https://orcid.org/0000-0003-2862-3771
Abstract
The objective of this work was to evaluate the anti-inflammatory activity of the aqueous extract, fractions and major compounds, which are isolated and identified from Passiflora edulis f. edulis (purple passion fruit) leaves extract.
For the isolation of the major compounds, reversed-phase chromatography and normal phase countercurrent chromatography were used. The separation was followed by thin layer chromatography and HPLC-DAD-ELSD. One-dimensional and two-dimensional NMR and ESI-TOF-MS/MS were used for structural elucidation. The anti-inflammatory activity was evaluated on a TPA multiple dose model of skin chronic inflammation in mice. Additionally, myeloperoxidase (MPO) and nitric oxide synthase (NOS) activity assays were performed as possible mechanisms of action studies.
The study of the butanolic fraction mainly showed the presence of saponins and flavonoids. Three minor flavonoids were detected; and three known saponins, cyclopassiflosides IX, XI and III were isolated and identified. This is the first unequivocal report of the presence of these compounds in P. edulis f. edulis leaves. The most favourable results of anti-inflammatory activity were obtained for the flavonoid-rich fraction. All the fractions and isolated compounds evaluated, presented high percentages of inhibition of nitric oxide synthase activity.
Introduction
The genus Passiflora represents more than 530 species distributed mainly in tropical regions and some reports in India, China, Australia and the Pacific islands. More than 170 species have been described for northern Andes.[1, 2] Chemical studies of plants of this genus allowed the isolation of alkaloids, phenols, cyanogenic compounds, saponins and glycosyl flavonoids.[3, 4]
Two forms of Passiflora edulis Sims are mainly cultivated in Colombia, due to the commercialization of its fruits: P. edulis f. flavicarpa O. Deg. (yellow passion fruit, maracuja) and P. edulis f. edulis (purple passion fruit, gulupa).[5] The latter is a semi-perennial climbing liana with alternative, three-lobed leaves of green colour. The flower is generally solitary, semi-erect, pentameric, and hermaphroditic, with a pleasant aroma, a length of 4.5–6.5 cm and a width of 4–5 cm. The fruit is a spherical or ovoid berry, with a hard, smooth pericarp and a spongy, white mesocarp. It turns from pale green to dark purple at maturity. The seeds are contained in a gelatinous yellow pulp that exhibits intense aroma and sweet-acid taste.[6, 7]
The concentrated juice manufacturing is the most impacting activity involving this species, and wide-scale processing inevitably generates big amounts of agro-industrial residues. Therefore, bioprospecting studies to transform these by-products, such as leaves, into useful high value-added products are required,[1] especially, given the numerous therapeutic applications studied for different parts of the plant. Ethnopharmacological uses of plants of this species around the world include digestive stimulants, treatment of dysentery and hypertension, menopausal symptoms, tonic, anti-diarrhoeal, anthelmintic, diuretic and sedative.[4] Pharmacological activities, such as sedative, anxiolytic, antitumour, antidepressant, antidiabetic, hypolipidemic, hepatoprotective and lung-protective, anti-hypertensive, antimicrobial, analgesic and anti-inflammatory, have been studied.[3, 8]
Within the composition of the fruits, seeds and leaves of plants of Passiflora edulis, the presence of pectin, fibre, polysaccharides, lipids, proteins, amino acids, alkaloids, sulforaphanes, carotenoids, flavonoids and triterpenoids has been described. The last two are considered the main chemical constituents[8] and are metabolites widely related to antibacterial, antioxidant, antihaemolytic and anticancer properties.[9, 10] Additionally, for presenting anti-inflammatory effects,[11–14] the case of interest in this study.
However, the intraspecific taxonomy of P. edulis Sims is contradictory, and studies on its botanical forms are very limited.[8] The chemical composition and pharmacological effects are not clearly differentiated between the two main forms, since the studies found in the literature of P. edulis Sims lack precision about the specific form that is being studied.[15–17] In this context, this research aimed to identify and clarify the saponin and flavonoid content of P. edulisf. edulis leaves and to study its anti-inflammatory properties as a possible application of such a by-product of the crops.
Materials and Methods
Materials
Ethyl acetate, n-butanol, ethanol and acetonitrile were purchased from Merck. Water was purified using a Milli-Q system (Millipore, Bedford, MA, USA). Deuterated pyridine pyr-d5 99.8% (compounds 1 and 2) and methanol-d4 99.8% (compound 3) were purchased from Merck. Dexamethasone standard was purchased in a GENFAR brand injectable dosage form. The inflammation inducer 12-O-tetradecanoylphorbol-13-acetate (TPA) was purchased from Sigma-Aldrich in reagent grade. Hydrogen peroxide and tetramethylbenzidine (TMB) were acquired from Sigma-Aldrich.
