In vitro antiplasmodial activity and identification, using tandem LC-MS, of alkaloids from Aspidosperma excelsum, a plant used to treat malaria in Amazonia
Abstract
Ethnopharmacological relevanceAspidosperma excelsum Benth. (Apocynaceae) is a native tree in the Brazilian Amazonia, where it is traditionally used to treat various diseases, including malaria.Aim of study:To investigate the chemical constitution of different samples obtained from A. excelsum by chromatographic techniques and to evaluate its antimalarial potential.Materials and methodsThe hydroethanolic extract and the alkaloid fractions from trunk bark of A. excelsum were tested against chloroquine-resistant Plasmodium falciparum (W2 strain), also their cytotoxicity was evaluated on HepG2 cells. Hyphenated chromatographic techniques, such as UPLC-DAD-ESI-MS/MS and HPLC-MS MicroTOF, were used to analyze different samples obtained, comparing their chemical composition to the literature data.ResultsAll the samples showed good activity against chloroquine-resistant P. falciparum and low cytotoxicity to HepG2 cells, indicating good selectivity to the parasite. Chromatographic techniques allowed the identification of 20 indole alkaloids, from which seven are here firstly described for this species. These results confirm the alleged anti-malarial use and disclose the A. excelsum bark as a promising source of a phytomedicine on the basis of the antiplasmodial activity of its alkaloids.
1.Introduction
Malaria remains one of the most prevalent infectious diseases worldwide and it is, therefore, a global health problem despite substantial efforts to control it over the past few decades. Approximately 3.3 billion people in 91 countries and territories are at risk of being infected with malaria and developing the disease. A total of 216 million malaria cases and 445.000 deaths were globally estimated in 2016 (WHO, 2017). In Brazil, the International Health Regulations (IHR) National Focal Point reported that between January and November of 2017, there were 174,522 malaria cases reported in the Amazon region, representing an increase in comparison to the same period of 2016 when 117,832 malaria cases were reported (PAHO/WHO, 2018).Malaria has been treated according to therapeutical schemes including quinine, chloroquine, primaquine, mefloquine and doxycycline. More recently, artemisinins, in combination with other anti-malarials (Artemisinin-based Combination Therapy – ACT) are recommended as the first choice treatment in several endemic countries. Resistance to existing anti-malarial drugs, including artemisinins, represent a major challenge in reducing mortality caused by Plasmodium spp, mainly by P. falciparum, the most virulent Plasmodium species, and reveals the need to discover a new generation of drugs to be used in the treatment of the disease (Dondorp et al., 2010; WHO, 2017). An alternative way to combat malaria is the scientific validation of herbal remedies traditionally used in endemic countries. Plants used in folk and traditional phytotherapy represent important resources for research on new treatments of malaria (Mariath et al., 2009 ; Willcox et al., 2011).Among the plant species with anti-malarial potential effects are those from Aspidosperma genus (Apocynaceae family), various of them being used in the Amazonian countries of the Central and South American (Brandão et al. 1992; Dolabela et al., 2012; Milliken and Albert, 1996; Oliveira et al., 2009). A literature survey reported the use of 24 Aspidosperma species to treat malaria/fevers and to 19 species that have had their extracts and/or alkaloids evaluated, with good results, for in vitro and/or in vivo anti-malarial activity.
Indole alkaloids are typical constituents of Aspidosperma species. However, only 20 out of more than 200 known indole alkaloids isolated from this genus have been assayed for anti-malarial activity (De Paula et al., 2014). These data support the potential of Aspidosperma species as sources of anti-malarials and the importance of research aiming to validating their traditional use in the treatment of human malaria. In Hispano-american and Brazilian Amazonia the tea prepared from trunk bark of Aspidosperma excelsum Benth., vernacular carapanauba, has an alleged use to treat liver disorder, fever and malaria (Pérez, 2002; Pinto and Barbosa, 2009; Oliveira et al., 2015).A total of 20 alkaloids is reported for this species (Banerjee et al., 1954; Benoin et al., 1967; Bolzani et al., 1987; De Paula et al., 2014; Pereira et al., 2007; Verpoorte et al., 1983). Alkaloids are possibly related to those indications, in particular to the anti-malarial activity, which is described for several species of the genus (Bourdy et al., 2004; Mesquita et al., 2007; Oliveira et al., 2009; Weniger et al., 2001). The presence of indole alkaloids in A. excelsum and the reported anti- parasitic activity of other species of the same genus motivated the present investigation. A bark ethanol extract of a specimen collected in Peru, where it is known as Remo caspi, showed a moderate growth inhibition (IC50 42 μg/mL) againstP. falciparum chloroquine-sensitive strain (3D7) and no cytotoxicity to phytohaemagglutinin (IC50 > 100 μg/mL) (Kvist et al., 2006). No evident signs of acute toxicity in oral dose of 5,000 mg/mL were observed for a tincture (EtOH 70 % extract) that disclosed a good in vitro activity against P. falciparum chloroquine resistant strain (W2) (Gomes, 2011).
