Chemicals and Reagents

Silica gel (100-200 mesh) used for column chromatography was from Sigma-Aldrich, St. Louis (Missouri, USA). Precoated TLC (silica gel 60 F254) and quercetin were purchased from Merck (Darmstadt, Germany). Rutin and β-sitosterol were purchased from Sigma-Aldrich (St. Louis, Missouri, USA). The solvents used for extraction and fractionation were of HPLC grade.

Extraction and isolation

Fresh fruit of L. siceraria was obtained from an agricultural farm of Pattoki (Punjab, Pakistan). The plant was identified by Dr. Khalid Rasib, associate professor of Biological Sciences, FC College University, Lahore. Fresh peel (7.85 kg) of the fruit was separated, ground, and macerated in methanol for 20 days (3 × 15 L) in a glass flask, at room temperature with regular shaking. The methanolic extract was filtered and pooled. The solvent was evaporated in vacuo using vacuum-assisted rotary evaporator (Buchi, Flawil, Switzerland) to obtain the methanolic extract as a gummy material. It was dried in an incubator at 40 °C for 2 days. The methanolic extract (458 g) was subjected to liquid–liquid extraction using solvents of increasing polarity using separating funnel. The extract was suspended in distilled water and exhaustively extracted by partitioning with hexane, chloroform, ethyl acetate, and n-butanolic successively leaving behind residual aqueous fraction. As the ethyl acetate fraction proved to be most active in our previous studies [4], it was selected for isolation and identification of chemical compounds using column and thin-layer chromatography.

Column chromatography

The ethyl acetate fraction (11.2 g) was subjected to CC using silica gel as the stationary phase. The dimensions of the column used were 57.96 cm × 1.54 cm. Wet slurry method was used for column packing. Weighed silica (24.8 g) was immersed in hexane overnight and was sonicated for 1 h to remove trapped air. A small piece of glass wool was deployed at its base to prevent silica from contaminating the eluents. The column was filled with hexane, and the slurry was poured from above leaving the tap open. Constant tapping by the side aided proper setting and packing of the column. Hexane was eluted 3 times before loading the ethyl acetate fraction, which was covered with a small amount of the silica gel slurry. Elution started with pure hexane and then gradually increasing ratio of ethyl acetate, followed by combinations of ethyl acetate and methanol with increasing polarity (Table 1).

Thin-Layer Chromatography (TLC)

Precoated TLC plates were used for further separation of the sub-fractions obtained from the column. The spots were scratched from the TLC plates and were dissolved in appropriate solvents and filtered. The filtrates were collected and analyzed using HPLC, GC–MS, and LC–MS. Various solvent systems were found appropriate for separation. The outcomes are shown in Table 2.

Locating agents used for TLC examination: (a) Folin–Ciocalteu reagent, (b) vanillin HCl (1 g vanillin + 10 mL conc. HCl), (c) vanillin H₂SO₄ (3 g vanillin + 95% CH₃OH [50 mL] + 2 M H₂SO₄ [15 mL]), and (d) 5% NaOH followed by drying for 1–2 min.

HPLC analysis

The compounds isolated from TLC were further purified and compared with their standards on HPLC. A flow rate of 1 mL/min was set on HPLC (Shimadzu [Kyoto, Japan] SPD-M10AVP) with an ultraviolet (UV) detector. Elution was done with mobile phase methanol–acetonitrile (90:10). The λ_max of the standard compounds was determined beforehand in a double beam UV spectrophotometer (Labomed Inc., Los Angeles, California, USA).

The λ_max of the standards was as follows: vanillin (250 nm), 3-tert-butyl-4-hydroxyanisole (290 nm), quercetin (254 nm), rutin (260 nm), β-sitosterol (206 nm), and stearic acid (245 nm).

LC–MS analysis

Liquid chromatography (LC) Agilent 1100 series (Bremen, Germany) fitted with binary pump, an auto-sampler with a 100 μL loop, a diode array detector set for recording at 254, 280, and 320 nm, and scanning from 200 to 600 nm was employed for the analysis. Ion-trap MS, fitted with an electrospray ionization (ESI) source, Bruker Daltonics HCT Ultra (Bremen, Germany) operating in full-scan with source temperature of 550 °C, spray voltage of −4500 V, declustering potential tuned to −255 eV, collision energy of −44 eV, entrance potential of −10 eV, cell exit potential of −35 eV, collision activated dissociation of 5 psi, curtain gas at 15 psi, mass dwell time of 400 ms, unit mass resolution of 0.1 m/z, operated in a quadrupole analyzer at <400 m/z, scan speed to 1000 m/z per second, and auto-MSn mode resulted in a fragment ion m/z pattern. The maximum fragmentation amplitude was set at 1 V, starting at 30% and ending at 200%. MS operating conditions were in negative mode and were optimized using 5-ocaffeoylquinic acid with a capillary temperature of 365 °C and a dry gas (nitrogen) flow rate of 10 L/min. The nebulizer pressure was set at 10 psi. Internal calibration was done with 10 mL of 0.1 M sodium formate solution injected through a six-port valve prior to each chromatographic run. Calibration was done using the enhanced quadratic mode. Solvent system used for elution was methanol–acetonitrile (90:10).

