey bar the macrophages in the presence of compound 15. The incubation was carried out for 72 h. Data represent mean ± standard error of three independent experiments done in triplicates.

As an additional parameter of cytotoxicity, the haemolytic activity of the compounds was evaluated by incubating a suspension of red blood cells in the presence of compounds 3 and 15. It was observed that both compounds were non-toxic, even at high concentration when compared to the controls (Table 2).

Table 2. Haemolytic activity of digoxin derivatives.

Compounds	Haemolysis (%)
TRITON	100
DMSO 1%	ND
3 [128 µM]	1.68 ± 0.91
15 [128 µM]	ND

ND = not detectable.

3.3. Anti-intracellular amastigote activity

To evaluate the effects of digoxin derivatives on amastigote forms of L. infantum, cellular infections were done using lineage RAW 264.7 of murine macrophages. As shown in Figure 4(A), compound 3 had no effect at reducing macrophage infection, whereas compound 15, even at the lowest concentration (20 µM), reduced infection by up to 40%. In relation to the number of intracellular amastigotes, compound 15 decreased the number of intracellular amastigotes by up to 78% (Figure 4B).

Figure 4. Effect of digoxin derivative on infected macrophages (A) and intracellular amastigotes (B) treated with compounds 3 and 15 for 72 h. The black bar represents the control of infected macrophages in the absence of compounds; striped bar represents the control with macrophages in the presence of 1% of DMSO; dark and light grey bars represent the macrophages treated with compound 3 and compound 15, respectively. Data represent mean ± standard error of three independent experiments done in triplicates. * Indicates a significant difference (p < 0.05).

The IC₅₀ of the compounds were calculated from previous results. It was observed that IC₅₀ values are variable in different parasite life stages. Table 3 shows that compound 15 had a lower IC₅₀ for the amastigote form than for the promastigote form.

Table 3. Promastigote and amastigote IC₅₀ and selectivity index (SI) values.

	RAW 264.7 CC50 (µM)^a	Promastigote IC50 (µM)^b	Amastigote IC50 (µM)^c	SI (CC50/ IC50)^d
Compound 3	300 ± 13.5	46	>100	ND
Compound 15	300 ± 7.5	42	17.6 ± 2.6	17.0

^a CC₅₀ = cytotoxic concentration to 50% of RAW 264.7 macrophages.

^b IC₅₀ = concentration to inhibit 50% of the parasite growth.

^c IC₅₀ = concentration to reduce 50% the number of intracellular amastigotes.

^d SI values (SI = amastigote CC₅₀/IC₅₀).

ND: not determined.

Compounds cytotoxicity for RAW 264.7 macrophages and promastigotes IC₅₀ was compared using the selectivity index (SI), the ratio between the CC₅₀ and IC₅₀ promastigote values. The SI results in Table 3 shows compound 15 is 17 times less toxic for RAW 264.7 cells than for parasites when compared to amastigotes.

3.4. Oxygen consumption

Since this parasite possesses a huge and unique mitochondrion, it was important to evaluate mitochondrial function of the parasites treated with digoxin derivatives, so a high-resolution respirometry assay was performed. Promastigotes treated with compounds 3 and 15 at IC₅₀ concentration showed reduction if respiration rates of 67% and 63%, respectively, when compared to the control (Figure 5).

Figure 5. Mitochondrial respiration of promastigotes in the presence of compounds 3 and 15. The black circles represent the control of promastigotes in the absence of compounds; grey circles represent the promastigotes treated with 46 µM of compound 3; grey triangles represent promastigotes treated with 42 µM of compound 15. Data represent mean ± standard error of three different experiments.

3.5. Ultrastructural analysis

TEM analysis on promastigotes in the presence of compounds showed ultrastructural changes after 72 h treatment. Some parasites treated at IC₅₀ values of compound 3 were disrupted completely while others showed mitochondrial crest disorganizations (Figure 6), which may corroborate the mitochondrial alteration observed in the respirometry assay. Parasites treated with compound 15 were showed the presence of flagellar pocket deformation and the presence of myelin figures from cellular organelles (Figure 7).

Figure 6. Transmission electron microscopy of L. infantum promastigotes. (A) Untreated control; (B and C) treated with compound 3 (46 µM). Observe the disruption caused by the compound (B) and the mitochondrial crest alterations (arrows in C).