Plant material, extraction and fractionation
The leaves of Passiflora edulis f. edulis (gulupa) were collected from a crop located in Nemocón (55 km from Bogotá), Colombia. The plant was identified by taxonomist Gustavo Morales on the basis of its morphological characteristics, and a voucher specimen was deposited at Herbario Nacional Colombiano, Universidad Nacional de Colombia (COL530661). The leaves were shade-dried at room temperature and grounded to a fine powder using a blade mill.
Fractionation of aqueous extract (AE)
The powdered vegetal material (100 g) was extracted by infusion with boiled distilled water (90°C, 1 : 10 w/v) for 10 min to simulate the extraction method used in popular medicine.[18] The extract was filtered and freeze-dried, yielding 25.0 g of AE. Then, as reported in previous phytochemical studies,[18, 19] a saponin and flavonoid-rich fraction was obtained by dissolving 20 g of AE in 50 ml of water to be partitioned with n-butanol (20 ml, 3 times); the organic solvent was removed under reduced pressure and then freeze-dried yielding 6 g of butanolic fraction (BF).
A portion of the BF (100 mg) was separated on a Sephadex LH-20 (27–163 µm, Pharmacia Biotech) column (60 × 15 mm), using methanol as eluent, collecting 3 ml each time, the first fractions yielded a fraction rich in saponins, while the last fraction was rich in flavonoids. All these fractions were analysed by UHPLC-DAD (ultra-high-performance liquid chromatography-diode array detection) and HPLC-DAD-MS (high-performance liquid chromatography coupled with diode-array detection and electrospray ionization tandem mass spectrometry).
Isolation and characterization of main compounds
The BF was subjected to high-speed countercurrent chromatography (HSCCC) using a P.C. Inc. apparatus. Five proportions of the solvent system ethyl acetate : n-butanol : water (1 : X : 1) were evaluated according to the partition coefficient calculated by HPLC-DAD-ELSD (high-performance liquid chromatography-diode array detector-evaporative light scattering detector, system A, X = 0.1; B, X = 0.3; C, X = 0.5; D, X = 0.7 and E, X = 1.2). Initially, the lower aqueous phase of system A was selected as the stationary phase, and the upper organic phases of systems A–C were used as the gradient mobile phase in tail-to-head mode at 1.0 ml/min and 600 rpm. After reaching hydrodynamic equilibrium, 1 g of the BF dissolved in 7 ml of a mixture of both phases (1 : 1 ratio) was injected. A total of 121 fractions of 3 ml each were collected and pooled according to their TLC profiles in fractions IA to VIIA. Fraction VIIA (546.5 mg) was further chromatographed with the same previous conditions but in isocratic mode of solvent system C to isolate 143.9 mg of compound 1 and 50.2 mg of compound 2. A second chromatographic separation, employing an isocratic mode of solvent system B for the separation of 1 g of BF dissolved in 5 ml of the lower aqueous phase, yielded 10 grouped fractions: IB-XB. The third fraction, IIIB (106.5 mg), was further purified by chromatographic RP-18 column eluted with acetonitrile/water (30–70% ACN) to yield 60.8 mg of compound 3.
UHPLC-DAD-ELSD analysis
An Ultimate 3000 Thermo Dionex UHPLC apparatus was used, equipped with diode array detection (DAD) and coupled to a Sedex 85 Evaporative Light-Scattering Detector (ELSD) operating with nitrogen gas at 60°C and 40 psi pressure. The DAD spectra were acquired between 190 and 450 nm. Kinetex C8 (100 mm × 4.6 mm, 2.6 µm) was used as stationary phase at 21°C. The injection volume was 10 µl. A gradient system of water (solvent A) and acetonitrile (solvent B) was used as mobile phase, 15–48% B (0–30 min) and 48–15% B (30–32 min). The flow rate was kept constant at 0.5 ml/min. Data were processed using the software Chromeleon Client version 6.80.
UHPLC-MS analysis
LC-MS analyses were carried out using an Ultimate 3000 UHPLC-TOF-MS coupled to a Bruker Daltonics micro-TOF-QII. Electrospray ion source (ESI)-MS spectra were acquired in positive and negative ion modes, and the interface and MS detector parameters were as follows: detector voltage, 3.8 kV; CDL voltage, 500 V; CDL temperature, 200°C; heat block temperature, 200°C and nitrogen as nebulizer gas at a flow rate of 11 L/min. Kinetex C18 column was used (150 × 2.0 mm, 1.7 µm). The gradient system consisted of acetonitrile (A) and 0.5% formic acid (B); 17–35% A in 20 minutes at 0.5 ml/min. The injection volume was 3 µl.
