HPLC method for quantifying verbascoside in Stizophyllum perforatum and assessment of verbascoside acute toxicity and antileishmanial activity

We report the chemical composition of the crude leaf extracts obtained from Stizophyllum perforatum (Cham.) Miers (Bignoniaceae), a simple high-performance liquid chromatography–diode array detection (HPLC-DAD) method based on mangiferin as an internal standard to quantify verbascoside, and the verbascoside acute oral toxicity and antileishmanial activity. HPLC–high-resolution mass spectrometry–DAD (HPLC–HRMS–DAD) analyses of the crude ethanol S. perforatum leaf extracts (CE-1 and CE-2) revealed that verbascoside was the major constituent in both extracts. CE-1 was purified, and verbascoside and casticin, among other compounds, were isolated. The developed HPLC-DAD method was validated and met the required standards. Investigation of the CE-2 acute toxicity indicated a lethal dose (LD50) greater than 2,000 mg/kg of body weight. Both CE-1 and CE-2 exhibited antileishmanial activity. The isolated compounds, verbascoside and casticin, also displayed antileishmanial activity with effective concentrations (IC50) of 6.23 and 24.20 µM against promastigote forms and 3.71 and 18.97 µM against amastigote forms of Leishmania amazonensis, respectively, but they were not cytotoxic to J774A.1 macrophages. Scanning electron microscopy of the L. amazonensis promastigotes showed that the parasites became more rounded and that their plasma membrane was altered in the presence of verbascoside. Additionally, transmission electron microscopy demonstrated that vacuoles emerged, lipids accumulated, kinetoplast size increased, and interstitial extravasation occurred in L. amazonensis promastigotes exposed to verbascoside. These findings suggest that S. perforatum is a promising candidate for further in vivo investigations against L. amazonensis.


Introduction
Stizophyllum perforatum (Cham.)Miers (Bignoniaceae), synonyms Bignonia perforata Cham.and Bignonia physaloides Cham., is a liana that grows in Mexico, Guatemala, Costa Rica, Panama, Guyana, and Brazil.The genus Stizophyllum comprises only three species, S. perforatum, Stizophyllum inaequilaterum, and Stizophyllum riparium, which are morphologically similar and occur in wet to dry forests and disturbed vegetation from Mexico to southern Brazil.Interestingly, Stizophyllum species are pioneers and are typically found in disturbed areas, forest margins, and secondary vegetation (Lohmann and Taylor, 2014;Beyer et al., 2017).Additionally, the crude leaf extract obtained from S. perforatum has been shown to exhibit trypanocidal activity (Silva et al., 2014).
Triterpenes and pregnane derivatives with cytotoxic action against P-388 lymphocytic leukemia have been isolated from S. riparium whole plant extract (Duh et al., 1987;Duh et al., 1991).In Peru and Panama, S. riparium is used as folk medicine to treat snakebites (Valadeau et al., 2010) and to dilate atrophied urethras in cases of anuria (Gupta et al., 1993).However, no information regarding the isolation of compounds from S. perforatum or its antileishmanial activity has been published.
To the best of our knowledge, no internal standard-based highperformance liquid chromatography-diode array detection (HPLC-DAD) method is available for quantitatively determining verbascoside in S. perforatum.Here, we validated such a method by following the Brazilian Health Surveillance Agency regulations.We also investigated the antileishmanial activity of verbascoside against Leishmania amazonensis, as well as the changes that verbascoside elicits in L. amazonensis promastigotes.Finally, we assessed toxicological aspects by evaluating acute oral toxicity.

Preparation of extracts, stock solution, standard solutions, and samples
The S. perforatum leaf materials from the first and second harvests were dried at room temperature and powdered.The powders (110 g and 665 g) were separately macerated with ethanol (400 mL × 3) through three consecutive extractions, at room temperature; each extraction lasted 72 h.The filtrates were concentrated under reduced pressure, which yielded 31 g and 43 g of the crude ethanol S. perforatum leaf extracts labeled CE-1 and CE-2, respectively.
Samples were prepared from CE-1 or CE-2 by weighing 1.5 mg of extract and dissolving it in 1 mL of methanol in an ultrasonic bath for 10 min.Next, the solutions were centrifuged in an Eppendorf centrifuge at 5,000 rpm for 5 min and transferred to vials.

