Leishmanicidal Activity of Betulin Derivatives in Leishmania amazonensis; Effect on Plasma and Mitochondrial Membrane Potential, and Macrophage Nitric Oxide and Superoxide Production

Herein, we evaluated in vitro the anti-leishmanial activity of betulin derivatives in Venezuelan isolates of Leishmania amazonensis, isolated from patients with therapeutic failure. Methods: We analyzed promastigote in vitro susceptibility as well as the cytotoxicity and selectivity of the evaluated compounds. Additionally, the activity of selected compounds was determined in intracellular amastigotes. Finally, to gain hints on their potential mechanism of action, the effect of the most promising compounds on plasma and mitochondrial membrane potential, and nitric oxide and superoxide production by infected macrophages was determined. Results: From the tested 28 compounds, those numbered 18 and 22 were chosen for additional studies. Both 18 and 22 were active (GI50 ≤ 2 µM, cytotoxic CC50 > 45 µM, SI > 20) for the reference strain LTB0016 and for patient isolates. The results suggest that 18 significantly depolarized the plasma membrane potential (p < 0.05) and the mitochondrial membrane potential (p < 0.05) when compared to untreated cells. Although neither 18 nor 22 induced nitric oxide production in infected macrophages, 18 induced superoxide production in infected macrophages. Conclusion: Our results suggest that due to their efficacy and selectivity against intracellular parasites and the potential mechanisms underlying their leishmanicidal effect, the compounds 18 and 22 could be used as tools for designing new chemotherapies against leishmaniasis.


Introduction
Leishmaniasis is endemic in 98 countries, in which 14 million people remain infected and more than 350 million are at risk [1,2]. Approximately 1.3 million people contract the disease annually; 300,000 cases correspond to visceral leishmaniasis (VL) and 1,000,000 to tegumentary leishmaniasis (TL). The estimated number of deaths from VL ranges between 20,000 and 50,000 each year [3]. However, these numbers reflect the tip of the iceberg, since only 40 countries where the disease is endemic regularly report their health statistics to the World Health Organization (WHO) [4].
In Venezuela, VL has a low incidence (0.03 to 0.24 per 100,000 inhabitants) [5], although a notable under-registration of cases remains [6]. The main focus of VL is located at the Herein, we evaluated in vitro the anti-leishmanial activity of betulin derivatives in Venezuelan isolates of Leishmania amazonensis [L. (L.) amazonensis], isolated from patients with therapeutic failure. Our results suggest that, due to their efficacy and selectivity against intracellular parasites and the mechanisms underlying their leishmanicidal effect, some of these compounds could be used as tools for designing new chemotherapies against leishmaniasis. The patients were diagnosed DCL and suffered therapeutic failure against meglumine antimoniate (Glucantime ® , Sanofi-Aventis (Suzano, SP, Brazil). Species identification was performed by PCR and molecular hybridization [31]. This project (CDCH UCV PI-09-8717-2013) was approved by the Ethics Committee from the Institute of Biomedicine (SAIB-MPPS-UCV), Caracas, Venezuela. Isolates were kept in liquid nitrogen until use, when parasites were thawed and cultured in blood agar-glucose enriched with human urine [32,33].

Materials and Methods
Leishmania promastigotes were grown at 26 • C in semisolid blood agar supplemented with glucose-NaCl medium (glucose 1.5%, NaCl 0.85%, w/v). Promastigotes at late log growth phase were collected by centrifugation at 125× g for 10 min at room temperature (RT). The medium was decanted and the cells were suspended in the appropriate medium up to the desired cell density.

Compounds
Twenty-eight semisynthetic betulin derivatives (20 structurally simple and 8 heterocyclic derivatives, compounds 1-20 and 21-28, respectively) were tested. The compounds, stable at RT, were dissolved in DMSO at a concentration of 10 mM and stored at −20 • C until use. Their synthesis and properties have been described by Alcazar et al. and references therein [28]. A summary of their chemical formulae, structures, and molecular weights is included in Supplementary Materials I.