NMR experiments
The NMR spectra were recorded on a Bruker Avance II 400 spectrometer, using pyridine-d5 and methanol-d4 as solvents. The residual solvent signals were used as the internal standard.
Animals
Female ICR mice (8 weeks old, 25–30 g) were supplied by the Pharmacy Department of Universidad Nacional de Colombia. The animals were kept under constant temperature conditions (22°C ± 1), 12-h light/dark cycles, with food and water ad libitum. The assays were carried out in accordance with the international and local ethical guidelines on the use and care of laboratory animals, with the approval of the local Research Ethics Committee (Act 03/2014 Faculty of Sciences, Universidad Nacional de Colombia). The following eight treatments were tested (n = 6): control (ethanol 70%), AE, BF, saponin fraction (SF), flavonoid fraction (FF), compound 1, compound 2 and dexamethasone (reference drug).
Anti-inflammatory assay
The chronic inflammation model mouse ear oedema induced by multiple topical application of TPA was used, following Navarro et al. and Stanley et al.’s procedure.[20, 21] Oedema was induced after the application of 10 µl of TPA in both ears (2 µg/ear) on alternative days (0, 2, 4, 7 and 9). The AE, BF, SF, FF and compounds 1 and 2 (isolated in sufficient amount), dissolved in 70% aqueous ethanol, and 10 µl were topically applied twice daily guaranteeing a dose of 0.5 mg/ear on days 7, 8, 9 and only in the morning at day 10. Dexamethasone dissolved in acetone, 10 µl, was applied guaranteeing a dose of 0.05 mg/ear under the same application schedule of the treatments. The control group was treated with 10 µl of 70% EtOH. The mice were sacrificed 6 h after the last product application and two ear punches (6 mm diameter) were taken from each animal. Each punch was weighted and froze for further enzymatic analysis. Swelling inhibition was expressed as weight reduction referred to the control group. The percentage inhibition of inflammation was calculated as: [(weight control–weight sample)/weight control] × 100.
Myeloperoxidase (MPO) assay
The enzymatic analysis was based on De Young et al.’s conditions with slight modifications.[22] Ears from each treatment were homogenized using a Kinematica Polytron Aggregate rotor in 750 µl 0.5% hexadecyltrimethylammonium and centrifuged at 12 000× g at 4°C for 10 min. The supernatant (50 µl × triplicate) was added to a mixture containing 150 µl of phosphate-buffered saline, 20 µl of 80 mM phosphate buffer pH 5.4 and 20 µl of H2O2 0.012% in a 96-well microtiter plate. The reaction was started after 5 minutes of pre-incubation at 37°C by adding 20 µl of 18 mM TMB in 8% aqueous dimethylformamide, then, it was incubated again for 3 min at 37°C and stopped with 30 µl of acetic acid (2 M, pH = 3.0). Enzyme activity was determined colourimetrically in Bio-Rad plate reader set to measure absorbance at 620 nm. It was expressed as the inhibition percentage of MPO levels, determined by the absorbance difference between the control group and the treated group compared with the absorbance observed in the control group.
Nitric oxide (NO) assay
NO release was indirectly measured by the quantification of nitrite (NO oxidation product) in Griess assay as described by Saleh et al.[23] The Griess reagent was composed of equal volumes of 1% sulfanilamide and 0.1% naphthyl ethylenediamine in 5% HCl. 80 µl of homogenized ear extract was transferred to a fresh 96-well plate and mixed with 80 µl of Griess reagent. After 10 min at room temperature, absorbance was measured at 540 nm using a Bio-Rad plate reader set and NO concentration was determined using a calibration curve.
Statistics
IBM SPSS Statistics software version 23.0.0.0 was used. Statistical significance was determined by one-way analysis of variance (ANOVA) followed by Dunnett’s t-test for multiple comparisons. Values were considered statistically significant in comparison with the control group at P ≤ 0.05.
Results
UHPLC analysis of the butanolic fraction
To characterize the main compounds of BF, it was analysed by UHPLC-DAD-ELSD. The flavonoid presence was initially identified due to their characteristic absorption at 280 and 340 nm, corresponding to the peaks in retention times 11–15 min of chromatograms (Figure 1). On the contrary, the absorption absence of some compounds at wavelengths >205 nm and their detection with ELSD indicated the presence of terpene compounds without chromophore groups, possibly saponins. Additionally, since it is a universal detector able to discriminate analytes according to their concentration in the eluted sample,[24] flavonoids were identified as minor metabolites and peaks 1–3 as major compounds of P. edulis f. edulis leaves extract.