2.Experimental
The botanical data (plant name, synonymy and authority) of A. excelsum were initially accessed, in 2009, using Tropicos database (http://www.tropicos.org), later, in 2014 and 2018, these data were checked again consulting The Plant List (http://www.theplantlist.org) as recommended by Elsevier.Trunk bark of species was collected from a specimen in the Santa Luzia forest, municipality of Acará, Pará State, Brazil (01°31’ S and 48°25’ W), in August, 29, 2009. The plant material was identified by Dr. Mário Augusto G. Jardim, botanist at the Museum Paraense Emilio Goeldi (MPEG), as Aspidosperma excelsum Benth (Apocynaceae) by comparison of a herborized material with authentic sample deposited in MPEG herbarium under the numbers MG 15,470; MG 15,860 and MG 82,422.A tincture of A. excelsum trunk bark was prepared, by maceration, from 850 g of the bark powder with 70 % ethanol, in a proportion of 1:10 (w:v), and an aliquote of 8,5 L concentrated in a rotatory evaporator under reduced pressure until complete removal of ethanol. The aqueous residue was then frozen and lyophilized, yielding approximately 70 g of ELAe (Ethanol/water Lyophilized Aspidosperma excelsum extract).In order to evaluate the alkaloid profile of A. excelsum trunk bark, five different procedures were performed. In the first one (figure 1A) four aliquots of an acidic extract of ELAe (5 g) in 5 % HCl (200 mL, pH 1) were extracted with dichloromethane at pH 1 following alkalinization to pH 4, 7, 10, with ammonium hydroxide, after a previous treatment with 50 mL n-hexane to remove non-alkaloidal components. Then, each sample was sequentially extracted at increasing pH values according to the Stas-Otto method, modified (Auterhoff and Kovar, 1985). Fractions at pH 7 (Fr7) and 10 (Fr10), totalizing 159 mg, were separated by middle pressurechromatography in Gilson Pump model 305, column 31×3 cm on normal phase silica gel <125 μm, flow rate of 2 mL/min and a gradient initiated with hexane/AcOEt (50:50) + 3% diethylamine – DEA, followed by AcOEt + 3% DEA, AcOEt/MeOH (75:25) + 3% DEA, AcOEt/MeOH (50: 50) + 3% DEA and, finally, MeOH + 3% DEA).
Two samples obtained (95 mg) afforded yohimbine (5.5 mg) by preparative TLC silica gel eluted with AcOEt-MeOH (75:25).In the second procedure, alkaloids acid-base extraction of ELAe (Figure 1B) was carried out according to the classical Stas-Otto method (Auterhoff and Kovar, 1985). Briefly, ELAe (40 g) as suspended in 1 L of 5 % aq. HCl, the suspension was sonicated for 10 min, then filtered to remove insoluble material, extracted with n- hexane (2x200 mL) to remove non-alkaloidal components and then with dichloromethane (3x200mL). Afterwards, the pH of the acid aqueous solution (pH 1) was successively adjusted with ammonium hydroxide to pH 4, 7 and 10, and after each pH adjustment, extractions with dichloromethane (2x300 mL) were performed. The dichloromethane solutions extracted from pH 1, 4, 7 and 10 were dried over anhydrous sodium sulfate and then concentrated in a rotavapor affording ELAe- FAlk1 (0.338 g), ELAe-FAlk4 (0.313 g), ELAe-FAlk7 (2.906 g), and ELAe-FAlk10 (0.102 g).In the third procedure (C), the hydroethanolic extract (ELAe) (1.9 g) was suspended in 2 N HCl (200 mL) and then extracted with dichloromethane (2x100 mL) to remove non-alkaloids. The acid aqueous solution (pH 1) was alkalinized with ammonium hydroxide to pH 10 and extracted with dichloromethane (4x100 mL) affording the total alkaloids from A. excelsum ELAe-Alk (0.136 g, 7.1%)In the fourth procedure (D), trunk bark powder (50 g) was extracted by maceration with 2 N HCl (2x200 mL). The aqueous acid solution (pH 1) was extracted with dichloromethane to remove non-alkaloids, the aqueous acid solution was alkalinized to pH 10 with ammonium hydroxide and extracted with dichloromethane (4x150 mL) giving total alkaloids Ae-Alk (0.468 g, 0.93%).In the fifth procedure (E), trunk bark powder (160g) was extracted by maceration with 2 N HCl (2x200 mL).