Figure 7. Transmission electron microscopy of L. infantum promastigotes. (A) Untreated control; (B, C and D) treated with compound 15 (42 µM). fp: flagellar pocket. Observe the presence of myelin figures (arrows in B and D) and flagellar pocket deformation (arrowheads in C).

3.6. Therapeutic efficacy of the best digoxin derivate (compound 15) and Glucantime® in Syrian hamster model

Since the compound 3 was less effective against intracellular amastigotes, we decided to continue with compound 15 only, for the next experiments. After 30 days of infection, the animals received 2.5 mg/kg of each drug intraperitoneally for 5 consecutive days. The drug of the first choice, Glucantime®, was used as a control of the treatment as well as the infected and untreated groups were used as a control of the infection. A group of normal animals that was neither infected nor treated was used for comparison purposes at the endpoint of the experiment. All groups showed classic splenomegaly as the infection progression (p < 0.05) (Figure 8A). We observed an increase of approximately twice the weight of the spleens in the normal hamsters. Although hepatomegaly is also a clinical sign of the course of infection in hamsters, no differences were observed (p > 0.05) (Figure 8B). The efficacy of the compound 15 was demonstrated in the limiting dilution assay (Figure 8C). A significant reduction in parasite load was shown (p < 0.05).

Figure 8. Determination of the splenomegaly (A) hepatomegaly (B) and parasite load in spleens (C). Hamsters were infected with 10⁷ promastigote forms. After 30 days of infection, the animals were treated for 5 days with 2.5 mg/kg of Glucantime® or compound 15. One day after the complete treatment, all hamsters were euthanized. The in vivo experiments were repeated three times (n = 3–4 animals per group). * Indicates statistical differences (p < 0.05).

We observed in the groups treated with Glucantime® and the compound 15, a reduction in 94.5% and 95.3%, respectively (Figure 8C). And surprisingly there were no differences between them. Compound 15 is shown to be as potent as Glucantime® in the treatment of visceral leishmaniasis.

3.7. Toxicity in vivo

Serious adverse reactions involving hepatotoxicity and nephrotoxicity are common manifestations in the classic treatment used in the treatment of visceral leishmaniasis. In this study, our results showed a lower level of urea in the sera of infected animals treated with compound 15 (p < 0.05) (Figure 9A). However, no differences were found between the levels of creatinine compared to the Glucantime® group (Figure 9B). Finally, all hepatotoxicity parameters analysed showed no significant differences (Figure 9C and D) (p > 0.05).

Figure 9. Analysis of nephrotoxicity and hepatotoxicity in infected hamsters after chemotherapy. Serum samples from the hamsters were collected after euthanasia to measure the levels of urea, creatinine, ALT, and AST. All analyses were repeated three times (n = 3–4 animals per group). *Indicates statistical differences (p < 0.05).

4. Discussion

Leishmaniasis is a disease caused by protozoa of the Leishmania genus and they are among infections with the highest human activity loss indexes per years, presenting a rate of DALYs (disability-adjusted life year) of 2.35 million, of which 2.3% corresponds to the Americas [37].

During the last 70 years, the standard treatment for leishmaniasis has been pentavalent antimonial [5]. However, treatment with this drug is expensive, long-term and no oral administration which, in addition to causing several side effects, leads some patients to withdraw from treatment. Other drugs, such as pentamidine, amphotericin B, miltefosine and paramomycin, have also been used to treat leishmaniasis [7]. Therefore, treatment options against leishmaniasis are limited, or in specific cases are ineffective.

Most drugs available to treat leishmaniasis also have several side effects, nephrotoxic, cardiotoxic, or may lead to renal and hepatic insufficiency, and also could be responsible for some cases of drug resistance. All this points to an urgent need to discover and study new antileishmanial compounds with low or no toxicity and side effects [13,7].

The use of digitalis has been described since 1785, and over the last 50 years this class of compounds has been studied and used for the treatment of heart diseases [38,39]. Digoxin is the main form of digitalis used today and acts inhibiting Na⁺/K ⁺-ATPase [14].