Promastigote In Vitro Susceptibility
The fluorometric method of Alamar Blue ® (resazurin) was used [35]. Promastigotes grown in biphasic agar-blood medium, at the exponential phase of growth, were seeded in flat-bottom 96-well culture plates, at a cell density of 1 × 10 6 parasites mL −1 (200,000 parasites/well), in 180 µL of RPMIc without phenol red. Subsequently, either a sole concentration (50 µM), or increasing concentrations of compounds or drugs were applied, according to each treatment group, in volumes of 2 µL. Plates were kept at 26 • C, Microorganisms 2021, 9, 320 4 of 18 in a humid chamber, for 72 h. Resazurin was then added (20 µL of 10% v/v), to reach a final concentration of 250 µM and plates were incubated for an additional 4 h. Live parasites are metabolically active and capable of reducing non-fluorescent resazurin (blue) to fluorescent resorufin (red). Fluorescence intensity is directly proportional to the number of viable cells. The fluorometric reading was carried out using the appropriate wavelengths (ex: 540 nm/em: 580 nm) in a spectrofluorometer Victor 2 Wallac (Perkin Elmer, Turku, Finland). At least two independent experiments were performed in duplicate. The effect of trivalent antimonial [SbIII, potassium antimonyl tartrate trihydrate, Sigma-Aldrich, (St. Louis, MO, USA)] was included as positive control of the assay. Based on growth inhibition caused by the tested compounds, the percentage of growth inhibition compared to untreated cells was used to calculate the GI 50 as the concentration that causes 50% inhibition of cell proliferation [36].

Cytotoxicity and Selectivity
Cytotoxicity of "active compounds" was evaluated by the fluorometric method of Alamar Blue ® [37]. In flat-bottom 96-well culture plates, J774.1 macrophages were seeded at a cell density of 5 × 10 4 mL −1 (10,000 macrophages/well) in 200 µL of RPMIc. After 24 h at 37 • C, 5% CO 2 , medium was replaced by 180 µL of RPMIc, without phenol red, and increasing concentrations of test compounds in 2-µL volumes. Plates were kept at 37 • C, 5% CO 2 for 72 h; then resazurin was added (20 µL, 10% v/v) to a final concentration of 250 µM and plates were incubated for additional 4 h. Fluorometric readings were carried out using the appropriate wavelengths (ex: 540 nm/em: 580 nm) in a spectrofluorometer Victor 2 Wallac (Perkin Elmer, Turku, Finland). At least two independent experiments were performed in duplicate. Based on the inhibition of cell growth caused by the tested compounds, the percentage of growth inhibition compared to untreated cells was calculated. With these percentages, the concentration that causes cytotoxicity in 50% of macrophages (CC 50 ) of macrophages was calculated. Selectivity indices (SI) were calculated, dividing the macrophage CC 50 by the parasites GI 50 (SI = CC 50 /GI 50 ) [38].

Macrophage-Amastigote Model
We used a previously described methodology [39], with some modifications. Macrophage monolayers, grown in RMPIc medium, for 3 days, were detached from the bottom of the culture chamber, by physical treatment (low temperature and light strokes). Obtained macrophages were seeded in flat-bottom 96-well culture plates, at a density of 2 × 10 5 cells mL −1 (40,000 macrophages/well), in a final volume of 200 µL. Plates were incubated for 2 h, and the supernatant was discarded to remove non-adhered macrophages. Medium was replaced with stationary-phase promastigotes suspended in 200 µL of RPMIc (2 × 10 6 parasites mL −1 , 400,000 parasites/well), at a macrophage: parasite ratio 1:10 [40]. The plates were incubated for 4 h. Afterwards the medium was replaced with fresh RPMIc and increasing concentrations of compounds or drugs tested were applied, in volumes of 2 µL, according to the different treatment groups. Plates were kept for 72 h, at 37 • C and 5% CO 2 . The medium was discarded and the wells were washed twice with PBS. Next, a Giemsa staining was performed in situ. It consisted of adding 200 µL of methanol/well, incubating for 5 min, discarding the fixative, and letting it dry at room temperature. Later, the Giemsa solution (10% in PBS w/v, 200 µL/well) was added, incubated for 10 min, and washed with PBS, allowing it to dry. Finally, the bottom of the plates was detached by percussion and the circular bottoms (wells) were fixed to slides with Permount synthetic resin, for later evaluation (Alcazar W., manuscript in preparation). Cell counting was performed with a 100× immersion objective. A minimum of 200 random macrophages were examined. The percentage of infected macrophages (% infection) and the average number of amastigotes present in infected macrophages ( amastigotes ) were determined. From these data, the infection index (inf-I) was calculated: inf-I = % infection × amastigotes . The infection rate represents the number of intracellular amastigotes per 100 macrophages evaluated [40,41]. Additionally, the GI 50 for the decrease in the infection rate was calculated.