The aqueous acid solution (pH 1) was extracted with dichloromethane to remove non-alkaloids. Afterwards, the solution was alkalinized to pH 10 with ammonium hydroxide and submitted to sequential extractions with dichloromethane for three times, each time with 2x200 mL, these fractions were dried with anhydrous sodium sulfate, filtered and finally concentrated to residue in a rotavapor affording fractions Ae-Alk1 (0.385 g, 0.24%), Ae-Alk2 (0.111 g, 0.06%), Ae-Alk3 (0.200 g, 0.13%).Fig. 1. Alkaloids separation by acid-base extractions and isolation of yohimbine from ELAe, where A and B: SOm: Stas-Otto method, C: Total alkaloids extraction from ELAe at pH10, D: Total alkaloids extraction from trunk bark at pH10, E: Fractional alkaloids extraction from trunk bark at pH10Silica gel TLCs (Merck®) of the tincture (ELAe) as well as of all fractions obtained (Figure 1) were run with AcOEt/MeOH (75:25) and CHCl3/MeOH/NH4OH (85:15:0.2) as mobile phases, the chromatograms were observed under UV light at 254 and 365 nm and were sprayed with Dragendorff reagent for alkaloids detection.Fingerprints of the dried tincture (ELAe) and alkaloid fractions (FAlk 1, 4, 7, 10, Ae-Alk and Ae-Alk1-3) were registered in a HPLC-DAD Waters 269 apparatus equipped with a UV-DAD detector (Waters 2996). A LiChrospher 100 RP-18 column (5 μm, 250 × 4 mm i.d.; Merck, Darmstadt, Germany) was employed at 40 °C, flow rate of 1.0 mL/min and detection at wavelengths of 220, 280 and 350 nm. Aliquotes of dried extract (10.0 mg/mL) and fractions (1.0 mg/mL) solutions prepared with HPLC grade methanol was sonicated for 15 min and then centrifugated at 10,000 rpm for 10 min. The supernatant was filtered through a Millipore membrane (0.2 μm) and injected (10.0 μL) onto the equipment. Elution was carried out with a linear gradient of water + 0.1 % phosphoric acid (A) and acetonitrile + 0.1 % phosphoric acid (B) (from 5 % to 95 % of B in 60 min), followed by an isocratic elution (95 % ACN from 60 to 65 min) and then returning to the initial elution at 65-70 min (5 % ACN).Analyses were performed at Faculty of Pharmacy, UFMG, Belo Horizonte, Brazil, in an UPLC Acquity (Waters) ion trap mass spectrometer equipped with an atmospheric pressure chemical ionization (APCI) interface operated in the following conditions: positive and negative ion mode; capillary voltage, 3500 V; capillary temperature, 320 °C; source voltage, 5 kV; vaporizer temperature, 320 °C; corona needle current, 5 mA and sheath gas, nitrogen, 27 psi. Analyses were run in the full scan mode (100-1500 u). The UPLC-ESI-MS/MS analyses were performed in an UPLC Acquity (Waters) with helium as the collision gas, and the collision energy was set at 40 eV. Chromatographic separation was done on an column ACQUITY UPLCTM BEH (1.7 μm, 50 x 2 mm i.d.) (Waters).