HPLC-HRMS analyses
CE-1 and CE-2 at 2 mg/mL were dissolved in methanol and analyzed by HPLC-HRMS.The conditions employed here were the same conditions described by Bertanha et al. in 2020.The HRMS data were obtained using the following conditions: capillary voltage = 3.5 kV, dry temperature = 220°C, nebulizer gas at 60 psi, dry gas at a flow rate of 10 L/min, mass range = 50-1,300 Da, and drying, nebulizing, and collision gas = nitrogen.The HPLC method consisted of a linear gradient of 0.1% acetic acid in water (solvent A) and methanol (solvent B), starting from 95% solvent A and 5% solvent B and going to 100% solvent B over 35 min, followed by 10 min with 100% solvent B; the flow rate was 1.0 mL/min.

Validation
The chromatographic conditions were established through trial and error.Various gradient elution conditions were tested primarily by altering the solvent composition and using CE-2.The optimal conditions were determined using an ODS Luna Phenomenex column (5 mm, 250 × 4.60 mm).Separation was achieved through gradient elution with solvent A (2% aqueous acetic acid) and solvent B (acetonitrile).The following linear gradient was applied: 2%-25% B over 40 min, followed by 25%-100% B in 5 min and 100% B for 5 min.Sample volumes of 20 µL were injected at a flow rate of 1.0 mL/min; the oven temperature was 40°C.The column was equilibrated at 2% B for 15 min.The UV-Vis spectra were recorded from 200 to 600 nm, and the chromatograms were monitored at l = 330 nm.
The method was validated by following the guidelines of the Brazilian Health Surveillance Agency (ANVISA) and involved assessing several parameters, including selectivity, linearity, limit of detection (LOD), limit of quantification (LOQ), accuracy, and precision (ANVISA, 2017).Selectivity was confirmed using a diode array detector, which allowed the UV-Vis spectra to be obtained and peak purity to be assessed during analyses of the crude ethanol S. perforatum leaf extracts.This also helped to verify the identity of the standard by comparing its retention time with the verbascoside standard retention time.To establish linearity, 20 µL of standard solutions of 1 at concentrations ranging from 0.8 to 60 mg/mL and containing 10 µg/mL IS was injected into the chromatographic system.A similar procedure was followed for IS.Linear regression analysis was applied to the obtained peak areas, and the analyses were performed in triplicate.The linear regression equation of the calibration curve was used to determine LOD and LOQ.Accuracy and precision were assessed using IS.CE-2 (1.5 mg in 1 mL of methanol) was spiked with IS at 50, 10, or 3 mg/mL, which was followed by sonication.After 10 min, the solution was centrifuged in an Eppendorf centrifuge 5804 at 5,000 rpm for 5 min and transferred to a vial.Accuracy was evaluated by back-calculation and expressed as the percent deviation between the experimental amounts of IS and the amounts added at the three examined concentrations.Precision was estimated using relative standard deviation (RSD).Repeatability and intermediate precision tests were conducted on a single day and three non-consecutive days using the samples from the recovery experiment.