Mitochondrial Membrane Potential (∆Ψm)
To evaluate the mitochondrial membrane potential (∆Ψm), the fluorescent compound JC-1 (5,5 ,6,6 -tetrachloro-1,1 ,3,3 -tetraethylbenzimidazolocarbocyanine iodide) was used, according to a modification of a previously described method [45]. Exponential phase promastigotes were suspended at 26 • C, in RPMIc without phenol red, in a volume of 1 mL, at a cell density of 1 × 10 7 parasites mL −1 . The parasites were left to grow for 48 h, in the presence and absence of the test compounds. Then, after centrifuging (500× g for 10 min) and washing the parasites twice with PBS, they were resuspended in 200 µL of HEPES-G, and placed in 96-well plates. In the depolarizing control wells, 1 µM carbonyl cyanide 4-(trifluoromethoxy) phenylhydrazone (FCCP) and 20 µM oligomycin were added. The JC-1 probe was then added, at a final concentration of 15 µM. Then, the plates were incubated for 30 min, at RT in the dark. Finally, fluorescence was measured in a Victor 2 Wallac spectrofluorometer (Perkin Elmer, Turku, Finland) at the appropriate wavelengths (ex: 485 nm/em: 535 and 580 nm). As the mitochondrial membrane is polarized, JC-1 produces intra-mitochondrial aggregates "J-aggregates" that emit a red fluorescence (λ 590 nm). Mitochondrial membrane depolarization induces JC-1 monomeric Microorganisms 2021, 9, 320 6 of 18 conformation in the cytoplasm, then the fluorescence turns green (λ 530 nm). The red/green ratio of fluorescence then is an indirect measurement of the mitochondrial membrane potential. A more depolarized mitochondria means a lower ratio of red/green fluorescence.
Relative fluorescence (RF) was calculated, where RF = 580 nm fluorescence/535 nm fluorescence (red/green). A decrease in the value of RF would indicate the collapse of the mitochondrial membrane potential. At least two independent experiments were performed in duplicate. Compounds used as depolarizing controls were: FCCP, an oxidative phosphorylation inhibitor and ionophore, which acts by decoupling the respiratory chain, interrupting the synthesis of ATP by transporting hydrogen ions through the mitochondrial membrane before they can be used as a source of energy; and oligomycin, an antibiotic that inhibits mitochondrial ATPase by binding to the F0 subunit and interfering with the transport of H + , thus inhibiting the synthesis of ATP [43].

Nitric Oxide (NO) Production
Following a previously described procedure [46,47], infected and uninfected macrophage cultures were treated with test compounds to analyze their effect in Leishmania host cell nitric oxide (NO) production. J774.1 macrophages, at the third day of growth, were collected and seeded in flat-bottom 96-well culture plates, at a density of 1 × 10 6 cells mL −1 (2 × 10 5 macrophages/well), in 200 µL of RPMIc. The plates were incubated for 2 h; once the supernatant was discarded, the medium was replaced with 200 µL of RPMIc mixed with infectious promastigotes in the stationary phase of growth (1 × 10 7 parasites/mL, 2 × 10 6 parasites/well), so that the ratio macrophage: parasites was 1:10 [41]. The plates were then incubated for 4 h, the medium was replaced with fresh RPMIc and the test compounds were applied, according to the different treatment groups. The plates were kept in incubation for 48 h, at 37 • C and 5% CO 2 .
From the culture supernatants, 50-µL aliquots were recovered and transferred to a new 96-well plate to perform the Griess reaction [48], 50 µL of Solution 1 [1% sulfanilamide dissolved in 2.5% phosphoric acid] was added, incubated for 5 min at RT and in the dark. Subsequently, 50 µL of Solution 2 [0.1% N-(1-naphthyl) ethylenediamine dihydrochloride dissolved in distilled water] was added, stirred for an additional 5 min, at RT and in the dark. Finally, the colorimetric reading was performed at 550 nm. Three independent experiments were performed in duplicate. The Griess reaction quantifies NO, using nitrite (NO 2 − ) as the most stable species. NO 2 − concentrations in culture medium reflect indirectly the levels of NO produced by macrophages. Levels of NO 2 − (expressed as µM of NO 2 − ) were calculated from a calibration curve constructed with increasing concentrations of sodium nitrite (NaNO 2 ). MLF (5 µM) and AmB (0.1 µM), known to induce NO production in macrophages, were included as controls.