The use of digoxin has decreased since 90s and has been extensively discussed in the literature because of its toxicity, which may arise during long-term therapy or after an overdose, and may occur even within the therapeutic limits [40]. Its toxicity is involved in the cases of anorexia, nausea, vomiting, neurological symptoms and cardiac arrhythmias [41]. Chronic toxicity is more common than acute intoxication and its clinical and laboratory manifestations occur more frequently and exacerbated [40].

Knowing the toxicity of this drug, several studies have made changes in the digoxin molecule, synthesizing derivatives to find interesting biological activity. Many authors have demonstrated the effects of various digoxin derivatives on different cell lines and observed their cytotoxic effects. Rocha et al. [42] reported the effects of adding a styrene group to the lactone ring of digoxin, where they observed virtually no Na⁺/K ⁺-ATPase activity, the inhibition of the multiple drug transporter Pdr5p, the regulation of activity and the expression of Na⁺/K ⁺-ATPase in cancer cells, such as HeLa (human cervical carcinoma) and RKO (colon carcinoma). Other studies have shown effects on the suppression of Th17 cell differentiation and in the selectivity for another Na⁺/K ⁺-ATPase isoform, which acts to potentially reduce intraocular pressure [43,44].

There is only one study showing potential anti-leishmanial effects of cardiac glycosides, which described an immunomodulatory effect of ouabain and digoxin. Ouabain reduced total cell numbers in the peritoneal cavity as a reflex of inhibition of neutrophil migration induced by Leishmania (L.) amazonensis. Ouabain also reduced TNF-α and IFN-γ levels, without cytotoxicity against peritoneal macrophages. The immunophenotype of human patients infected by Leishmania guyanensis (L. guyanensis) was studied by Hartley et al. [25]. They found significant association between the inflammatory cytokine IL17A, the presence of a cytoplasmic virus within L. guyanensis parasites (LRV1), and disease chronicity that could lead to extensive open ulcers that metastasize into secondary sites and are often resistant to standard therapies. In addition, the study confirmed that digoxin does not have IL-17A independent parasite toxicity in vivo. As IL17A has been shown to be pathologic in other forms of leishmaniasis, digoxin may hold therapeutic potential across a broader range of leishmanial species [25].

In this work, digoxin derivative compounds were tested, evaluating their effects against promastigotes and amastigotes of L. infantum, to find possible antileishmanial activity. Initially, screening with these compounds was performed to check which would be able to inhibit promastigote growth. Using a concentration of 100 µM, only four derivatives (compounds 3, 11, 13 and 15) showed strong inhibition of promastigote growth (Figure 2), showing some specificity related to primarily digoxin structure, the standard and model for synthesis of the derivatives.

The dose–response curves of the four compounds that showed the greatest ability to inhibit the growth of promastigotes provided IC₅₀ values (Table 1), between 42 and 70 µM. To our knowledge, the anti-parasitic effect of these compounds has not been described in the literature. Some studies show the effects of digitalis glycosides on viruses [21] and bacteria [45].

Despite descriptions of the toxicity of digoxin in the literature [40], the toxicity tests with derivatives that were made using digoxin as a model showed that compounds 3 and 15 were not toxic to RAW 264.7 macrophages even in high concentrations. This allowed the evaluation of their effects on intracellular amastigote forms, without any damage to host macrophages. An important criterion in the search for new compounds for leishmaniasis treatment is to determine the absence of toxic effects in the host cells. These results are especially interesting because they show that compounds 3 and 15 are less toxic to macrophages than to parasites.

When evaluating effects of these compounds on amastigote forms, we observed reduced number of infected macrophages and intracellular amastigotes when treated for 72 h with compound 15. The IC₅₀ for this biological form was also found to be lower than for the promastigotes.

The cytotoxicity of the compounds for the RAW 264.7 macrophages and for parasites was compared using the selectivity index (SI), the ratio between the CC₅₀ of the RAW 264.7 macrophages and the IC₅₀ for amastigotes. The data shows compounds 3 and 15 are less toxic to macrophages than to parasites. Compound 15 had an SI of 17.0, suggesting that the compound is more selective for the parasite (Table 3). The criterion of success and leadership in the discovery of drugs for infectious diseases is an ideal SI value above 10 [46].