Chromatogram of P. edulis f. edulis leaves butanolic fraction. Graph obtained by (A) DAD at 340 ηm, (B) DAD at 205 ηm, (C) evaporative light scattering detector. The compounds 1–3 correspond to the isolated compounds.
Flavonoid characterization
The low concentration of flavonoids did not allow to isolate them, and it was necessary to conduct a UHPLC-TOF-MS analysis. The study of UV profile and MS/MS fragmentation allowed to identify three flavonoids: vicenin-2 (retention time 5.35 min and [M − H]−m/z 593), 6,8-di-C-glucosylchrysin (retention time 5.81 and [M − H]−m/z 577) and spinosin (retention time 8.67 and [M − H]−m/z 607). In addition, an isomer of 6,8-di-C-glucosylchrysin was also detected (retention time 8.41 and [M − H]−m/z 577). The retention times correspond to those of Figure 2 and the [M − H]− pseudomolecular ions were compared with the data reported by Zucolotto et al.[16]
Chromatogram of P. edulis f. edulis leaves butanolic fraction. Graph obtained by: (A) UV at 256 ηm, (B) total ion chromatogram. 1: vicenin-2, 2: 6,8-di-C-glucosylchrysin, 3: unidentified isomer, 4: spinosin.
Characterization of the isolated compounds
As was suggested before, the main compounds 1–3 correspond to saponins (Figure 1). Each compound was purified by chromatographic procedures. Compound 1 was isolated as a colourless solid ([α]25D + 21.2, c 2.0, MeOH) and has a molecular formula C43H72O17, assigned on the basis on its HRESIMS pseudomolecular ion at m/z 861.4843 [M − H]+ (error 0.6 ppm), and congruent with seven degrees of unsaturation. The 1H NMR (400 MHz, pyridine-d5) and 13C (100 MHz, pyridine-d5) NMR were indicative of a saponin-type structure. The APT spectra (100 MHz, pyridine-d5) allowed to identify 43 carbons, which were assigned as: 6 methyls, 12 methylenes, 18 methines (1 sp2) and 7 quaternary carbons, including an ester carbon at δ C 177.0 (s, C-28). The identification of two anomeric methine carbons at δ C 96.6 (C-1′/δ H 6.51, 1H, d, J = 8.0 Hz, H-1′) and δ C 105.8 (C-1″/δ H 4.95, 1H, d, J = 7.7 Hz, H-1″) confirms the presence of two β-glucopyranoside residues. Considering the number of oxygenated methines[13] and oxygenated methylenes,[3] it was suggested the presence of 5 hydroxylated carbons on the triterpene nucleus, corresponding to one methylene, three methines and one quaternary carbon. Finally, the presence of 5 signals for methyl carbons at δ C 9.9 (q, C-29), δ C 17.4 (q, C-26), δ C 17.6 (q, C-27), δ C 18.7 (q, C-18), δ C 19.7 (q, C-21) and δ C 20.5 (q, C-30) suggests the presence of cycloartenol-type steroid. The cyclopropane feature was evidenced by the signals at 1H NMR spectrum at δ H 0.54 (1H, d, J = 3.5 Hz, H-19a/δ C 30.4) and δ H 0.76 (1H, d, J = 3.5 Hz, H-19b/δ C 30.4).[25] The protons for carbinol residues at δ H 5.59 (1H, dd, J = 11.9 y 3.9 Hz, H-3/δ C 70.9), 4.68 (1H, dd, J = 11.5 y 7.8 Hz, H-16/δ C 72.5), 3.99 (2H, d, J = 9.9 Hz, H-31/δ C 74.8) and 3.90 (1H, br s, H-1/δ C 71.9) confirmed the presence of three secondary alcohols and one primary alcohol. The previous analysis and the comparison of these data with those previously published allowed identifying compound 1 as cyclopassifloside IX (Figure 3).[26]
Structure of main compounds isolated from P. edulis f. edulis leaves (purple passion fruit, gulupa).