The mobile phases consisted of water 0.1 % formic acid (solvent A) and acetonitrile 0.1 % formic acid (solvent B). The elution protocol was 0-11 min, linear gradient from 5% to 95% B. The flow rate was0.3 mL min-1, and the sample injection volume was 4.0 μL. The UV spectra were registered from 190 to 450 nm. Mass spectrometry. Analysis was performed on a quadrupole instrument fitted with an electrospray source in the positive mode. Ion spray voltage: -4 kV: orifice voltage -60 V.ELAe at 9 mg/mL was analysed at the Faculty of Pharmacy, UFPA, Belém, Brazil, in an UHPLC-Triple Quadrupole equipment Agilent®, employing a Zorbax® C 18 – 3,5 µm (50 x 2.1 mm) column and as mobile phase; H2O-Formic acid at pH 3 (solvent A) and Acetonitrile-Formic acid 0.1 % (solvent B), in isocratic elution mode for 20 minutes (40 % A and 60 % B), under the following conditions: column temperature at 27 °C (± 1 °C), flow rate of 0.4 mL/min and injection volume 2 µL. Chromatograms were registered at 280 nm. Spectrograms were registered in an Agilent 6400 Series Triple Quadrupole with a Dual ESI source using Agilent Jet Stream Technology (AJS ESI+), Nitrogen Gas, Dry Gas: 9 L/min, Capillary: 4500V, Nebulizer: 27.6 psi, Dry Temp: 300 °C.ELAe-Alk (total alkaloids) was analysed at Central Analítica, USP, São Paulo, SP, Brazil, in a HPLC equipment CBM-20A Shimadzu, with pumps LC-30AD Shimadzu, detector SPD-20A Shimadzu, oven CTO-20A at 40 °C Shimadzu and autoinjector SIL 30AC Shimadzu, column 12.5 x 4.5 cm, LiChroCART ® RP 18 – 5 µm, mobile phases were H2O-Formic acid 0.1 % (solvent A) and Acetonitrile-Formic acid 0.1 % (solvent B), with elution in a linear gradient (0-11 min), starting with 5 to 95 % B. Conditions: column temperature at 27 °C (± 1 °C), flow rate of 0.3 mL/min. Chromatograms were registered at 220 nm, 280 nm and 350 nm. Spectrograms were registered in Maxis 3G - Bruker Daltonics, Capillary: 4500V, Nebulizer: 27.6 psi, Dry Gas: 9 L/min, Dry Temp: 300 °C, ESI+.In vitro cytotoxicity was evaluated against HepG2 cells derived from a primary human hepatoblastoma (Mosmann, 1983; Varotti et al., 2008). Initially, concentration of cells suspension was adjusted to 4x105, which was transferred to a 96 wells plates (100 µL in each well) and then incubated under CO2, at 37 °C for 24 h; 100 µL of the complete medium containing the test samples at 1 µg/mL, 10 µg/mL, 100 µg/mL and 1000 µg/mL were added to each well, in triplicate, incubating again for 24 h. Following, 18 µL of MTT, at 20 mg/mL, are added to each well.
After 90 min incubation with MTT, 100 µL DMSO were added to each well and the absorbance at 492 nm registered in a spectrophotometer (Denizot and Lang, 1987). Calculation of inhibition rate of cell growth uses the recorded absorbance values were 100% absorbance is observed in wells containing cells cultured without the presence of any sample. The results were tabulated using Origin 8.0, determining dose-response curves with sigmoidal fit. Cytotoxic concentrations that inhibit 50% of the cell growth (CC50) were determined in relation to controls without sample.In vitro evaluation of the anti-malarial activity was carried out using erythrocytes infected with Chloroquine resistant P. falciparum (W2), by the LDH method (Makler et al., 1993). P. falciparum W2 strain was maintained in continuous culture on human erythrocytes (blood group A+) in RPMI medium supplemented with 10% human plasma (complete medium), as previously described (Trager and Jensen, 1976). Synchronization of the parasites was achieved by Sorbitol treatment (Lambros and Vanderberg, 1979) and the parasitemia was determined microscopically in Giemsa-stained smears.A preliminary screening was performed: aliquots of 20 µL of a DMSO solution containing 25 and 50 µg/mL, respectively, of each sample were added to 180 µL of a suspension of infected erythrocytes (1 % hematocrit, 2 % parasitemia). This procedure was performed in triplicate in a 96-wells microplate. Chloroquine and a suspension of infected erythrocytes were used as positive controls and uninfected erythrocytes were the negative control.
The 50% in vitro anti-malarial effect (IC50) of test samples and controls wasmeasured by the pLDH assay as described previously (Makler et al., 1993), with minor modifications. Briefly, ring-stage parasites in Sorbitol-synchronized blood cultures (Lambros and Vanderberg, 1979) were added to 96-well culture plates at 2% parasitemia and 1% hematocrit and then incubated with the test drugs that were diluted in complete medium, from 50 mg/mL stock solutions in DMSO, to a final concentration of 0.002% (vol/vol) and stored at -20°C. After a 48h incubation period the plates were frozen (-20°C for 24 h) and thawed for the pLDH assay. The hemolyzed cultures were transferred to another 96-wells culture plate with simultaneous addition of Malstat® and NBT/PES reagents. After 1h of incubation at37 ºC in the dark, the absorbance was read at 570nm in a spectrophotometer (Infinite®200 PRO, Tecan). Statistic analyses of the results were performed with the software Microcal Origin 8.5 for determination of the dose-response curves plotted with sigmoidal fit (Varotti et al., 2008). The 50% inhibitory concentration growth of the parasites (IC50) was determined by comparison with positive control (and without drugs.According to the IC50 values, the samples were classified as very active (< 1 µg/mL), active (1-15 µg/mL), moderately active (15.1-25 µg/mL), weakly active (25.1- 50 µg/mL) or inactive (> 50 µg/mL). The selectivity index (SI) was calculated by the ratio between the CC50 in HepG2 cells and the IC50 against P. falciparum (SI = CC50/IC50).
