­Chemical profiling and biological activity of Peperomia blanda (Jacq.) Kunth

Plant collection and extraction

Peperomia blanda (Jacq.) Kunth was collected from the Socotra Island, Yemen (Diksam-sikand), with approval number 120025 and, 120026 from the General Directorate of Plant Protection, Yemen, and identified by Dr. Abdul Wali Al-Khulaidi, Department of Botany, Taiz University. The voucher specimen number PPLS/2013/90/1 was placed at the Faculty of Pharmacy, Sana’a University. Plant samples were oven-dried for 7 days at 30 °C. The dried sample was milled and stored at room temperature for further analysis. Dried plant sample weighing 50 g was successively extracted by Soxhlet extraction using solvents with different polarities, starting with non-polar petroleum ether, then dichloromethane and methanol. The resulting solvent extracts were concentrated under reduced pressure to obtain crude extracts for further biological studies and compound isolation. All solvents used were of analytical grade (Merck, Darmstadt, Germany).

Chemical profiling determination

Chemical profiling of the different solvent extracts was carried out using gas chromatography mass spectrometer (Narayanamoorthi, Vasantha & Maruthasalam, 2015) QP2010 Plus (Shimadzu, Kyoto, Japan) with an auto sampler. The petroleum ether and dichloromethane extracts were dissolved in chloroform, and the methanol extract was dissolved in methanol, and 1 µL of each sample (100 µg/mL) was added to the column. A fused HP5-MS silica capillary column (30 m × 0.32 mm, film thickness 0.25 µm) was used. The carrier gas was helium, and 1:50 split ratio was used. The oven temperature was maintained at 50 °C for 5 min, after which it was slowly increased at a rate of 5 °C per min to 240 °C and maintained at 240 °C for 5 min. The injection port temperature was 250 °C and the spectra of the constituents of the sample solution were matched with the spectra of identified compounds kept in the internal library (Wiley; mass spectral library).

Isolation of compounds

The in vitro cytotoxic and antioxidant assays of the crude solvent extracts were evaluated and showed activities. Therefore, to isolate bioactive compounds, petroleum ether and dichloromethane extracts were fractionated using column chromatography (50 × 3.5 cm) on silica gel and run with solvent gradients of chloroform:EtOAc (100:0, 95:5, 90:10, 85:5, 80:20, 75:25, 70:30 until 0:100 %v/v, each 400 mL), EtOAc:MeOH (100:0, 95:5, 90:10, 85:5, 80:20, 75:25, 70:30 until 0:100 %v/v, each 400 mL), and finally with 800 mL MeOH. A total of 200 sub-fractions (50 mL) were collected based on their TLC (8:2 of CHCl3:EtOAc, solvent system) into four main fractions of petroleum ether and three fractions of dichloromethane. Fraction 2 of the petroleum ether extract was further chromatographed with an n-hexane:EtOAc solvent system on silica gel to afford small needle crystals of N,N′-diphenethyloxamide and phytosterol mixture (stigmasterol and β-sitosterol). Dichloromethane fraction 2 was subjected to silica gel column chromatography for fractionation and purification using the gradient elution method with an n-hexane:EtOAc solvent system. Recrystallization was done by dissolving crystal in small amount of ethyl acetate and drops of hexane were added till turbid solution start to appear. Then solution left to dry and crystal start to form.

Structural elucidation

Chemical structures of isolated compounds were elucidated using FT-IR/FT-FIR (Perkalin), GCMS (QP2010 Plus, Shimadzu, Kyoto, Japan) and 1D and 2D NMR (Nuclear Magnetic Resonance, JEOL, Peabody, MA, USA) spectroscopy at 500 MHz with deuterated chloroform as solvent. The crystal compounds were identified via single crystal X-ray structure determination (Oxford Diffraction Xcalibur Gemini S diffractometer; Agilent, Santa Clara, CA, USA) equipped with CuKα radiation, (k = 1.5418 Å). The data were processed using CrysAlis software and Empirical absorption correction using spherical harmonics was implemented using SCALE3 ABSPACK scaling algorithm (Xcalibur, 2007). The figures were produced using MERCURY software (Macrae et al., 2008).