Acute toxicity test
Female Wistar Hannover rats (Rattus norvegicus) aged between 8 and 12 weeks were obtained from the Animal Facility at the University of São Paulo in Ribeirão Preto.The animals were kept in individual isolators under controlled temperature (23°C ± 2°C) and humidity (50% ± 10%), under a 12-h light-dark cycle, with ad libitum access to water and food.The CE-2 acute toxicity and LD 50 were obtained by following OECD 425 (2022), which classifies compounds in a Global Harmonized System (GHS).CE-2 was diluted in dimethyl sulfoxide and orally administered via gavage; each animal received a single dose of 2,000 mg of CE-2/kg of body weight.Then, the animals were observed for 14 days to monitor possible deaths and behavioral changes.Each animal had its body mass measured each week starting from the oral administration.At the end of the experimental period, the animals were euthanized through intraperitoneal injection of sodium thiopental (Thiopentax 1.0 g; 840 mg/kg of body weight), and a necropsy was conducted to search for macroscopically detectable pathologies.Following macroscopic pathological necropsy, each animal had the liver and kidneys removed for histopathological analysis to evaluate ulceration, necrosis, angiogenesis, and inflammatory infiltrate in the connective tissue, including polymorphonuclear cells like neutrophils, eosinophils, and basophils, as well as mononuclear cells such as monocytes and lymphocytes.The intensity scores 0 (absent), 1 (mild), 2 (moderate), and 3 (intense) were used.Connective tissue fibroplasia was also assessed.
The antileishmanial activity was also assessed against L. amazonensis amastigotes; the method described by Oliveira et al. ( 2022) was followed.Briefly, a suspension of 2 × 10 5 J774.1 macrophages in supplemented RPMI 1640 medium (Gibco ® ) was seeded on glass coverslips in a 24-well microplate and incubated at 37°C under 5% CO 2 for 2 h.After incubation, the adherent macrophages were infected with L. amazonensis promastigotes at a concentration of 1 × 10 6 cells/well (10:1 ratio) at 37°C for 4 h.Non-internalized L. amazonensis promastigotes were washed, and the infected cultures were incubated with one of the tested samples at concentrations ranging from 3.12 to 50 µg/mL or 3.12 to 50 µM for 48 h.Amphotericin B (0.095 to 1.56 mM) and RPMI 1640 medium containing 0.1% dimethyl sulfoxide served as the positive and negative controls, respectively.Next, coverslips were prepared and examined under an optical microscope (Nikon) by counting 200 macrophages to determine the number of L. amazonensis amastigotes within each infected cell.In both experiments, the 50% effective concentration (EC 50 ) values were determined by nonlinear regression analysis.The experiments were performed in triplicate and repeated twice.
Transmission electron microscopy (TEM) and scanning electron microscopy (SEM) analyses were performed by following the protocols of Alvarenga et al. (2016) and Pereira et al. (2015) with some modifications.L. amazonensis promastigotes (1 × 10 6 /mL) were incubated with 1 at 8.92 or 89.7 mM (EC 50 and EC 90 ), RPMI 1640 containing 0.1% dimethyl sulfoxide (negative control), or amphotericin B (positive control) at 24°C for 24 h.After incubation, L. amazonensis promastigotes were centrifuged and fixed with 3% glutaraldehyde in phosphate buffer (0.1 M, pH 7.2) at 37°C for 1 h and at room temperature for 1 h.The solution was removed, and L. amazonensis promastigotes were washed twice with 0.1 M phosphate buffer and processed for TEM or SEM analyses.The samples were analyzed using a JEOL Model JEM-100CXII transmission electron microscope (Tokyo, Japan) or a Joel JSM-5200 scanning electron microscope (Tokyo, Japan) at the Multi-User Microscopy Laboratory of the Ribeirão Preto Medical School.

Cytotoxic activity
The cytotoxic activity of the samples was assessed using the murine macrophage cell line (J774A.1)cultivated in RPMI 1640 medium.The cytotoxicity assay used 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium bromide (MTT; Sigma-Aldrich); the protocol described by Valerino-Dıáz et al. ( 2020) was followed.The macrophages were seeded at 2 × 10 5 cells/well in 96-well culture plates and incubated at 37°C under 5% CO 2 for 24 h.Following incubation, the tested sample was added to the plate at final concentrations ranging from 100 to 3.12 µg/mL or µM.The positive control contained 25% dimethyl sulfoxide, while the negative control contained 0.1% dimethyl sulfoxide.The plates were further incubated for 48 h.The concentration that reduced cell viability by 50% (CC 50 ) was determined through non-linear regression curve analysis.The experiments were performed in triplicate and repeated twice.