3.Results and discussion
A tincture (ELAe, EtOH-water 70:30) and alkaloid extracts were obtained fromA. excelsum trunk bark, according to Figure 1. In the first one (Figure 1A), a small ammount of the tincture (5 g, hydroethanolic extract, ELAe) was submitted to alkaloid extraction by the Stas-Otto method, modified, followed by middle pressure chromatography and preparative TLC (AcOEt-MeOH 75:25) allowing to obtain yohimbine (5.5 mg), whose characterization was based on spectroscopic data compared to those published for an authentic sample of the same substance (Nascimento, 2013).The Stas-Otto method, modified, (Auterhoff and Kovar, 1985) was performed (Figure 1B) with higher amount of ELAe (40 g), as described in the experimental section, affording alkaloidal fractions FAlk1, FAlk4, FAlk7, FAlk10 in which the numbers are related to the pH of the aqueous phases from which each one was extracted. This procedure can afford selective extraction of alkaloids according to their basicity (Auterhoff and Kovar, 1985; Nascimento, 2013). Therefore, lower yield of alkaloids for FAlk1 and FAlk4 might be expected since at these pH values, most of the alkaloids should be in their salt form and, therefore, would be insoluble in dichloromethane. Higher alkaloid contents would be expected for FAlk7 and FAlk10 in which total and non-phenolic alkaloids would be in the free basic form and, therefore, more soluble in dichloromethane. Indeed, the yield was higher for FAlk7: FAlk1 (0.338 g, 0.84%), FAlk4 (0.313 g, 0.78%), FAlk7 (2.906 g, 7.26%), and FAlk10(0.102 g, 0.25%). The third procedure (Figure 1C) afforded a relatively high yield of the total alkaloids extract separated by acid-base extractions from ELAe (ELAe-Alk, 7.1%).
In the fourth procedure (Figure 1D), a total alkaloids extract was obtained directly from the trunk bark by extraction with dil. HCl followed by alkalinization with ammonium hydroxide to pH 10 and extraction with dichloromethane (Ae-Alk 0.468 g, 0.93%). The fifth procedure (Figure 1E), started also by direct acid extraction of the trunk bark but alkaloid fractions were separated by sequential extraction with dichloromethane, aiming to separate alkaloids according to their solubility in dichloromethane, affording Ae-Alk1 (0.385 g, 0.24%), Ae-Alk2 (0.111g, 0.06%) and Ae-Alk3 (0.20g, 0.13%).The reason for separation of alkaloidal fractions by different procedures was to obtain alkaloid mixtures with possible different compositions, to evaluate their anti- malarial activity and phytochemical composition, as a strategy aiming to develop phytopreparations containing alkaloids that are supposed to be responsible for the observed anti-malarial activity of A. excelsum. Analyses of these fractions by hyphenated chromatographic techniques (HPLC-DAD, UPLC-DAD-ESI-MS/MS) would allow to meet the quality standards required by drug authorities along with the guarantee of good clinical practice (GCP), conforms to clinical trials requirements, as recomended by Wagner (2004). The first steps in this direction are here reported. Therefore, the aiming was to answer the following questions: 1- Would a tincture (ELAe, hydroethanolic extract) of A. excelsum trunk bark, that usually contains alkaloids and non-alkaloids, be effective as anti-malarial? 2- Would alkaloid fractions prepared by different procedures disclose distinct anti-malarial activity? 3- Would it be possible to have an efficient anti-malarial alkaloidal fraction, which could become a candidate to the development of a phytomedicine? Besides these questions, and taking into account the need to develop more accessible phytotherapeuticals, simple and low cost methodologies like acid-base extractions, were performed.TLC analyses of the tincture (ELAe) and alkaloid fractions revealed the presence of reactive zones to the Dragendorff reagent in all samples and predominantly in FAlk 7 and FAlk 10.