Superoxide (O 2 − ) Production
The intracellular oxidative burst was evaluated by the production of the superoxide anion O 2 − , using the method of reduction of ferricytochrome c, previously described [46]. Macrophages from preceding experiments were used after removing the supernatant; the cells pretreated with compounds or AmB as positive control of the experiment were resuspended in 100 µL of Hanks saline (HBSS: 137 mM NaCl, 5.4 mM KCl, 1 mM MgSO 4 ·7H 2 O, 1.3 mM CaCl 2 ·H 2 O, 0.3 mM Na 2 HPO 4 , 0.44 mM KH 2 PO 4 , 4.2 mM NaHCO 3 , 5 mM glucose, pH 7.2) and incubated for 1 h in the presence of 2 mg mL −1 (160 µM) of cytochrome c type III, at 37 • C and 5% CO 2 . The supernatant was saved to measure the change in optical density at 550 nm. Three independent experiments were performed in duplicate. The superoxide anion concentration was calculated using the reduced molar extinction coefficient of cytochrome c, 21 mM −1 cm −1 . Therefore, [O 2 − ] = (∆OD/6.3) × 100, where ∆OD is the variation of the optical density (reading-blank), and the result obtained corresponds to the concentration of superoxide anion expressed in nanomoles of O 2 − released/2 × 10 5 cells. MLF (5 µM) and amphotericin-B (AmB, 0.1 µM), known to induce NO production in macrophages, were included as controls.

Statistics
GraphPad Prism 8 software (2018, San Diego, CA, USA) was used. The Kolmogorov-Smirnov normality test and the Barlett test for homogeneity of variances, with a 95% confidence, were used to test the null hypothesis that the data were taken from a Gaussian distribution. Since the value of p > 0.05 showed that the data did not deviate from the normal distribution, the variance analysis and the Bonferroni multiple comparisons test were used to compare the results between different test conditions [49]. Results were expressed as mean ± standard error of the mean. In all tests, a value of p < 0.05 was considered statistically significant. GI 50 and CC 50 values were calculated by non-linear regression of the sigmoid response dose curve with variable slope [50]. Supplementary Materials III includes tables of the statistical analysis performed for each of the methods used to analyze the potential mechanism of action of compounds.

Selection of the Tool Compounds
To select compounds that may be interesting for their potential anti-leishmanial activity, 20 structurally simple (compounds 1-20) and 8 heterocyclic (compounds 21-28) betulin derivatives [28,29] were evaluated by their capacity to inhibit reference strain (LTB0016) growth at one sole concentration of 50 µM (upper limit to consider their clinical value) [38]. For details on their chemical formula, structure, and molecular weight, please consult Supplementary Materials I.
A total of eight compounds that reduced at least 75% of parasite growth were forwarded to the next step of analysis as "active compounds". These compounds were tested at increasing concentrations to find out their GI 50 against the Leishmania reference strain.
To quantify the leishmanicidal effect of the best compounds of both groups of betulin derivatives, GI 50 was determined in promastigotes of LTB0016, at increasing concentrations from 1 to 100 µM and 72 h of incubation (see Table 1). One-way ANOVA analysis, followed by Bonferroni test, allowed identification of significant differences (p < 0.05) of the tested compounds vs. SbIII [GI 50 = 5.51 (IL = 4.92-SL = 6.08 µM] against promastigotes. Compounds 17, 18, 22, 26, and 27 were as effective as the reference compound to inhibit LTB0016 growth. However, a significant difference against SbIII (p = 0.0413) was only detected for 18, being even stronger than it. Potential cytotoxicity was evaluated in J774.1 macrophages. Due to their low cytotoxicity, 18 and 22 presented the best SI (SI > 15) and were forwarded to the next step of research.