Compound 2 (50.2 mg) was obtained as a white solid ([α]25D + 15.1, c 2.0, MeOH). The (+)-ESI-TOF spectrum showed a [M + H]+ pseudomolecular ion at m/z 861.4850 (error 0.8 ppm), corresponding to a molecular formula of C43H72O17. These data indicate that compound 2 is an isomer of 1. In general, NMR signals for compound 2 were similar to those obtained for compound 1. The main differences were observed in the 13C NMR (100 MHz, pyridine-d5) signals between 72 and 77 ppm, which correspond to carbons assigned to D ring. Additionally, in the 1H NMR spectrum (400 MHz, pyridine-d5), for the carbinolic proton H-16, which for the case of 2 appears at δ H 4.30 (1H, br m, H-16/δ C 77.3), whereas for compound 1, it appears at δ H 4.68 (1H, dd, J = 11.5 and 7.8 Hz, H-16/δ C 72.1), indicating an opposite configuration between them. The analysis of proton coupling constants in compound 2 shows that proton H-16 is pseudo-equatorial, with a small J value. The comparison of the NMR data here obtained with those published in the literature allowed, concluding that 2 corresponds to cyclopassifloside XI (Figure 3), previously reported by Yoshikawa et al.[26]
Compound 3 was isolated as a colourless solid, [α]25D + 24.8 (c 2.0, MeOH), and showed at (+)-ESI-TOF spectrum a [M + Na]+ pseudomolecular ion at m/z 867.4720 (error 11.7 ppm) suggesting a molecular formula C43H72O16 (7 degrees of unsaturation). Thus, it was established that compound 3 has one oxygen atom less than compounds 1 and 2. The APT spectrum (100 MHz, methanol-d4) of compound 3 allowed to identify 43 carbons, which were assigned to 6 methyl, 13 methylenes, 17 methines and 7 quaternary carbons. 13C NMR chemical shifts reveal the presence of 16 oxygenated sp3 carbons. All NMR shifts (1H and 13C) were similar to those shown for the two compounds previously described. The main differences are seen in the signals attributed to D-ring members, suggesting the absence of hydroxyl at the C-16 carbon for compounds 3. The comparison of these data with those published in the literature allowed, establishing that 3 corresponds to cyclopassifloside III (Figure 3) and was reported by Yoshikawa et al.[27]
Anti-inflammatory activity
As can be seen in Figure 4, all fractions and compounds of P. edulis f. edulis leaves, but not the AE, exhibited a statistically significant reduction of the mouse ear oedema compared with the control group. The inhibition of the inflammation by the AE, BF, SF, FF, compounds 1 and 2 was: 9.5, 15.1, 10.7, 18.3, 16.3 and 10.7%, respectively. Dexamethasone used as the positive control showed 38.1% inhibition.
Anti-inflammatory effect of the aqueous extract, fractions and compounds of P. edulis f. edulis. Values are expressed as mean ± SEM (n = 6). One-way ANOVA, Dunnett’s post-hoc test: *P ≤ 0.05, **P ≤ 0.01 compared with the control group. The percentage inhibition of inflammation of each treatment is shown above each column. AE, aqueous extract; BF, butanolic fraction; FF, flavonoid fraction; SF, saponin fraction.
Several mediators, inhibitors and regulatory molecules are involved in triggering complex responses during the inflammatory process. MPO and NO are characterized for presenting some pro-oxidative and pro-inflammatory properties. The first one is considered as a marker of neutrophil content in inflamed tissues[28] and the second one, as an index of nitric oxide synthase (NOS) activity.[29] Therefore, after evaluating the anti-inflammatory activity in a sub-chronic mouse ear oedema model, a homogenate was obtained from a portion of each treated ear. The inhibition of MPO and NOS activity was evaluated. Most of the treatments exhibited MPO inhibition percentages under 50% but inhibited NOS activity above 70%. The results of this evaluation are shown in Table 1.
Effect of P. edulis f. edulis leaves fractions and compounds on myeloperoxidase and nitric oxide synthase activity
| Treatment | Percentage inhibition of myeloperoxidase activity | Percentage inhibition of NOS activity |
|---|---|---|
| Dexamethasone | 10.5 ± 3.2 | 77.2 ± 1.9* |
| AE | 15.3 ± 3.6* | 82.8 ± 3.5* |
| BF | 0.0 ± 0.0 | 83.2 ± 2.2* |
| SF | 16.7 ± 2.5* | 92.2 ± 1.6* |
| FF | 21.7 ± 1.2* | 88.2 ± 3.1* |
| Compound 1 | 59.2 ± 1.8* | 71.4 ± 5.3* |
| Compound 2 | 24.0 ± 0.6* | 87.0 ± 0.3* |
| Treatment | Percentage inhibition of myeloperoxidase activity | Percentage inhibition of NOS activity |
|---|---|---|
| Dexamethasone | 10.5 ± 3.2 | 77.2 ± 1.9* |
| AE | 15.3 ± 3.6* | 82.8 ± 3.5* |
| BF | 0.0 ± 0.0 | 83.2 ± 2.2* |
| SF | 16.7 ± 2.5* | 92.2 ± 1.6* |
| FF | 21.7 ± 1.2* | 88.2 ± 3.1* |
| Compound 1 | 59.2 ± 1.8* | 71.4 ± 5.3* |
| Compound 2 | 24.0 ± 0.6* | 87.0 ± 0.3* |
Data are expressed as mean ± SEM, n = 6. One-way ANOVA, Dunnett’s post-hoc test.