GC–MS analysis

GC–MS analysis of the isolated compounds was done on GC–MS Varian (Palo Alto, California, USA). Column specification was the following: 30 m × 0.25 mm and 0.25 μm thickness packed with 5% diphenyl–95% dimethyl polysilosane cross bonded liquid phase (DB-5) in a capillary column. The carrier gas was research grade helium (99.999%) sent with a flow rate of 28.6 cm/s (1 mL/min). GC–MS was run in splitless mode with a run time of 47 min and a restabilization time of 5 min. The samples were kept at room temperature before analysis. They were heated in the initial compartment at 80 °C for 5 min. The MS scan mode was set at m/z 40 to 650, and the GC injector temperature was 260 °C. Main oven temperature was 50 °C (1 min), and it was programmed to increase to 260 °C at a ramping rate of 5 °C/min.

HPLC analysis

The results of HPLC analysis of isolated compounds are shown in Table 3. The comparison of the retention times (RTs) of standard vanillin (RT = 3.131 min) and isolated one (RT = 3.121 min) indicated their identity, which was further confirmed by GC–MS.

The HPLC scan of standard stearic acid (RT = 7.417 min) and isolated one (RT = 7.413 min) showed similarity in the retention times. This was reconfirmed by GC–MS data. The saw-like pattern in the UV scan (380 to 570 nm range) is a prominent characteristic of fatty acids [13].

The compound 3-tert-butyl-4-hydroxyanisole and its standard had similar retention times (3.529 and 3.489 min, respectively). The structure was verified by GC–MS. Comparing the HPLC data of standard quercetin (RT = 6.377 min) and the isolated one (RT = 6.352 min) suggested its presence, which was verified by LC–MS. Quercetin, a natural antioxidant, is a flavonoid in nature. It has been previously isolated from the mesocarp of L. siceraria [5, 14].

The HPLC comparison also indicated the presence of rutin, which is a quercetin glycoside. It was obtained at RT of 3.077 min, which matched with the standard [14, 15]. The compound was further established using the MS fragmentation pattern from LC–MS.

Similarly, the HPLC chromatograms of isolated and standard β-sitosterol matched. The λ_max of both the standard and the isolated compound was at 206 nm, which further confirmed their identity. This compound has been reported from the fruit in the past [14].

GC–MS analysis

Four compounds isolated by TLC and HPLC were clearly identified by GC–MS analysis. They were vanillin, 3,4-dimethoxybenzaldehyde methyl monoacetal, octadecanoic acid methyl ester, and 3-tert-butyl-4-hydroxyanisole (Table 3).

LC–MS analysis

The LC–MS has been found especially useful to identify phytochemicals present in small amounts not sufficient for spectroscopic techniques [16]. In the present work, structures of quercetin and rutin were elucidated by LC–MS. The isolated compounds were dissolved in methanol. The solvent system used for the LC–MS analysis was methanol–acetonitrile (90:10. The data are tabulated in Table 4.

Quercetin (LC–MS RT = 5.95 min) showed a peak at m/z = 303.24. This was M + 1 peak. The fragmentation also showed M + 2 peak and another one at m/z = 613.30, which may be due to a dimeric adduct having a sodium atom and water molecule (302 + 302 + 23 − 18).

Rutin was identified as follows. The LC–MS peak at RT = 5.93 min showed a fragmentation pattern with a peak at m/z 303.28, indicating a quercetin unit, the aglycone of rutin. The peak at m/z = 334.00 was due to a condensation product of two rhamnose moieties. The m/z = 465.07 was due to quercetin attached to one rhamnose moiety (302 + 164). The peak at m/z 633.29 was possibly an Na salt of rutin.

NMR Spectral Study

Only four isolated compounds were in sufficient quantity to obtain their proton nuclear magnetic resonance (¹H-NMR) (300 MHz; CDCl₃) spectra, which matched with the reported ones and, hence, confirm their structures. They were stearic acid, vanillin, 3-tert-butyl-4-hydroxyanisole, and β-sterol.

GC–MS analysis of column eluate Frac4

The sub-fraction Frac4 obtained form the column was chosen for GC–MS analysis, as it gave a very prominent fluorescent spot on the TLC. Upon GC–MS screening, a number of compounds were clearly identified (Table 5). Most detected compounds were non-polar in nature or have long non-polar chain. Mostly, they were fatty acids and their esters.

The compounds containing phenolic rings are important for explaining the anti-bacterial effect of the peel, and this lures scientific interest as the plant is believed to be an anti-bacterial agent against mild bacterial infections.