Isolation and identification of compounds
CE-1 exhibited antileishmanial activity and was initially processed by liquid-liquid partition.At 50 µg/mL, the three resulting fractions showed activity against L. amazonensis.However, HF and EAF had stronger activity than HMF, so they were used in further purification steps.EAF was submitted to sequential re-fractionation on a Sephadex LH-20 column to isolate verbascoside (1).Subfractions 63 and 64 were then purified by preparative TLC to isolate penduletin (2) and casticin (3).Another EAF aliquot was processed by SPE to obtain four SPE subfractions (EAF-1, EAF-2, EAF-3, and EAF-4).Bioassays against L. amazonensis revealed strong activity for EAF-3, so it was purified by silica gel column chromatography and preparative HPLC to isolate casticin (3), penduletin 4′-methyl ether (4), ursolic acid (5), and oleanolic acid (6).HF was purified on a silica gel column to give mixtures of stigmasterol (7) and b-sitosterol (8), as well as 5 and 6.HMF also exhibited antileishmanial activity, albeit lower.Nevertheless, it was purified by SPE to isolate 1 in fraction 2 after a preparative HPLC step.

Phytochemical analyses by HPLC-HRMS-DAD
We conducted phytochemical fingerprinting of CE-1 and CE-2 by HPLC-HRMS-DAD; we employed electrospray ionization in the positive and negative modes, and we set UV-Vis at 200-800 nm.We obtained the chromatograms using a linear gradient of methanol/water (+0.1% acetic acid) from 5% to 100% methanol for 35 min and 100% methanol for 10 min; the flow rate was 1.0 mL/ min.The data provided information about the similar chemical profile of S. perforatum CE-1 and CE-2 and helped us to select the chemical marker.The DAD detector provided chromatograms (Figures 2A, B) for CE-1 and CE-2; Figures 2C-F present the total ion chromatogram (TIC) of CE-1 and CE-2 in the negative and positive modes.Combined information from the UV-Vis and mass spectra revealed nine main peaks: at R t of 10.5, 10.9, 11.7, 12.0, 12.3, 22.5, 22.9, 23.4, and 27.9 min for CE-1 and at R t of 9.2, 10.5, 10.8, 11.7, 14.2, 14.9, 16.6, 17.2, and 23.7 for CE-2.Table 1 describes the possible molecular formulas of the most intense ions with a mass error lower than 5 ppm.Both CE-1 and CE-2 presented peaks at 10.5 and 11.7 min, respectively, which were assigned to C 21 H 20 O 11 and C 29 H 36 O 15 , respectively.The peak at 11.7 min confirmed that 1 was present in the TIC of CE-1 and CE-2 in the negative and positive modes.Mass spectrometry (Supplementary Material) helped to identify the structures of 2 and 3 in CE-1 and CE-2 and flavonoid 4 in CE-1 only.

Validation and quantitative analyses
The chromatographic condition (Figures 3A-E; linear gradient of solvent A (2% aqueous acetic acid) and solvent B (acetonitrile)) was as follows: 2%-25% B over 40 min, followed by 25%-100% B over 5 min and 100% B for 5 min) allowed 1 (R t of 38.68 min) to be separated from the other compounds in CE-1 and CE-2; mangiferin (R t of 28.89 min) was the IS.We also obtained the UV-Vis spectra (Figures 3F-I) in this condition.Peak purity, obtained with the aid of the DAD detector, was 0.999 for both 1 and IS, confirming that no coelution occurred.We investigated linearity using external and internal standardization methods.At 330 nm, the method was linear for 1 at concentrations between 0.8 and 60.0 µg/mL.The regression equation was y = 34,237x − 23,016, the correlation coefficient (r) was 0.9920, and the RSD was less than 5% for triplicate analyses.At 330 nm, the IS analytical curve was also linear in the same concentration range.The regression equation was y = 24,680x − 31,022, r was 0.9948, and RSD was less than 5% for triplicate analyses.As for the internal standardization calibration curve, the equation was y = 0.1685x − 0.1157, with r = 0.9958 and RSD < 5%.The LOD and LOQ were 0.30 and 1.01 µg/mL, 0.05 and 0.17 µg/mL, and 0.36 and 1.19 µg/mL, respectively.IS recovery from the CE-2 solutions spiked with IS at 3.0, 10.0, and 50.0 µg/mL was 102.82%, 108.51%, and 111.04%, respectively, with RSD < 5%.Regarding repeatability, RSD was 0.64%, 1.48%, and 1.76%, respectively.With respect to intermediate precision, RSD was 3.15%, 0.60%, and 1.50%, respectively.We determined that the concentration of 1 in CE-1 and CE-2 was 3.51 ± 0.20 and 3.78 ± 0.14 µg/mL, or 0.59% and 0.63% relative to the dry crude ethanol S. perforatum leaf extract, respectively.