*The mean difference is significant at P ≤ 0.05 level compared with the control group.
Effect of P. edulis f. edulis leaves fractions and compounds on myeloperoxidase and nitric oxide synthase activity
| Treatment | Percentage inhibition of myeloperoxidase activity | Percentage inhibition of NOS activity |
|---|---|---|
| Dexamethasone | 10.5 ± 3.2 | 77.2 ± 1.9* |
| AE | 15.3 ± 3.6* | 82.8 ± 3.5* |
| BF | 0.0 ± 0.0 | 83.2 ± 2.2* |
| SF | 16.7 ± 2.5* | 92.2 ± 1.6* |
| FF | 21.7 ± 1.2* | 88.2 ± 3.1* |
| Compound 1 | 59.2 ± 1.8* | 71.4 ± 5.3* |
| Compound 2 | 24.0 ± 0.6* | 87.0 ± 0.3* |
| Treatment | Percentage inhibition of myeloperoxidase activity | Percentage inhibition of NOS activity |
|---|---|---|
| Dexamethasone | 10.5 ± 3.2 | 77.2 ± 1.9* |
| AE | 15.3 ± 3.6* | 82.8 ± 3.5* |
| BF | 0.0 ± 0.0 | 83.2 ± 2.2* |
| SF | 16.7 ± 2.5* | 92.2 ± 1.6* |
| FF | 21.7 ± 1.2* | 88.2 ± 3.1* |
| Compound 1 | 59.2 ± 1.8* | 71.4 ± 5.3* |
| Compound 2 | 24.0 ± 0.6* | 87.0 ± 0.3* |
Data are expressed as mean ± SEM, n = 6. One-way ANOVA, Dunnett’s post-hoc test.
*The mean difference is significant at P ≤ 0.05 level compared with the control group.
Discussion
Seven specimens of P. edulis have been described in the literature, but it is not clear whether they are recognized as different intraspecific categories within the species or as synonymous names.[30, 31] At least two specimens have been differentiated as two botanical forms: P. edulis f. flavicarpa O. Deg and P. edulis f. edulis.[2, 6] A form or forma is a category below variety in the botanic nomenclature, which refers to different phenotypes with no geographic, ecologic or phylogenetic integrity. On the contrary, varieties are coherent evolutionary subsets of a species that imply such characteristics.[32]
Some phenotypical differences between P. edulis f. flavicarpa O. Deg and P. edulis f. edulis are the flowers, fruit colour/shape (yellow and purple passion fruit, respectively), dark brown seeds in the first case and black seeds in the second one. Additionally, according to the genetic study of de Andrade Aukar et al., there is a low rate of similarity between both forms.[5]
According to the review published by He et al., more than 110 compounds have been reported for different parts of P. edulis Sims plants, of which, 33 correspond to flavonoids and 29 to triterpenoids; being impossible to identify the chemical composition between the forms since it has not been specified.[8] However, for greater clarity of the information reported for the drug studied in this research, the Supplementary Material shows the compilation of articles found about the analysis and identification of flavonoids and triterpenoids present in the leaves of P. edulis Sims (without specification of the form studied) and P. edulis f. edulis (purple passion fruit, purple maracuja, gulupa).
Although most of the studies performed the identification of flavonoids by mass spectrometry, the lack of differentiation in the studied form in the literature makes it difficult to analyse the chemical composition of flavonoids between species. Comparisons of the fragmentation patterns of flavonoids found in P. edulis Sims or P. edulis f. flavicarpa O. Deg. were indistinctly made in most of the references that used such technique (see Supplementary Material).