Physical, chemical and spectroscopic data of the compounds

Biological activity

Total phenolic and flavonoid contents determination

The total phenolic contents extracted were estimated by the Folin–Ciocalteu reagent (Narayanamoorthi, Vasantha & Maruthasalam, 2015; Wei et al., 2011). A seven-point standard curve (0–50 mg/L) for gallic acid (≥ 97%, GA) was built to calculate total phenolic percent as Gallic acid equivalent (mg GA/g dry extract). Analyses were performed in triplicates. The extract total flavonoid contents were evaluated based on the aluminum chloride colorimetric method (Wei et al., 2011; Xcalibur, 2007), using quercetin as standard. Seven-point standard curve (0–1,000 mg/L) of quercetin was used and the levels of extracts total flavonoid content were calculated as quercetin equivalent (mg quercetin/g dry extract).

Chemical profiling

The results obtained from GCMS investigation of the crude extracts led to the identification of several compounds. The different extracts showed 53–57 peaks (Fig. 1).

The major peaks are summarized in Table 1 and the most commonly identified compounds were fatty acids, sesquiterpenes, palmitic acid, vitamins and steroids. Those compounds have been reported to have diverse biological activities, including antioxidant, antimicrobial, anticancer and anti-inflammatory activities (Narayanamoorthi, Vasantha & Maruthasalam, 2015). In addition, hydroquinone was identified, and has been reported to be used for the treatment of hyperpigmentation.

It can be seen from the data of GCMS in Table 1 that the presence of N,N′-diphenethyloxamide, β-sitosterol and stigmasterol with molecular ion peaks at 296, 414 and 412 (m/z), respectively, could be detected in petroleum ether with (peak number, retention time (RT)) values of (22, 36.23), (52, 57.33) and (56, 58.58) min. Moreover, GCMS analysis of the dichloromethane extract showed the presence of peperomin A with a molecular ion peak, peak number and RT of 414, 52 and 58.58 min, respectively.

Isolation of compounds

The active petroleum ether and dichloromethane extracts of P. blanda were repeatedly chromatographed on silica gel and this successfully led to the isolation of three compounds (Fig. 2), and the chemical structures were elucidated using 1D and 2D NMR, GCMS and X-ray diffraction for crystals obtained.

Single crystal X-ray structure determination

Peperomin A was isolated in crystal form for the first time, and the X-ray crystal structure was deposited with the Cambridge Crystallographic Data Centre (CCDC) with deposition number CCDC 1529063. The structure of peperomin A was resolved in the centrosymmetric triclinic space group P 2₁ 2₁ 2₁, with one molecule in the asymmetric unit (Table 2). Lattice parameters have been determined as 11.3200 Å, b 11.6312 Å, c 14.5155 Å and the volume of the unit cell is found to be 1911.19 ų. R-configuration is present at the chiral center C3 and S-configuration at the chiral center C5 of 2-methyl-butyl lactone ring. The x-ray molecular structure of peperomin A showed the same information found in solution by NMR spectroscopy that were also comparable with data reported by Felippe et al. (2011).

From 1D and 2D NMR data, it is apparent that ¹H NMR spectrum displayed resonances of four meta-coupled aromatic hydrogens at 6.41 (H-6′,6″, d, J = 1.95 Hz), 6.33 (H-2′,2″, J = 1.5 Hz), two methylenedioxy groups at 5.90–5.92 (d, J = 1.2 Hz), and two O-methyl singlets at 3.87. This set of signals characterized two 3,4-methylenedioxy-5-methoxy phenyl rings with magnetically non-equivalent methylenedioxy protons as a result of the anisotrophy from aromatic ring (Felippe et al., 2011). A doublet at δ 0.90 (H-6, d, J = 8.0 Hz) was assigned to the methyl group, and multiplets at 3.76 (H-4a) and 4.27 (H-4b) were due to the methylene group of the butyrolactone moiety. The resonances at δ 2.31 (H-2, m), 2.83 (H-3, m), and 3.54 (H-5, d, J = 10.9 Hz), due to three methine groups, were then observed. A Ϩ-butyrolactone ring was confirmed via HMBC between H-2, H- 4, and H-6 and the lactone carbonyl carbon at 179.9 ppm (C-1). The ¹³CNMR data corroborated the presence of the butyrolactone system, and all signals involved were accordingly assigned based on the HMBC data.