Effect of Compounds on Intracellular Amastigotes
To confirm the anti-leishmanial effect of the compounds 18 and 22, GI 50 was determined for intracellular amastigotes of reference strain LTB0016 and patient isolates, using the amastigote-macrophage model. Thus, infected macrophages were exposed to increasing concentrations of 18 and 22 (0.1, 0.5, 1, 5, and 10 µM). The infectious index data is included as Supplementary Materials (Supplementary Materials II). Final results are summarized in Table 2. Compounds 18 and 22 showed GI 50 < 1 µM and can be considered safe in their toxicity against J774.1 macrophages (SI > 20), both for reference strain LTB0016 and patient isolates.

In the Search of the Mechanism of Action of Betulin Derivatives
In the search of the mechanism of action of the selected compounds, their effect on the plasma membrane potential and mitochondrial membrane potential was evaluated in promastigotes. In non-infected and infected macrophages, the effect of selected compounds was evaluated on nitric oxide production and on superoxide production.
The resting ∆Ψp of parasites was then established (LTB0016: −181.97 ± 19 mV; VE98MR = −166.75 ± 11 mV and VE2000MM = −158.30 ± 8 mV). The isolates presented more depolarized resting membrane potentials than the reference strain (see Supplementary Materials III for statistics). Figure 1 demonstrates that only 18 was able to significantly depolarize the ∆Ψp in all evaluated strains when compared to untreated controls (p < 0.05). The effect was similar to that depicted by valinomycin and concentration dependent, except in the case of VE2000MM. This strain was not sensitive to valinomycin.

Mitochondrial Membrane Potential (∆Ψm) in Promastigotes
To evaluate if the effect of the betulin derivatives on parasites was related to their action on mitochondria, or related to programmed cell death, their effect on the mitochondrial membrane potential (∆Ψm) was evaluated. Promastigotes were exposed to two concentrations of the compounds, equivalent to their GI 50 and twice GI 50 (as before, 18: 1 and 2 µM; 22: 2 and 4 µM), during 48 h. As controls, untreated cells, as well as cells treated with depolarizing compounds (FCCP 1 µM and oligomycin 20 µM) were included. As the label for ∆Ψm, the lipophilic cationic probe JC-1 was used. Figure 2 demonstrates how 18 depolarizes ∆Ψm in a concentration-dependent way when compared to untreated cells (p < 0.05) in all studied cells. The effect of 18 at its higher concentration (2 µM) in LTB0016 and VE98MR was comparable to that of oligomycin 20 µM, a compound that quickly inhibits the mitochondrial ATPase and ATP synthesis. 22 also inhibited ∆Ψm, and a significant difference (p = 0.025) was only present at the highest concentration used (4 µM) and for LTB0016 (see Supplementary Materials III for statistical details).

Mitochondrial Membrane Potential (ΔΨm) in Promastigotes
To evaluate if the effect of the betulin derivatives on parasites was related to their action on mitochondria, or related to programmed cell death, their effect on the mitochondrial membrane potential (ΔΨm) was evaluated. Promastigotes were exposed to two concentrations of the compounds, equivalent to their GI50 and twice GI50 (as before, 18: 1 and 2 µM; 22: 2 and 4 µM), during 48 h. As controls, untreated cells, as well as cells treated with depolarizing compounds (FCCP 1 µM and oligomycin 20 µM) were included. As the label for ΔΨm, the lipophilic cationic probe JC-1 was used. Figure 2 demonstrates how 18 depolarizes ΔΨm in a concentration-dependent way when compared to untreated cells (p < 0.05) in all studied cells. The effect of 18 at its higher concentration (2 µM) in LTB0016 and VE98MR was comparable to that of oligomycin 20 µM, a compound that quickly inhibits the mitochondrial ATPase and ATP synthesis. 22 also inhibited ΔΨm, and a significant difference (p = 0.025) was only present at the highest concentration used (4 µM) and for LTB0016 (see Supplementary Materials III for statistical details).