The UHPLC-DAD profile of the tincture (ELAe) at 280 nm (Figure 2A) is rather simple disclosing one major peak (RT 4.15 min, 63.75% calc. area) besides less intense peaks, such as RT 2.49 min (4.49 % area), 3.17 min (11.2 % area), 4.55 min (5.22 % area), 5.57 min (4.56 % area) and 5.73 min (4.87 % area). The UV spectrum (λmax 216, 238, 276 nm) registered online for the major peak indicates that it corresponds to an indole alkaloid (Sangster and Stuart, 1965). The peak at RT 4.55 min (5.22% area) was attributed to yohimbine what was confirmed on basis of the online registered UV spectrum (λmax 221, 272 nm), co- injection with a reference sample under the same chromatographic conditions (data not shown) and LC-MS that gave the expected molecular mass of [M+H]+ 355. Furthermore, yohimbine was isolated and characterized by TLC direct comparison with a reference sample. The presence of yohimbine in A. excelsum (Benoin et al., 1967; Verpoorte et al., 1983) as well in other Aspidosperma species, such as A. oblongum (Palmer, 1964), A. ramiflorum (Marques et al., 1996) and A. ulei (Torres et al., 2013) was previously reported.Aiming to identify components of ELAe (tincture, hydroethanolic extract) and alkaloid fractions, by UPLC-ESI-MS, in a reasonable time, a high gradient slope was used, with formic acid in the mobile phase with UV detection at 280 nm. The proposed method was deemed acceptable, and adequate the objective. The analyses were then carried out by reversed-phase ultra high performance liquid chromatography coupled to tandem mass spectrometry (UPLC-ESI-MS/MS), a variant of HPLC that affords significant advantages in resolution, speed, and sensitivity for analytical determinations, particularly when coupled to mass spectrometers capable of high-speed acquisitions, allowing studies on the fragmentations of each component under electrospray conditions (Churchwell et al., 2005) and comparison with previously reported data on MS fragmentation for indole alkaloids (Aguiar et al., 2010; Joule et al. 1965; Li-Mei et al., 2015).
The intense peak at 4.15 min (63,75 % area) in the ELAe chromatogram by UHPLC-DAD (Figure 2A) corresponds to the major constituent that was identified as corynan-17-ol (dihydrocorynantheol, 3) on the basis of its [M+H]+ 299 along with its fragmentation pattern by UPLC-ESI-MS/MS and by HRMS (Table 1). Figure 2B represents the UPLC-DAD for ELAe and its spectrogram (TIC) by UPLC-ESI-MS is shown in Figure 2C. Minor alkaloids, including yohimbine (14), were also identified (Table 1). A marked difference is observed between the chromatograms registered by UHPLC-DAD and UPLC-DAD, the last one detecting minor components and, therefore, giving much more information on the composition of complex matrices such as plant extracts, as can be clearly seen in Figures 2A and 2B. Total ions chromatograms (TIC, full scan mass spectra) were registered for ELAe (Figure 2), and all alkaloid fractions in positive and negative ion modes. Sensitivity was greater for chromatograms in positive than in negative ion mode, in which few peaks were observed and, as consequence, subsequent analyses were performed in the positive ion mode. These data coupled to exact molecular mass determined by HRMS (UPLC-ESI-MS MicroTOF) for ELAe, fragmentation patterns by tandem mass spectrometry (UPLC-ESI-MS/MS) and to literature data on known alkaloids allowed the identification of 20 indole alkaloids (Tables 1 and 2) whose structures are shown in Chart 1.Previous reports on A. excelsum phytochemical studies described the isolation and identification of 16 alkaloids: 1, 6, 7, 9, 11, 12, 13, 15, 16, 18, 19, 21-25 (Benoin et al., 1967; Bolzani et al., 1987 ; Verpoorte et al., 1983). However, eight of these (1- 5, 10, 14 and 20), were not detected in the present analysis and primary references were not found for cytations on 12 and 13 (Tables 1 and 2, Chart 1). Attention is called for the detection of seven alkaloids that are here firstly reported for A. excelsum although they have been isolated from other Aspidosperma species: geissoschizol (2), corynan-17-ol (3), kopsanol (or epikopsanol) (4), condylocarpine(5),yohimban-16-carboxylic acid (10), aspidocarpine(16) and 6,7-dihydroobscurinervine (22) (Tables 1 and 2, Chart 1) (Pereira et al., 2007).UPLC-DAD chromatograms and spectrograms registered by UPLC-ESI-MS ofA. excelsum trunk bark alkaloidal fractions (FAlk1,4,7,10; ELAe-Alk, Ae-Alk and Ae- Alk1-3) allowed to identify the major alkaloids and several of the minor ones in each sample.