On the other hand, although previous studies highlight differences in the chromatographic profiles between the forms flavicarpa and edulis,[15–17] Li et al. focused mainly on the chemical study of P. edulis f. flavicarpa O. Deg.[17]; in Zucolotto et al., it was not possible to identify by LC-MS the peaks observed for P. edulis f. edulis[16]; and Ayres et al proposed the presence of vitexin-2″-O-rhamnoside and luteolin-7-O-glucoside in P. edulis var edulis by co-injection with commercial standards.[15] In our research, the low concentration of flavonoids in the extracts of P. edulis f. edulis leaves prevented their isolation. However, the analysis by LC-MS allowed to determine the presence of vicenin-2; 6,8-di-C-glycosylchrysin and spinosin. Although these flavonoids have been described for P. edulis f. flavicarpa O. Deg.,[15–17] here for the first time, they are identified in P. edulis f. edulis leaves.
Six of the 13 articles in the Supplementary Material, for the undefined species P. edulis or P. edulis Sims, reported the identification of various cycloartane triterpenes such as cyclopassifloic acids,[27] passiflorine, a derivative of these,[33] and a wide variety of cyclopassiflosides.[26, 27, 34, 35] Most of them describe the presence of a large number of triterpenoids for the same unidentified species, such as the reports of Yoshikawa et al.[26, 27] where a total of 18 triterpenoids were isolated from P. edulis. On the contrary, our results indicate a low diversity of saponins (three) within P. edulis f. edulis composition. Thus, the results of this research represent the first identification of the saponins found specifically in the leaves of P. edulis f. edulis.
Regarding anti-inflammatory activity (Figure 4), the lower effect in reducing oedema obtained for the AE in comparison with the BF was expected due to the concentration of compounds in the BF. Additionally, the higher percentage inhibition of inflammation for the FF over the SF suggests that the flavonoid-rich mixture has greater potential as an anti-inflammatory than the saponin mixture.
Positive results for the flavonoid-enriched fraction could be expected, considering that the identified compounds, vicenin-2 and spinosin, have shown significant anti-inflammatory effects in other studies.[36, 37] But in the case of the fraction enriched in saponins, the results were of particular interest since, it has been reported that according to the type of triterpenoid skeleton and the substitution pattern of the same can vary the interaction with different target enzymes related to the inflammatory process, such as COX-1 and -2; LOX-5; MPO, PLA2 and iNOS.[38, 39, 40] Specifically, for cycloartane-type triterpene glycosides, very few studies of anti-inflammatory activity were found, especially focused on the species Astragalus membranaceus.[41–43] The group of compounds identified for this species, called agroastragalosides, show significant differences in the substitution pattern to the cycloartane glycosides identified in P. edulis f. edulis, the cyclopassiflosides. Less interaction between these types of compounds and target enzymes could explain the lower anti-inflammatory activity observed for this fraction, compared with FF. However, these results would represent the first study of anti-inflammatory activity carried out for cyclopassiflosides. On the contrary, some mechanisms of action have been proposed for the C-glycosyl flavonoids, such as the reduction of leucocytes, neutrophils, mononuclears, the inhibition of TNF-α expression, and NO production, mediated by vicenin-2[36, 37, 44]; or the inhibition of leucocytes and neutrophils by spinosin.[36]
Since the BF showed similar results compared with the isolated compounds, it is proposed that no purification process of the metabolites of P. edulis f. edulis leaves is required to obtain anti-inflammatory activity from this drug but rather an enriched fraction in flavonoids and saponins.
MPO is a heme-peroxidase enzyme mainly expressed by neutrophils and monocytes. It is released from the azurophilic granules of neutrophils and when reacting with superoxide or hydrogen peroxide, produces highly reactive oxidants such as chloride, hypochlorous acid and chloramines. These products exert an important microbicidal effect during phagocytosis in the inflammation process.[45, 46] Although most of the P. edulis f. edulis treatments evaluated in these assays showed significant statistical difference to the control group (Table 1), only compound 1 presented high inhibition of MPO (higher than 50%). However, it is once again noticed that the SF presented a lower effect than the FF. This could be related to the fact that the saponins identified for P. edulis f. edulis do not have the structural characteristics described in the literature for a good theoretical binding between triterpenoids and MPO, where it has been seen that the type of substituents is more important than the type of skeleton for their interaction. For example, carboxylic groups in positions 21, 28 or 30 and the presence of voluminous aromatic groups in position 3.[40] Additionally, the result obtained for the FF agrees with the ineffectiveness found by Zucolotto et al., for vicenin-2 and spinosin in inhibiting MPO activity.[36] Considering the results of MPO activity, it is inferred that this drug does not inhibit neutrophil migration to the inflamed area as part of its anti-inflammatory effect.
