General Methods

Column chromatography was carried out on silica gel (230-400 mesh, Macherey Nagel, Düren, Germany) and macro-porous resin DIAION HP-20 (Mitsubishi Chemical Corporation, Tokyo, Japan). Analytical high performance liquid chromatography (HPLC) was carried out on an Agilent 1100 Series liquid chromatograph equipped with an Agilent DAD-G1315A detector for the UV spectrum acquisition (Agilent Technologies, California, United States) and the data acquisition was done using ChemStation (Version B.04.03). Columns RESTEK Ultra AQ C18 (3 µm, 100 x 3.2 mm, RESTEK, Pennsylvania, United States) and Agilent ODS Hypersil (5 µm, 250 x 4 mm, Agilent Technologies, California, United States) were used for analytical purpose. Thin-layer chromatography (TLC) was conducted over pre-coated silica gel 60F₂₅₄ plates (Merck, Darmstadt, Germany) and spot detection was performed under UV light (λ = 254 and 366 nm) and then spraying with 1 % AlCk in ethanol (EtOH) (Sigma-Aldrich, Missouri, United States). Nuclear magnetic resonance (NMR) spectra were recorded on a Bruker DRX-400 NMR Spectrometer (Bruker Biospin Gmbh, Rheistetten, Germany) with solvent residual peaks of dimethyl sulfoxide (DMSO)- ^_d6 at δH 2.5 and δc 39.52 as references.

Antioxidant colorimetric assays using DPPH ^• and ABTS ^{• +} (Sigma-Aldrich, Missouri, United States) and phenolic and flavonoid quantifications, using Folin-Ciocalteu & AlCl₃ (Sigma-Aldrich, Missouri, United States), respectively, were assessed on a microplate spectrometer Multiskan GO (Thermoscientific, Massachusetts, Unites States). Infrared Spectroscopic data were acquired in an Agilent Cary 630 FTIR (Agilent Technologies, California, United States) with a Windows compatible version of MicroLab Software (Agilent Technologies, California, United States).

Plant Material

Leaves and twigs of Alchornea coebphylla (Euphorbiaceae) were collected in a natural protected zone known as Bremen-La Popa at the coordinates 4° 40’ 48.6’’ North and 75° 37’ 32.7’’ West at 6737 feet above sea level, near to the municipality of Filandia (Quindio, Colombia), this procedure was completed under the contract of access to genetic sources for scientific research without commercial interest number 56 of February 28^th of 2016 granted by the Ministry of Embironment. A voucher specimen was deposited at the Herbarium of the Universidad de Antioquia (Medellin, Colombia), labelled as FJR-3969. The botanical identification was done by Professor Francisco Javier Roldan from Universidad de Antioquia, Medellin, Colombia.

Extraction and Isolation

The collected plant material (leaves and twigs) was dried in a laboratory stove at 50 °C and further ground with a hammer mill (Nogueira (0: 5.1 mm)). Afterwards the dried material (1.212 g) was exhaustively extracted at room temperature allowing it to stand overnight using 10 L of n-hexane (HEX), dichloromethane (DCM) and methanol (MeOH), with an increasing polarity order. The concentration of the respective solutions under vacuum produced the respective dried extracts.

Fifteen grams of methanolic leaf extract were subjected to open column chromatography with DIAION HP-20 resin (200 g, 74.5 x 90 mm) and eluted with MeOH-H2O mixtures (0:100, 20:80, 40:60, 60:40, 80:20, 100:0 (v/v) with 1.5, 1.25, 5.55, 4.6, 3.7 and 2 L, respectively). After pooling related fractions according to their TLC profiles, 29 fractions (DHP-1 to DHP-29) were finally obtained. Afterwards, a first exclusion step was applied based on the mass availability of each fraction. In this sense, 17 fractions with a mass greater than 300 mg were selected to continue the bio-guided isolation process. Thus, based on the correlation among mass availability and further antioxidant data, fraction DHP-11 (750 mg) were seeded onto a silica gel column (22.5 g, 20 x 150 mm) and then washed with approximately 10 bed volumes (BV) of chloroform (CHCl₃) and ethyl acetate (EtOAc) to eliminate low polarity impurities, to finally elute the sample with EtOAc-MeOH-AcOH-H₂O (70:20:5:5) collecting fractions DHP-11-A to DHP-11-G. The presence of the target compound in the fraction DHP-11-C was confirmed by analytical RP-HPLC using a 100 % MeOH system as the mobile phase at a flow rate of 1 mL. min^-1 at a retention time of 1.010 min. Spectral data were also recorded.