Nitric Oxide Production (NO) in Macrophages
To evaluate if the anti-leishmanial activity of compounds 18 and 22 is related to modulating effects on the microbicidal activity of macrophages, non-infected and infected cultured macrophages were exposed to two concentrations of the compounds, promastigote GI 50 and twice GI 50 (18: 1 and 2 µM; 22: 2 and 4 µM). The NO production was indirectly determined through the evaluation of nitrite (NO 2 − ) levels by the Griess reaction. Compounds 18 and 22 were not capable of inducing NO production in any of the strains tested ( Figure 3). Supplementary Materials III details the associated statistics. Figure 2. Effect of betulin derivatives on mitochondrial membrane potential (ΔΨm) in L. (L.) amazonensis reference strain LTB0016 and isolates VE98MR and VE2000MM promastigotes isolated from patients, at two increasing concentrations (a and b): 18 (1 and 2 µM), 22 (2 and 4 µM), incubated for 48 h. Each bar represents the average (n = 3 independent experiments in duplicate) with its respective standard error (SEM). * p < 0.05; ** p < 0.005; *** p < 0.0005 with respect to the control without treatment. Control: untreated control; FCCP: carbonyl cyanide 4-(trifluoromethoxy) phenylhydrazone, 1 µM; OLIG: oligomycin, 20 µM.

Nitric Oxide Production (NO) in Macrophages
To evaluate if the anti-leishmanial activity of compounds 18 and 22 is related to modulating effects on the microbicidal activity of macrophages, non-infected and infected cultured macrophages were exposed to two concentrations of the compounds, promastigote GI50 and twice GI50 (18: 1 and 2 µM; 22: 2 and 4 µM). The NO production was indirectly determined through the evaluation of nitrite (NO2 − ) levels by the Griess reaction. Compounds 18 and 22 were not capable of inducing NO production in any of the strains tested (Figure 3). Supplementary Materials III details the associated statistics.

Superoxide Production (O2 − ) in Macrophages
Additionally, the O2 − was evaluated in healthy macrophages, as well as in macrophages infected either with the reference strain LTB0016 or with the patient isolates VE98MR or VE2000MM and treated for 48 h with 18 and 22 (18: 1 and 2 µM; 22: 2 and 4 µM; respectively). Cells were further incubated for 1 h in the presence of cytochrome C-III (CYT-CIII 160 µM). By using the method of cytochrome reduction and the molar extinction coefficient of CYT-CIII, superoxide production, expressed in nanomoles of O2 − released by 2 × 10 5 cells, was determined. In healthy macrophages, betulin derivatives did not induce superoxide production. However, in LTB0016-and VE98MR-infected macrophages, the betulin derivative 18 induced superoxide production ( Figure 4). Supplementary Materials III details the associated statistics.