N,N′-diphenethyloxamide is a symmetric oxamide that have synthesized before to serve as a simple model for hydrogen bonding interaction that could be recoded in peptides and proteins (Martínez-Martínez et al., 1998). However to our knowledge, this is first time that it is isolated from the natural source. In this study, N,N′-diphenethyloxamide is the second crystal isolated for the first time from Peperomia species and the crystal structure was deposited with the CCDC with deposition number CCDC 1529062. The structure of N,N′-diphenethyloxamide (Fig. 2) was resolved in the centrosymmetric monoclinic space group P 2₁/c, as summarized in Table 2. The crystal oxamide group has two independent amides that disregard a π-conjugation through the C-C central bond, as shown by the average bond length of the oxalyl (OC-CO) of 1.2306 Å and torsion angle value of −180.0 °, in comparison to the oxalyl (OC-CO) average bond length of 1.541 Å and torsion angle of 180.0 for a trans conformation of N,N′-bis(2-hydroxyphenyl)oxamide and torsional angle between 90–115° for other oxamide derivatives reported previously (Martínez-Martínez et al., 1998). In crystal packing, the nitrogen atoms (N and N′) of oxamide was hydrogen bonded to the oxygen atom of the carboxylic acids through N-H⋯O intermolecular interaction establishing a sheet along the axes (Fig. 3) and the diffraction data details are summarized in Table 2.

The data for ¹HNMR chemical shifts of Ar-CH₂ and, CH₂-N appeared at 2.24 and 3.50 ppm, and aryl hydrogen signals that are found at 7.22 ppm that are supported by previous data recorded for the synthesis of N,N′-diphenethyloxamide (Martínez-Martínez et al., 1998). UV absorption bands were obtained at 290, and 382 nm. The IR spectrum recorded the presence of different bands at 2993 to indicate the presence of NH group, and 1635 (C=O), 1459, 1044 (C-H) 1364 (C-N). These results are consistent with the data obtained from IR spectrum of synthesized N,N′-diphenethyloxamide that recorded IR ν_max (KBr)/cm⁻¹ at 3,305, 1,653, 1,520, 1,495, 1,224, 1,189 and 697. Another important finding was the peak record from; UPLC/Q-TOF-MS m/z [M+1]⁺ at 296.300 for C₁₈H₂₀N₂O₂. The X-ray molecular structure of N,N′-diphenethyloxamide supported that data obtained from ¹HNMR, IR, mass spectroscopy and the previous recorded study for the synthesized N,N′-diphenethyloxamide.

Biological activity

Preliminary screening of the extracts using MTT assay showed cytotoxic activity against different cell lines. The petroleum ether extract exhibited IC₅₀ values of 9.54 ± 0.30 and 4.3 ± 0.90 µg/mL against HL-60 and WEHI-3 cells, respectively. The dichloromethane extract showed IC₅₀ values of 14.42 ± 1.52 and 15.58 ± 1.17 µg/mL against HL-60 and WEHI-3 cells, respectively. The methanol extract showed lower activity against HL-60 and WEHI-3 cell lines. However, methanol and petroleum ether extracts showed promising activity against MCF-7, with IC₅₀ values of 10.49  ± 0.79 and 5.39 ± 0.34 µg/mL, respectively. The extracts showed low cytotoxicity against normal cells (MCF10A, WRL-68 and HDFa), with IC₅₀ values greater than 100 µg/mL (Table 3).