Antioxidant Activity

The antioxidant activity was evaluated along the isolation process to ensure the isolation of a biologically active flavonoid. In this sense, biological activities were assessed on the crude extract and fractions of the first chromatographic purification step following the methodology described by Brand-Williams & Berset ¹⁴ and Re (15) for the antioxidant assays of DPPH ^• and ABTS ^{• +}, respectively, with minor modifications in order to perform the experiments in a 96-well microplate. Measurements were done by triplicate and with two different repetitions in order to obtain statistically homogeneous results. IC₅₀ determinations and calibration curves were constructed in the same way and analysed with GraphPad Prism V. 5.01.

Methanolic crude extract and chromatographic fractions were evaluated at concentrations of 1,000 and 500 µg mL^-1, respectively. A 1,000 µg mL^-1 methanolic solution of hydroquinone, prepared the same day of the test, was used as positive control. Analogously a photometric blank to each sample extract was employed additionally with the photometric blank (the respective solvent mixture).

DPPH ^• Radical Assay

100 µL of a 20 µg. mL^-1 DPPH solution in MeOH (prepared just before the assay) were thoroughly mixed with 25 µL of the sample. The reaction was allowed to stand for 30 minutes in the absence of light and then the absorbance was measured at a wavelength of 517 nm. The photometric blank for the plant extracts consisted of 25 µL of the sample and 100 µL of MeOH.

ABTS ^{• +} Radical Assay

An aqueous 3.5 mM ABTS ^{• +} and 1.25 mM potassium persulfate solution was allowed to react 12 hours before the evaluation, then the absorbance of the final solution was adjusted to 0.7 with EtOH. Finally, 294 µL of the adjusted solution was added to each well where 6 µL of the sample extract had been transferred previously. The reaction was allowed to stand for 30 minutes in the darkness and at that time the absorbance was measured at 732 nm. The photometric blank for plant extract consisted in 6 µL of the sample and 294 µL of EtOH.

In order to be able to compare and report antioxidant activity results as equivalents of the same reference standard, calibration curves of 6-hydroxy-2,5,7,8-tetramethylchroman- 2-carboxylic acid (TROLOX) were constructed with concentrations ranging from 1 µM to 100 µM, using it as antioxidant reference compound ¹⁶.

Additional to the determinations of the biological activity, quantifications of phenolic compounds and flavonoids were considered, in order to determinate which could be the most promising fraction to continue the isolation process.

Quantification of Phenolic Compounds

This determination was completed following the Folin-Ciocalteu method, described by Magalhães ¹⁷. A. coelophylla extracts and fractions were evaluated at 50 µg. mL^-1 in methanol. The Folin-Ciocalteu reactive was also diluted in a proportion 1:50 with distilled water to adjust the absorbance of the solution into the range of 0.2 to 0.8. Then, 100 µL of NaOH 0.35 M were transferred to the same wells where previously 50 µL of the plant extracts and 50 µL of the diluted Folin-Ciocalteu reactive had been mixed. Afterwards, the reaction was allowed to stand for three minutes in the absence of light and the absorbance was measured at 760 nm in a microplate spectrometer MultiSkan Go.

In this determination the blank for each sample was also used and the photometric blank corresponded to 200 µL of water. Finally, to express the results as equivalents of some comparable parameter ¹⁷, a calibration curve of gallic acid was constructed at concentrations of 0, 2, 4, 8 and 16 µg. mL^-1.

Quantification of Flavonoid Content

The flavonoid content was assessed following the methodology described by Kim ¹⁸. The sample extracts were evaluated at a concentration of 100 µg. mL^-1 in methanol. First, 7.5 µL of 5 % aqueous NaNO2 were transferred to each well where the extracts were going to be evaluated. Then, 20 µL of each plant extract and 115 µL of distilled water were then pipetted. After 5 minutes of reaction, 30 µL of 2.5 % AlCl₃ were added while shaking; then, once 5 minutes of reaction took µLace, 50 µL of NaOH 1 M and 50 µL of distilled water were added and thoroughly mixed. Finally, 5 minutes after the NaOH addition the absorbance was measured at 500 nm.

Finally, in the same way of the phenolic compounds determination described above and to obtain these results as equivalents of some comparable parameter, a calibration curve with kaempferol was constructed at concentrations of 0, 0.05, 0.1, 0.2, 0.4 and 0.8 µg. mL^-1.

Determination of Antioxidant Activity and Bio-guided Isolation

The dried and grounded plant material of Alchornea coebphylla (1.212 g) subjected to successive extraction by maceration with hexane, dichloromethane and methanol, gave 91.0997 g of methanolic dried extract.