Acute toxicity
In this study, animals treated with 2,000 mg of CE-2/kg of body weight for 14 days did not die or have significantly altered behavior.All the animals presented an average weight gain of 31.00 ± 2.35 g in this period.The macroscopic pathological necropsy did not show alterations that could indicate pathologies in organs and tissues.The histopathological analysis did not indicate alterations typical of pathologies in the kidney and liver (Table 2) .On the basis of the GHS classification for acute toxicity of chemicals, we classified CE-2 in category 5 (LD 50 > 2,000-5,000 mg/kg of body weight).
To assess whether 1 promoted morphological and structural changes in L. amazonensis promastigotes, we performed SEM and TEM analyses of L. amazonensis promastigotes incubated with 1 at 8.92 µM (EC 50 ) or 89.7 µM (EC 90 ) for 24 h.SEM analyses (Figure 4) showed morphological changes in L. amazonensis promastigotes incubated with amphotericin B (positive control, Figure 4D) or 1 (Figures 4F, H) when compared to the negative control (0.1% dimethyl sulfoxide, Figure 4B).In the negative control group, L. amazonensis promastigotes presented an elongated shape (Figure 4B).Compared to the negative control, L. amazonensis promastigotes incubated with 1 (Figures 4F, H) underwent changes such as a reduced number of parasites, acquisition of a rounded shape, and altered plasma membrane.TEM revealed an elongated body and intact organelles (nucleus, plasma membrane, and kinetoplast) in the negative control (Figure 4A).L. amazonensis promastigotes incubated with amphotericin B and positive control (Figure 4C) presented vacuolization of the cytoplasm, swollen kinetoplasts, lipid bodies, and altered nuclear membrane.As for L. amazonensis promastigotes incubated with 1 at 8.92 µM (Figure 4E) or 89.7 µM (Figure 4G), they displayed vacuolization of the cytoplasm, lipid bodies, altered nuclear membrane, interstitial extravasation, and swollen kinetoplasts.