On the other hand, the results of inhibition of NOS activity showed to be more promising since all the treatments presented significant percentages of inhibition, most of them even higher than the positive control dexamethasone (77%). NO synthase inhibitors are potential beneficial agents for the treatment of conditions associated with an overproduction of NO, such as neurodegenerative disorders, septic shock or inflammation.[47] On the latter, NO is considered as a mediator of both acute and chronic inflammation. In early stages, it is produced by the endothelial isoform of NOS (eNOS) as an immune cell function and oedema formation regulator; while in chronic inflammation, the inducible isoform (iNOS) activity is increased, leading to indirect effects of NO derived reactive species that cause deleterious/pro-inflammatory properties.[47–49] Therefore, given the complex biological role of NO, other tests are required to further characterization of P. edulis f. edulis anti-inflammatory mode of action.
Several articles have been published, highlighting the anti-inflammatory potential of P. edulis Sims, most of them without clarification of the studied form. For example, Cazarin et al. have evaluated the effects of passion fruit peel flour intake in different in-vivo models of colitis[50, 51] and the antioxidant/anti-inflammatory activity of the AE of the leaves.[52–54] All five related studies show significant results apparently about P. edulis f. flavicarpa O. Deg. since one of them referenced it as the yellow passion fruit.
On the other hand, the researchers Benincá et al.,[55] Montanher et al.,[56] Vargas et al.[57] and Zucolotto et al.[36] focused on studying the anti-inflammatory action of the AE, fractions and isolated compounds (C-glucosylflavones) from the leaves of P. edulis f. flavicarpa O. Deg. finding inhibition of leucocytes, neutrophils, MPO, NO, TNF-α and IL-1β as significant mechanisms.[36, 55–57] Furthermore, the pectin and rinds (epicarp and mesocarp) of the same species, significantly decreased neutrophil filtration[58] and peroxidase activity of MPO,[59] respectively. A polysaccharide fraction of the fruits possibly reduces inflammation by modulation of histamine and serotonin liberation or synthesis.[60]
Contrastingly, few articles described the anti-inflammatory activity of P. edulis f. edulis and none related to the activity of the leaves. According to Herawaty and Surjanto, the peel ethanolic extract at 400 mg/kg body weight showed significant anti-inflammatory effects in a paw oedema model with pletismometer.[61] More recently, Carmona et al.[62] determined that the polyphenolic extracts obtained from the pulp fruit of P. edulis f. flavicarpa O. Deg., P. edulis f. edulis and P. ligularis Juss, significantly inhibited the inflammation-induced intestinal barrier dysfunction in Caco-2 cell monolayers. In this sense, the presented results could indicate the first report of the anti-inflammatory potential of P. edulis f. edulis leaves.
Conclusions
This research constitutes the first study that uses both chromatographic and spectroscopic techniques for the chemical analysis specifically of P. edulis f. edulis leaves. This allowed us to propose for the first time the presence of three flavonoids: vicenin-2, 6,8-di-C-glycosylchrysin and spinosin. In addition, three cycloartan-type triterpenic saponins known as cyclopassiflosides IX,[1] XI[2] and III[3] were identified.
The anti-inflammatory activity of the AE, fractions and pure compounds isolated in sufficient amounts from the leaves of P. edulis f. edulis was evaluated. The BF and isolated compounds were active in the mouse ear oedema model induced by multiple topical application of TPA, highlighting the activity of the flavonoid-rich fraction. It is suggested as a possible mode of action for the evaluated substances – the inhibition of the NOS activity.
Author Contributions
N.U., L.C., F.A.R., and G.M.C. conceived, designed and analyzed the chemical experiments; N.U., M.A. and L.F.O. conceived, designed and analyzed the pharmacological experiments; N.U. performed all the experiments; P.S., M.A., and L.C. wrote the paper. All authors contributed to the final version of the manuscript. All authors have read and agreed to the published version of the manuscript.
Funding
The ANLA and Ministerio de Ambiente y Desarrollo Sostenible granted permission to collect samples and perform this research ‘contrato de acceso a recursos genéticos y sus productos derivados No. 244’. The authors acknowledge Fondo Nacional de Financiamiento para la Ciencia, la Tecnología y la Innovación, Francisco José de Caldas for funding this project through contract No. 0459-2013, Red Nacional para la Bioprospección de Frutas Tropicales-RIFRUTBIO. The project ‘Contribución al estudio fitoquímico y a la evaluación de la actividad antiinflamatoria de algunas especies del género Passiflora’ (HERMES code 23562) was funded by Universidad Nacional de Colombia..
Conflict of Interest
None declared.
References