A phytochemical screening of the plant extracts revealed that the major flavonoid content occurred in the most polar methanolic leaf extract, which was fractionated in a first chromatographic partition based on polarity and molecular size with DIAION HP-20 resin, through which 29 fractions (DHP-1 to DHP-29) were collected. Mass and yield fractions are shown in Table 1.

Antioxidant activities were assessed over the 29 fractions through both DPPH and ABTS ^{• +} assays, in order to determine which fraction could be the most promising for the isolation process. Similarly, total phenolic and flavonoids contents were determined. Figure 1 summarizes the values of above mentioned evaluations for all the chromatographic fractions.

Table 1 Masses and yields for fractions of the first DIAION HP-20 chromatographic step. 

Figure 1 Determinations of the first column chromatographic fractions. A) Antioxidant activity through ABTS ^{• +}; B) Antioxidant activity through DPPH*; C) Total phenolic content and D) Total flavonoid content. 

After data collection, fraction 11 was selected and subjected to a second Silica gel 0.04-0.063 mm/230-400 mesh column to obtain 7 fractions (DHP-11-A to DHP-11-G). Finally the target compound was confirmed to be with a high grade of purity (> 98 % based on HPLC data) in fraction DHP-11-C, fraction that was then used in the spectroscopic data acquisition. Masses and yields of fraction of the second column chromatography are presented in Table 2. IC₅₀ values of the compound present in fraction DHP-11-C were 7.528 and 379.7 µg. mL^-1, through DPPH and ABTS ^{• +} assays, respectively.

Structural Elucidation of Isolated Compound

Apigenm-8-C-(α-L-rhamnopyranosyl-(1-›2)-β-D-glucopyranoside). Obtained as a yellowish powder. IR cm^-1: 3400, 2970, 2925, 1700, 1650, 1600, 1560, 1500, 1430, 1350, 1300, 1260, 1200, 1175, 1025, 830, 800. ¹ H NMR (DMSO-d6, 400 MHz) δ: 6.74 (1H, j, H-3), 6.28 (1H, j, H-6), 8.01 (2H, d, J=8.464 Hz, H-2’, H-6⁵), 6.93 (2H, d, J=8.475 Hz, H-3’, H-5⁵), 4.77 (1H, d, J=10.041 Hz, H-1”), 4.95 (1H, s, H-1””), 2.854.10 (10H, m), 0.46 (3H, d, J=5.911 Hz, H-6’”). ^B C NMR (DMSO-d ₆ , 100 MHz) δ: 164.00 (C-2), 102.43 (C-3), 182.10 (C-4), 161.20 (C-5), 98.29 (C-6), 162.34 (C-7), 104.47 (C-8), 155.83 (C-9), 104.22 (C-10), 121.6 (C-1⁵), 128.99 (C-2⁷), 115.89 (C-3% 160.67 (c-4⁵), 115.89 (C-5⁵), 128.99 (C-6⁷), 70.67 (C-1”), 75.09 (c-2”), 79.90 (C-3”), 70.46 (C-4”), 81.81 (C-5”), 61.17 (C-6”), 100.32 (C-1’”), 70.26 (C-2’”), 68.22 (C-3’”), 71.68 (C-4’”), 71.49 (C-5’”), 17.71 (C-6’”).

Table 2 Masses and yields for fractions of the second Silica gel chromatographic step. 

Determination of Antioxidant Activity and Bio-guided Isolation

The methanolic crude extract obtained by maceration gave a yield of 7.52 %.

The MeOH leaf extract was dissolved in methanol and fractionated by open column chromatography with DIAION HP-20 resin to obtain 29 fractions (DHP1 to DHP29) which then were classified and selected by mass, yielding 17 fractions with an availability greater than 300 mg. Afterwards, the antioxidant activity of the selected fractions was evaluated by the DPPH and ABTS ^{• +} methods at 500 gg mL^-1; in addition, the above quantifications were assessed to obtain more information that could suggest the most promising fraction to continue the isolation process.

With all the collected information regarding biological activities, and taking into account the amount available of each fraction, the 11th fraction was selected as the most promising to continue the isolation process with a mass of 955.53 mg, with 51.85 % of antioxidant activity under the DPPH assay, 99.75 % of antioxidant activity under the ABTS ^{• +} test, 28.49 gg mL^-1 of gallic acid equivalents and 0.42 gg mL^-1 of kaempferol equivalents. As a consequence, this fraction was subjected to a silica gel column fractionation yielding fractions DHP-11-A to DHP-11-G, which finally by analytical HµLC revealed the presence of a flavonoid glycoside in the fraction DHP-11-C.