Superoxide Production (O 2 − ) in Macrophages
Additionally, the O 2 − was evaluated in healthy macrophages, as well as in macrophages infected either with the reference strain LTB0016 or with the patient isolates VE98MR or VE2000MM and treated for 48 h with 18 and 22 (18: 1 and 2 µM; 22: 2 and 4 µM; respectively). Cells were further incubated for 1 h in the presence of cytochrome C-III (CYT-CIII 160 µM). By using the method of cytochrome reduction and the molar extinction coefficient of CYT-CIII, superoxide production, expressed in nanomoles of O 2 − released by 2 × 10 5 cells, was determined. In healthy macrophages, betulin derivatives did not induce superoxide production. However, in LTB0016-and VE98MR-infected macrophages, the betulin derivative 18 induced superoxide production ( Figure 4). Supplementary Materials III details the associated statistics.
Although plant extracts and compounds against L. (L.) amazonensis [52][53][54][55] have been thoroughly evaluated, the betulin derivatives herein described have been tested against New World Leishmania spp. but not against patient isolates with therapeutic failure. In the present work, two L. (L.) amazonensis Venezuelan isolates, obtained from patients with therapeutic failure against SbV, were used to characterize the activity of betulin derivatives.
The corresponding promastigote GI50 against the reference compounds (AmB, PNT, SbIII, MLF) were defined. As pentavalent antimonials (SbV) constitute pro-drugs that need biological reduction to their trivalent form (SbIII) to be leishmanicidal [56], we used the latter as potassium antimonyl tartrate trihydrate instead of the pentavalent form, meglumine antimoniate.
Although plant extracts and compounds against L. (L.) amazonensis [52][53][54][55] have been thoroughly evaluated, the betulin derivatives herein described have been tested against New World Leishmania spp. but not against patient isolates with therapeutic failure. In the present work, two L. (L.) amazonensis Venezuelan isolates, obtained from patients with therapeutic failure against SbV, were used to characterize the activity of betulin derivatives.
The corresponding promastigote GI 50 against the reference compounds (AmB, PNT, SbIII, MLF) were defined. As pentavalent antimonials (SbV) constitute pro-drugs that need biological reduction to their trivalent form (SbIII) to be leishmanicidal [56], we used the latter as potassium antimonyl tartrate trihydrate instead of the pentavalent form, meglumine antimoniate.
Data not detailed in this manuscript suggest that the isolates' susceptibility against NaNO 2 was such that GI 50 was four-fold higher (VE98MR, GI 50 = 3.28 µM; VE2000MM, GI 50 = 4.79 µM) than the reference strain (LTB0016, GI 50 = 0.91 µM), thus suggesting that the Leishmania isolates, obtained from patients suffering therapeutic failure against meglumine antimoniate, can be classified as chemoresistant as has been described for antimonialresistant L. (L.) infantum isolated from Brazilian patients with therapeutic failure [57]. Additionally, the GI 50 against K 2 Cr 2 O 7 (>2.5 mM) for both isolates was significantly higher than that obtained for the reference strain (LTB0016 GI 50 = 1.06 mM); similar results have been published for SbV-resistant L. (L.) donovani tolerant to K 2 Cr 2 O 7 [58].
In the amastigote-macrophage model, the isolates (GI 50 : VE98MR = 1.76 µM, VE2000MM = 0.72 µM) were less susceptible to 18 than the reference strain LTB0016 (GI 50 : = 0.20 µM). Similar results were obtained for 22; the less susceptible isolate, VE2000MM, demonstrated a GI 50 of 1.69 µM, and the reference strain of 0.21 µM. VE98MR and VE2000MM might be less susceptible due to potential cross-resistance mechanisms between SbV and betulin derivatives as more than 3-fold doses were needed to reach similar effects in isolates vs. the reference strain.
These concentrations did not affect macrophages' viability (SI > 20), suggesting that 18 and 22 are worthy of being analyzed as potential candidates against parasites refractory to conventional treatment, as has been tested against L. (L.) donovani susceptible to SbV or selected as resistant to the same compound [60], as well as against Plasmodium falciparum (clone W2), resistant to chloroquine [61] for which betulin derivatives like O-disuccinoyl betulin, betulonic acid, and other betulinic acid derivatives have proved to be effective.
In L. (V.) braziliensis promastigotes, the most potent compound was 26 (3,28-diformylbetulin triazolodione) with a GI 50 < 1 µM [29]. Likewise, a betulin aldehyde derivative reduced infection up to 88% at 136 µM and 58% at 68 µM of intracellular L. (L.) amazonensis amastigotes [63]. Herein, 26 was also within the initially selected compounds. However, as previously mentioned, only 18 and 22 showed significant differences that made them worthy of further analysis. These results support our data regarding the moderate activity identified for betulinic aldehyde (12) that at 50 µM inhibited L. (L.) amazonensis growth by 67% and suggest that there might be species variances that could explain the differences herein found when compared to previous data in L. (V.) braziliensis.