Thus it is possible that compound 15, which had a more potent effect against amastigotes, acts directly on parasites and not on macrophages, demonstrating selectivity for these forms. To confirm this selectivity and rule out whether this activity could be attributed to an immunomodulatory effect of host cells, assays evaluating cytokine production should be performed. Martins et al. [47] observed that guanidinic compounds, which showed strong activity against L. infantum amastigotes and Trypanosoma cruzi trypomastigotes, did not induce nitric oxide production, and suppressed the production of cytokines such as TNF, IFN-γ and MCP-1 in macrophages infected with Leishmania, suggesting the selective elimination of the parasites. Such cytokines could be analysed in the future and contribute to increase knowledge about the effects of digitalis in these parasites.

To evaluate mitochondrial function, a respirometry test was performed. In this assay, the titration of KCN on the promastigotes was done to observe and compare oxygen consumption under normal conditions and in the presence of the compounds. KCN acts as a specific inhibitor of the electron transport chain IV complex (cytochrome c oxidase), inhibiting oxidative phosphorylation and, in consequence, ATP synthesis [48]. The promastigotes were treated with compounds 3 and 15 for 48 h. After this time, promastigotes in the presence of both compounds 3 and 15 showed reduced respiration rates, suggesting these compounds may cause damage to the electron transport chain.

Transmission electron microscopy was used to observe the organelles affected by compounds 3 and 15. Mitochondrial crest alteration and presence of myelin figures were observed in promastigotes in the presence of compounds 3 and 15 suggesting that they may be acting in the mitochondria and in the membrane, respectively. Compound 15 showed the deformation of the flagellar pocket and the presence of myelin figures, which could indicate autophagy due to a nutrient shortage. Autophagy can promote parasite clearance, so more tests should be performed to better understand this process.

Our in vitro experiments generated great expectations regarding its promising activity in the in vivo model. Thus the digoxin derivative that presented the best in vitro performance was used in the treatment of hamsters infected with L. infantum. The Syrian hamster model is known to be quite susceptible to visceral leishmaniasis and to present the evolution of leishmaniasis in a similar way to what happens in humans [49,50]. Therefore, it is the best model for the study of promising drugs in the treatment of visceral leishmaniasis. Our findings showed significant splenomegaly in all groups of animals that were infected except in the control group of normal animals, which corroborates what is described in the literature [49,51]. It is noteworthy that compound 15 was as potent as Glucantime® in treating animals, both drugs showed more than 90% reduction in parasite load found in the spleens of infected and treated animals.

One of the most serious problems caused by Glucantime® is its strong toxicity [52,53]. In this study, the group of animals treated with our compound 15 had significantly lower urea levels compared to the group treated with Glucantime®. This finding is of great importance because it presents a new drug as potent as Glucantime® in the ability to reduce parasite load, but not nephrotoxic. Therefore, compound 15 achieves the goals of a new drug presenting therapeutic efficacy and low toxicity.

Although this study was focused only on the treatment of visceral leishmaniasis, the need to study these derivates digoxin compounds against cutaneous leishmaniasis is evident. We hypothesize that digitalis derivatives will show excellent protective response and successful therapeutic efficacy in tegumentary leishmaniasis.

5. Conclusion

In summary, the results obtained so far show that compounds 3 and 15 strongly inhibited the growth of L. (L.) infantum promastigotes, and this effect may be related to the inhibition of mitochondrial metabolism. Compound 15 also exerts antileishmanial activity against amastigotes, which is essential to restrain the infection, since this is the intracellular form of the parasite. Further, it is noteworthy to highlight the in vivo effect of compound 15, a digoxin derivative that may be effective in treating leishmaniasis without displaying the adverse effects of the current first-choice treatment. These results are promising since they show these compounds have potent anti-leishmanial activity and could be a new tool to treat the dangerous diseases caused by L. infantum.

Acknowledgments

The authors thank Dr. Thomas Goreau (CEO of Global Coral Reef Alliance) and Dr. Daniel Clemente de Moares (IMPG – UFRJ) for critical review of this manuscript. The authors also like to thank Dr. Alane Beatriz Vermelho, Dr. Eliana Barreto-Bergter and Dr. Lucy Seldin for the valuable financial support.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This study was financed by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior – Brazil (CAPES) – Finance Code 001, FAPEMIG APQ 01915–22 and APQ 00855-19.