Thus, based on the antioxidant percentage values obtained for fraction DHP-11-C and using the TEAC calibration curves, total equivalents of Trolox (µmol. mg-1) were found to be 108.7 and 564.67 for the DPPH and ABTS ^{• +} assays, respectively.

Figure 2 IC50 assay with DPH-11-C for A) ABTS^•+ and B) DPPH^•. 

Once the purity of the isolated compound was determined, the median inhibitory concentration (IC₅₀) was calculated for both of the above mentioned assays. The biological activity was assessed on the isolated compound at concentrations of 500, 250, 100 and 50 µg. mL^-1 for both the DPPH^• and ABTS^•+ methods. IC50 values of 7.528 and 379.7 pg. mL^-1 were obtained for the DPPH^• and ABTS^•+ determinations, respectively. Those results agree with previous investigations in which it was found that the methanolic crude extract of A. coelophylla had an IC₅₀ even lower than the reference control (Hydroquinone), which revealed IC₅₀ of 41.14 and 151.19 µg. mL^-1 for both antioxidant colorimetric assays and also that the antioxidant response was notoriously higher when evaluated through the ABTS^•+ determination than the antioxidant values obtained for the DPPH^• determination ¹⁹. Curves constructed for IC₅₀ determinations of the isolated compound are in Figure 2.

Structural Elucidation of Isolated Compound

The isolated compound, present in fraction DHP-11-C, was obtained as a yellowish powder in which 1H NMR spectrum showed two singlets at δ 6.74 and 6.28 (1H each, H-3 and H-6, respectively), four aromatic protons at δ 6.93 (2H, d, J=8.475 Hz, H-3’, H-5’) and 8.01 (2H, d, J=8.464 Hz, H-2’, H-6’) indicating para-substitution pattern at B aromatic ring, suggesting the presence of the apigenin moiety (See Figure 3). Additionally, the spectrum displayed two anomeric protons at δ 4.77 (1H, d, J= 10.041 Hz, H-1”) and 4.95 (1H, s, H-1’”) and one methyl signal at δ 0.46 (3H, d, J=5.911 Hz, H-6’”) indicating the presence of a β-glucosyl and α-rhamnosyl moieties ²⁰. The ¹³C NMR spectrum displayed signals of a carbonyl group at δ 182.10 (C-4), 14 aromatic carbons ranging 98.29 - 162.34, 2 anomeric carbons at δ 70.67 (C-1”) and 100.32 (C-1’”), 8 oxygenated carbons ranging from δ 68.22 to 79.9, one methyl group at δ 17.71 (C-6’”) and one aliphatic methylene at δ 61.17 (C-6”) (with negative phase (ꜜ) in DEPT 135° spectrum). These data indicated that the compound correspond to an apigenin diglycoside ^21,22. Shifting of C-8 signal downfield up to 104.47, absence of a H-8 signal and anomeric shifting of glucose suggest that glycosidic linkage is a C-linkage over such position. Finally, these findings could be supported with literature reported for the structure described and related compounds found in other plant species ^23,24,25,26.

Figure 3 2D NMR interactions observed in HMBC experiment for the isolated compound. 

As displayed in Figure 3, in the heteronuclear multiple-bond correlation (HMBC) spectrum, long range correlations were detected from H-1’’ (δ 4.77) to C-7 (δ 162.34), C-8 (δ 104.47) and C-9 (δ 155.83), from H-1’’ (δ 4.77) and H-1’’’ (δ 4.95) to C-2’’ (δ 75.09). Correlations involving aromatic protons were also observed. Thus, H-3 (δ 6.74) correlates to carbonyl (δ 182.10) and H-6 (δ 6.28) detected C-7, C-5 (δ 161.20) and C-4. Also symmetric correlations inside the B aromatic para-substituted system H-2’,6’ (δ 8.01) to H-3’,5’ (δ 6.93).

Finally, based on the spectroscopic data, bibliographic revisions and comparisons with theoretical predictions of the NMR spectra, the structure of the target compound would be apigenin-8-C-(α-L-rhamnopyranosyl-(1→2)-β-D-glucopyranoside), illustrated in Figure 4. This compound could also receive the name of Apigenin-8-C- neohesperoside and vitexin-2’’-0-rhamnoside.

Figure 4 Apigenin-8-C-(α-L-rhamnopyranosyl-(1→2)-β-D-glucopyranoside) isolated from MeOH extract of Alchorena coelophylla (Euphorbiaceae). 

Based on the proposed structure upon NMR spectroscopic data, this article is the first report of isolation of this compound from any Euphorbiaceae species ^{24,25,26,27,28,29,30}.