Discussion
The phytochemical study of the crude ethanol leaf extract obtained from S. perforatum indicated that this species biosynthesizes triterpenes, phenylethanoid glycosides, and flavonoids.Among these compounds, only 5 had previously been isolated from S. riparium, a morphologically similar species (Duh et al., 1987;Beyer et al., 2017).To the best of our knowledge, this is the first report of 1-4 and 6-8 in Stizophyllum.
Chromatographic fingerprinting of CE-1 and CE-2 showed some differences in their chemical composition, well established and influenced by factors such as plant source, growth, harvest, season, and water availability (Chaves et al., 1997;Copaja et al., 2003;Soni et al., 2015;Gobbo-Neto et al., 2017;Prinsloo and Nogemane, 2018;Sampaio and Da Costa, 2018;Gimenez et al., 2020).However, we also observed some similarities, including the peak at R t of 11.7 min, which revealed that 1 is present in both CE-1 and CE-2.Thus, we selected 1 as the chemical marker for S. perforatum.Interestingly, 1 has also been described in other Bignoniaceae species, such as Cuspidaria pulchra and Xylophragma harleyi (Lima et al., 2003;Alvarenga et al., 2015).The phytochemical study of S. riparium resulted in the isolation of triterpene and pregnane    derivatives (Duh et al., 1987).Additionally, we compared the formulas obtained from the HPLC-HRMS characterization with the formulas of previously described compounds, but we were unable to identify any of them.Nevertheless, we were able to identify 5 by NMR data after it was isolated.
The developed HPLC method is suitable for controlling the quality of herbal materials from S. perforatum, and according to Indrayanto (2022), method validation in plant science is crucial for developing modern drugs from herbal medicines.
Our results concerning the antipromastigote and antiamastigote activities of CE-1 and CE-2 are more promising than the results reported by Maquiaveli et al. (2016b), who isolated 1 and isoverbascoside from the n-butanol fraction of Stachytarpheta cayennensis, tested them against L. amazonensis promastigotes and  amastigotes, and obtained EC 50 of 51 and 32 µg/mL, respectively.In addition to that, a previous study reported that the crude leaf extract obtained from S. perforatum exhibits trypanocidal activity with an EC 50 of 20.2 µg/mL (Silva et al., 2014).Another study, which also isolated 5 from Ajuga laxmannii, showed that the crude methanol extract of this plant has EC 50 of 30.1 µg/mL against Leishmania donovani axenic amastigotes (Atay et al., 2016).Here, we determined that the concentration of 1 in CE-1 and CE-2 is 3.51 ± 0.20 and 3.78 ± 0.14 µg/mL, or 0.59% and 0.63% relative to the dry crude ethanol leaf extract, respectively.However, the concentrations of 1 in the plant extracts mentioned in the literature were not established.Verbascoside (1) has various biological activities, including antioxidant, anti-inflammatory, antineoplastic, wound-healing, neuroprotective, and antileishmanial activity properties (Alipieva et al., 2014;Atay et al., 2016;Maquiaveli et al., 2016a;Maquiaveli et al., 2017). 1 has an EC 50 of 19 mM against L. amazonensis promastigotes and is a competitive arginase inhibitor (Ki = 0.7 mM) (Maquiaveli et al., 2016a).Additionally, 1 inhibits L. amazonensis amastigotes (EC 50 = 32 mM) by selectively inhibiting arginase, which reduces the protective oxidative mechanism of the parasite and impairs trypanothione synthesis.Arginase belongs to the polyamine biosynthesis pathway, which is important for parasite infectivity (Maquiaveli et al., 2017).Our results are more promising-1 has EC 50 of 6.23 and 3.71 mM for L. amazonensis promastigotes and amastigotes, respectively.The antileishmanial activity of 1 has also been established against L. donovani axenic amastigotes (8.7 µg/ mL) (Kirmizibekmez et al., 2004;Atay et al., 2016).
The presence of these compounds in CE-1 and CE-2 can contribute to antileishmanial activity.These findings are often considered to result from a synergistic or additive effect of the constituents (Caesar and Cech, 2019).Our results provide evidence of the antileishmanial activity induced by S. perforatum and additional evidence of the antileishmanial activity of 1 and 3. Further in vivo studies should be conducted to assess the antileishmanial activity of the S. perforatum crude extract against Leishmania.
HPLC-HRMS data showed that 1 is present in both CE-1 and CE-2, confirming that the extracts exhibit similar but not identical chemical composition.
The evaluated samples, CE-1, CE-2, EAF, EAF-3, 1, and 3, display antileishmanial activity against L. amazonensis promastigotes and amastigotes.The antiparasitic effect of 1 was confirmed by ultrastructural changes in the parasite.Nevertheless, the samples showed low cytotoxicity in acute oral tests.Thus, S. perforatum is a suitable candidate for further in vivo investigations against L. amazonensis.

FIGURE 1
FIGURE 1 FIGURE 2 (A, B) Chromatograms recorded with UV-Vis detection from 200 to 800 nm.(C, D) Base peak chromatogram in the negative mode.(E, F) Base peak chromatogram in the positive mode of crude ethanol Stizophyllum perforatum leaf extracts CE-1 and CE-2.Analytical conditions: methanol/water (+0.1% acetic acid) linear gradient from 5 to 100% methanol for 35 min and 100% methanol for 10 min.The flow rate was 1.0 mL/min, and the ODS column was Phenomenex Luna.

TABLE 1
Major compounds identified by high-performance liquid chromatography-high-resolution mass spectrometry (HPLC-HRMS) data of crude ethanol Stizophyllum perforatum leaf extracts CE-1 and CE-2.

TABLE 2
Histopathological analysis data of liver and kidney treated with crude ethanol Stizophyllum perforatum leaf extract.

TABLE 3
Antileishmanial and cytotoxic activities of crude ethanol Stizophyllum perforatum leaf extracts and isolated compounds.