Here are shown the UPLC-DAD profiles and the mass spectrograms (TIC) for ELAe (tincture, Figure 2) and ELAe-Alk (total alkaloids obtained by acid-base extraction from ELAe, Figure 3). The most intense peak in the ELAe and ELAe-Alk chromatograms (Figures 2 and 3) correspond to corynan-17-ol (dihydrocorynantheol) (3, Chart 1) based on UV data, molecular masses, fragmentation patterns and HRMS (Table 1, Chart 1). LC figures for other samples (FAlk1,4,7,10) are available as supplementary data. All the samples presently assayed against P. falciparum chloroquine resistant W2 strain caused parasite growth inhibition > 50 % in the concentrations of 50 and 25 µg/mL, excepting Ae-Alk3 (Table 3). Most of the presently evaluated samples disclosed IC50 < 25 µg/mL, excepting Ae-Alk3 whose IC50 was not determined because they disclosed low percentage of parasite growth inhibition in the primary screening. The hydroethanolic extract (tincture, ELAe) disclosed a moderate activity (IC50 23.68 ± 3.08 µg/mL) and the total alkaloids fraction derived from it (ELAe-Alk) was more than two times active (IC50 9,93 ± 1,28 µg/mL), that is an indication of a significant contribution of alkaloids for the antiplasmodial activity. On the other hand, alkaloid fractions obtained from ELAe by the classical Stas-Otto extraction procedure (Figure 1, FAlk1, 4, 7, 10) disclosed higher IC50 values, in the range of 13.10 ± 1.20 to 18.52 ± 0.60 µg/mL. For alkaloid fractions obtained directly from the trunk bark by acid-base extractions (Ae-Alk, Ae-Alk1-3) the IC50 values ranged from 8,75 ± The major component of the hydroethanolic extract (tincture, ELAe) (Figure 2B) is corynan-17-ol (dihydrocorynantheol, 3) and the following other alkaloids were identified on it: geissoschizol (2), kopsanol/epi-kopsanol (4), tubotaiwine (6), N- acetilaspidospermidine (7), demethylaspidospermine (9), yohimbine (14), aspidocarpine (16), aricine (17), and excelsinine/10-methoxy-yohimbine (18/19). The presence of the known anti-malarial alkaloids demethylaspidospermine (9) (Mitaine- offer et al., 2002) and aspidocarpine (16) (Andrade-Neto et al., 2007; Chierrito et al., 2014) might explain ELAe activity (IC50 23.68 ± 3.08 µg/mL). Alkaloids peak areas can be better related to the content of each alkaloid in the total alkaloids fraction (ELAe-Alk) (Figure 3A) obtained from ELAe, since it practically does not (or should not) contain non-alkaloids. ELAe-Alk disclosed higher activity (IC50 9.93 ± 1.28 µg/mL) than the tincture (ELAe) (IC50 23.68 ± 3.08 µg/mL) what points to alkaloids as responsible for the observed anti-malarial effect. Geissoschizol (2) and corynan-17-ol (3) are the major constituents of ELAe-Alk (Figures 3A and 3B). Minor alkaloids identified in this fraction are yohimbine (14), aspidocarpine (16), aricine (17), excelsinine/10-methoxy-yohimbine (18/19), 6,7- dihydroobscurinervine (22), and ochrolifuanine (21) (Table 1, Chart). Peaks attributed to two unidentified alkaloids, [M+H]+ 295 and 293, possibly related to geissoschizol (2), are seen in the UPLC-ESI-MS MicroTOF spectrogram (Figure 3B). Aspidocarpine (16) and ochrolifuanine (21), with reported IC50 of 19 nM and 100 - 500 nM, respectively, (Andrade-Neto et al., 2007, Frederich et al., 2002; Frederich et al., 2008) might explain the anti-malarial effect of ELAe-Alk (IC50 9.93 ± 1.28 µg/mL) that was obtained from the tincture (IC50 23.68 ± 3.08 µg/mL) (Table 3). Ae-Alk2, whose in vitro anti-malarial activity (IC50 8.75±2.26 µg/mL) is statistically equivalent to that one of ELAe-Alk (IC50 9.93 ± 1.28 µg/mL), concentrates geissoschizol (2), corynan-17-ol (3), yohimbine (14) and 6,7-dihydroobscurinervine (22) (Figures 4A and 4B) and contains also the very active alkaloids aspidocarpine (16) and ochrolifuanine (21).Alkaloid fractions obtained by the sequential Stas-Otto procedure can clearly be considered in two groups: FAlk1 and FAlk4 that were extracted from acid aqueous condition (pH 1 and pH 4, respectively) were more active than FAlk7 and FAlk10, and statistically equivalents (IC50 13.10 ± 1.20 and 12.03 ± 0.48 µg/mL, respectively). FAlk7 and FAlk10 with IC50 18.30 ± 1.91 and 18.52 ± 0.60 µg/mL, respectively, are also statistically equivalents. Similar UPLC-DAD profiles were registered for FAlk1 and FAlk4, both of them containing as major constituent an unidentified alkaloid with [M + 1]+ 427, followed by the highly active aspidocarpine (16) and the moderately active yohimbine (14) (Robert et al., 1983) besides corynan-17-ol (3). FAlk7 and FAlk10 show similar UPLC-DAD with corynan-17-ol (3) and aspidocarpine (16), the last one being highly active (IC50 0.019 ug/mL) against P. falciparum K1 strain and therefore might be responsible for the activity of both of these fractions (supplementary material).Considering that alkaloids are expected to be the active compounds in Apocynaceous plant species (De Paula et al., 2014; Pereira et al., 2007) the lower effect of ELAe (IC50 23.68 ± 3.08 µg/mL) might be related to the usual low content of alkaloids in a tincture/hydroethanolic extract than in an alkaloid extract. In the present case, the content of alkaloids in the ELAe (tincture) is estimated to be approximately 7.1% that is the alkaloids yield to afford ELAe-Alk, as described in the third extraction procedure. Furthermore, the ELAe spectrogram shows that this peak corresponds to two alkaloids with [M+H]+ 297.38 and 299.34 that were identified as geissoschizol (2) (or 3-epigeissoschizol) and corynan-17-ol/dihydrocorynantheol (3, Chart 1, Table 1). Minor alkaloids identified were kopsanol/epikopsanol (4), tubotaiwine (6), N- acetylaspidospermidine (7), demethylaspidospermine (9), yohimbine (14), aspidocarpine (16), excelsinine or 10-methoxy-yohimbine (18/19) (Table 1, Chart 1).In summary, three highly anti-malarial alkaloids, namely aspidocarpine (14), demethylaspidospermine (9) and ochrolifuanine (21) were identified but it seems that anti-malarial data are not available for the other alkaloids occurring in A. excelsum. ELAe-Alk and Ae-Alk2 disclosed favourable values for IC50, CC50 and SI (Table 3). Considering the antiplasmodial activity, Ae-Alk2 can be highlighted as the most active herbal preparation and its obtention is simpler and less expensive, but its toxicity still needs to be evaluated. Previous testing of acute oral toxicity of the tincture (ELAe) was carried out according to the OECD Fixed Dose Procedure when no obvious signs of toxicity were observed in the dose of 5,000 mg/mL (Gomes, 2011). Finally, it should be mentioned that A. excelsum Benth., the accepted name according to Tropicos (http://www.tropicos.org/Name/1800114?tab=synonyms), has as synonym A. nitidum Benth. ex Müll. Arg. A recent evaluation of trunk bark extracts from a specimen collected in Manaus, state of Amazonas, Brazil, were active againstP. falciparum (W2) and showed a low in vitro cytotoxicity against HepG2 cells, whereas the leaves, branch extracts and the pure alkaloid braznitidumine were inactive. The authors raised the possibility that a mixture of alkaloids detected in the active fractions of the crude extract of this species are probably responsible for the anti-malarial activity (Coutinho et al. 2013). However, the identification of these alkaloids was not reported. 4.Conclusion The anti-malarial activity of A. excelsum against chloroquine resistant P. falciparum W2 strain contributes for the validation of its use in the Amazonian folk phytotherapy to treat malaria and reinforces the potential of Aspidosperma species from Amazonia to provide new therapeutic options for the treatment of malaria. A total of 20 alkaloids were identified in the A. excelsum trunk bark. Although phytochemically investigated, seven out of the presently identified alkaloids were not previously reported for this species and, surprisingly, the major alkaloid, corynan-17-ol (3), is among these. Several unsuccessfull attempts to isolate this alkaloid were performed what might be explained by the occurrence of several minor alkaloids with close RTs, as can be observed in the LC chromatograms, and that co-eluted even by preparative TLC (data not shown). Therefore, the present work is a demonstration of the relevance of LC analysis coupled to spectroscopic analytical methods in the investigation of bioactive extracts. So far, only yohimbine (14) was presently isolated and was shown to be moderately active. Previously reported highly active alkaloids, namely demethylaspidospermine (9), aspidocarpine (16), and ochrolifuanine (21) were identified and they must contribute for the observed in vitro antimalarial activity of the alkaloid extracts. These findings allowed to answer the three questions posed in the final part of results and discussion: 1- Yes, the trunk bark tincture (ELAe) showed a reasonable in vitro anti-malarial activity against a chloroquine resistant P. falciparum strain supporting, in part, the popular use of A. excelsum to treat malaria in the Amazonia. 2- Total alkaloids extract obtained from the tincture (ELAe-Alk) and the alkaloid fraction Ae-Alk2 obtained directly from trunk bark pointed out for their activity (IC50 < 10 µg/mL). 3- Ae-Alk2 should be highlighted as the most promising out of the assayed samples because its preparation is simpler and less expensive. Further investigations must be performed aiming the development of an efficient and safe phytomedicine with rigorous control of composition what is nowadays carried out by LC-MS procedures. In vivo anti-malarial evaluation and non Itacnosertib clinical toxicity studies are the next steps in this direction. Efforts to isolate active constituents for anti- malarial evaluation and quantifying the herbal preparations for quality control are on progress.