Compound 22 is a heterocyclic adduct between 3,28-di-O-acetyllupa-12,18-diene and 4-ethylurazine, while the structurally related 25 is another heterocyclic adduct between 3,28di-O-acetyllupa-12,18-diene and 4-methylurazine [55]. Compound 25, although cytotoxic at all evaluated concentrations, is an effective derivative with a GI 50  The anti-leishmanial activity of betulin derivatives associates in Old World Leishmania with inhibitory effects over enzymes fundamental for parasite survival, such as topoisomerase IB [64] and trypanothione synthetase, TryS [65] as well as on the promoted parasite oxidative stress that ends up in programmed cell death [66]. However, specific targets and molecular mechanisms involved in their action still are not completely understood [67], and for New World species are not known at all. Herein, we explored their mechanism of action.
In organisms like protozoa, bacteria, pathogenic fungi, and tumor cells, betulin derivatives' target seems to be located at the plasma membrane [68]. Examples also exist in Leishmania that illustrate the effect of compounds on the membrane, specifically affecting the ∆Ψp, i.e., AmB in L. (L.) infantum [69], clomipramine and imipramine in L. (L.) donovani and L. (L.) major [70] or peptides like cecropin-A and melittin in L. (L.) donovani [71], or lupeol, an anti-leishmanial triterpene from Sterculia villosa for L. (L.) infantum [72], and L. (L.) donovani [73]. In this line of evidence, we report that betulin derivative 18, induced, in a very short time (15 min), the collapse of the plasma membrane potential (∆Ψp) in a concentration-dependent way. The effect was comparable to that imposed by valinomycin (1 µM) and cytochalasin B (300 µM).
Besides that, medicinal plants and their secondary metabolites may have immunomodulatory properties [74] [superoxide (O 2 − ) and nitric oxide (NO) production], potentially useful against leishmaniasis [75]. L. major amastigote elimination depends preferentially on NO production [75] and L. (V.) braziliensis abolition seems to strictly depend on both O 2 − and ROS generation [76]. Of note, in macrophages infected with L. (L.) donovani, lupeol induced NO production and increased the production of proinflammatory cytokines, decreasing the production of anti-inflammatory cytokines [73].
Nitric oxide is essential for macrophage activation and is produced in Leishmaniainfected macrophages [77,78]; additionally, metacyclic L. amazonensis promastigotes produce NO and axenic amastigotes produce NO levels 10-fold higher than those produced by metacyclic promastigotes as has been described by Acuña and collaborators (2017) [79] and references within. However, similar to what we herein described in both the reference strain and patient isolates and in L. amazonensis-infected macrophages, there is a lower production of NO than in un-infected macrophages. L. amazonensis seems to have an adaptive strategy to escape from host defense. This strategy seems to be produced by the activation of the NF-κB repressor complex p50/p50, a specific host transcriptional response that negatively regulates the expression of inducible nitric oxide synthase. This feature seems to favor the establishment and success of L. amazonensis infection [80].
In our hands, 18 and 22 were not capable of inducing NO production for any of the strains tested. However, compound 18 seems to influence the oxidative response of J774.1 macrophages infected with L. (L.) amazonensis LTB0016 and VE98MR, inducing an oxidative burst and thus increasing O 2 − production. Considering the structural similarity between lupeol and betulin derivatives, both terpenoids may be addressing the same cellular target. Molecular docking analysis [73] suggests that these compounds may be capable of attachment to the parasite surface protein GP63, thus affecting the cells' signaling processes that lead to macrophage activation [81]. On the other hand, an in silico analysis of a 58-member library of betulin derivatives (including 18) identified the membrane transport protein D1 as a potential target for this compound in L. (L.) donovani [67].
Finally, whether molecular functional events associated with the production of noncaspase-dependent apoptosis in L. (L.) donovani [64] and produced by O-disuccinoyl betulin [66] or inhibition of TryS [65]) are affected by betulin derivatives like compound 18 in L. (L.) amazonensis has to be further analyzed.
Altogether, our results suggest that diverse targets might be addressed by 18 in L. (L.) amazonensis. Therefore, further experiments are needed to fully elucidate the final mechanism of action of this compound.