Peperomin A demonstrated cytotoxic activity against different tested cell lines with selectivity index values greater than 1 for cancer cells (Jokhadze et al., 2007). Peperomin A exhibited cytotoxic activity against MCF-7 with an IC₅₀ of 5.58 ± 0.47 µg/mL, compared to a standard drug, Taxol that has an IC₅₀ of 3.00 ± 0.07 µg/mL (Table 4). Moreover, peperomin A showed cytotoxic activity against HL-60 and WE-HI cells with IC₅₀ values of 16.36 ± 0.62 and 4.62 ± 0.03 µg/mL, respectively. The isolated mixture of phytosterol (stigmasterol and β-sitosterol) showed anti-proliferative activity against WEHI-3, HL-60 and MCF-7, with IC₅₀ values of 8.06 ± 0.16, 9.84 ± 0.61 and 8.94 ± 0.05 µg/L, respectively, as summarized in Table 3. The in vitro cytotoxic activities of different solvent extracts of P. blanda, peperomin A and phytosterol reported herein are in accordance with previous reported studies in which, compounds such as secolignans and polyketides, previously isolated from P. ducluoxii, showed moderate growth inhibitory activity against VA-13 cells (malignant lung tumor cells) (Li et al., 2007) and secolignans isolated from P. pellucida, inhibited the growth of HL-60 (human promyelocytic leukaemia), MCF-7 and HeLa (cervical cancer cells) (Xu et al., 2006). The secolognin and the polyketide isolated from P. sui showed cytotoxic activity against HONE-1 and NUGC-3 cell lines (Cheng et al., 2003). Peperomin E showed anti-proliferative activity against human gastric carcinoma SGC-7901, BGC-823 and MKN-45 cell lines, and against non-small-cell lung cancer (NSCLC) cell lines. It also inhibited the proliferation of gastric cells by apoptosis induced via the mitochondrial and PI3K/Akt signalling pathways and exhibited anticancer activity in lung cancer by inhibiting DNMT1 expression and activity (Wang et al., 2016).

In addition, peperomin A and B showed moderate inhibitory effects on HIV-1 IIIB growth in C8166 cells (i.e., a cytopathic effect, CPE) and cytotoxic activity against the C8166 cell line. Previous studies have also reported the chemo protective activity of β-sitosterol in breast and colon cancer cell lines through inhibition of proliferation of the cancer cells, lowering the expression of β-catenin and PCNA, or activating Fas signaling (Awad et al., 2007; Baskar et al., 2010; Chai, Kuppusamy & Kanthimathi, 2008).

Radical scavenging activity using the DPPH assay was tested for all extracts, peperomin A and phytosterol mixture. The inhibitory concentrations (IC₅₀) are summarized in Table 5.

The methanol extract exhibited antioxidant activity with an IC₅₀ of 36.81 ± 0.09 µg/mL, followed by the dichloromethane extract at 61.78 ± 0.02 µg/mL and 203.80 ± 0.19 µg/mL for the petroleum ether extract. The ferric reducing capability (FRAP) of the three crude extracts was also determined. The extracts exhibited weak ferric reducing ability ranging from 162.2 ± 0.80 to 381.5 ± 1.31 µg/mL, as compared to the standards, vitamin C and BHA, with values of 2603.96 ± 10.54 and 1738.8 ± 9.53 µg/mL, respectively (Table 5).

The peperomin A and phytosterol mixture displayed weak free radical scavenging activity and ferric reducing power indicating that the antioxidant activity recorded in the crude extracts may be due to the presence of other secondary metabolites such as flavonoids and vitamin E that were observed from GCMS and analysis of flavonoid content in the crude extracts. Other reported compounds from Peperomia, such as two C-glycosyl-flavones isolated from P. blanda (Velozo et al., 2009), phenylpropanoid, benzopyran, chromone, prenylated quinone, secolignan and acylcyclohexane-1, 3-dione could also be responsible for the antioxidant activities observed.

The total phenolic content ranged from 16.42 ± 1.30 mg GAE/g in the methanol extract to 31.00 ± 0.44 mg GAE/g in the petroleum ether extract (Table 5). Moreover, the obtained results showed that total flavonoid content in the dichloromethane extract (232.0 ± 1.9 mg QE/100 g) was higher than in the petroleum ether extract (214.3 ± 1.3 mg QE/g) and the methanol extract (199.0 ± 1.7 mg QE/g). The phenol content varied due to the different solubility of phenol in each solvent extract. The more lipophilic phenols were extracted more readily in petroleum ether and dichloromethane, and their results were similar, while the polar phenols were more readily extracted by methanol. The phenol and flavonoid contents in each solvent extract led to different observations of antioxidant activities. Natural antioxidants like phenolic acids and flavonoids are interesting alternatives to modulate inflammation, and to inhibit the related oxidative processes (Guilhon-Simplicio et al., 2017). Consequently, further isolation of new flavonoids from this plant are warranted for drug discovery.