Introduction

Hippophae rhamnoides L commonly called Sea Buckthorn (SBT) is a valuable and unique natural sources gained attention worldwide not only for its medicinal properties but also for its nutritional properties. It is a multi-medicinal tree shrub of the genus Hippocea belongs to the family Elaeagnaceae. Hippophae rhamnoides L. is mostly found in the moderate geographic locations of the north hemisphere, categorized into six species and 12 subspecies in the plant taxonomy [1]. It is a deciduous and nitrogen-fixing plant of cold arid region, native to Asia and Europe. All parts of SBT are considered as a rich source of numerous bioactive constituents with excessive nutritional and medicinal properties [2-4].  Due to these properties, H. rhamnoides L. is domesticated around the globe [5,6]. It is a very important medicinal remedy which is considered to be a rich source of a large number of phytochemicals, nutrients and bio-active constituents [7,8]. The bio-active compounds include vitamins, carotenoids, phytosterols, poly unsaturated fatty acids, organic acids, mineral components and essential amino acids [7-10]. SBT has been used in Europe and Asia for pharmaceuticals, foods and therapeutic purpose for centuries. SBT leaves, berries and seeds are well known for their medicinal properties [11-14]. The known medicinal properties of SBT include immune-modulatory, anti-stress, radio protective, antioxidant, tissue regeneration, anti-atherogenic, hepato-protective property and improving functions of blood circulation and digestive system [4,15-26]. It is interesting to note that the aqueous extract of SBT has recently been reported to possess hypoglycemic activity [27]. SBT contains a lot of different bioactive constituents with multiple properties which have ability to prevent postprandial hyperglycemia [28]. Flavonoids from fruits and seeds of SBT can cause hypolipidemia and hypoglycemia [29]. The leaves of methanolic extracts of SBT contain compounds showing alpha-glucosidase inhibition activity which may be used in diabetes because it has the ability to control inhibition activity [30].

Methodology

General instrumentation

UV spectra were obtained using a Hitachi-U-3200 spectrophotometer. IR spectra were recorded on a Jasco A-302 spectrophotometer. 1D- and 2D NMR spectra were obtained on AM-500, Bruker spectrometers in CD₃OD and CDCl₃, using tetramethylsilane (SiMe4) as internal standard. Mass spectra (EI-MS and HR-EIMS) were analyzed on a Mass AB SCIEX spectrometer QSTAR xl. Column chromatography was carried out by using silica gel 60 (Merck, 70–230 mesh).TLC was carried out on silica gel 60 PF254 (Merck), with detection by UV at 254 nm and 366 nm.

Plant material

The Sea buckthorn berries (8.5kg) were collected randomly in August-September from Gilgit, north of Pakistan in September 2015. The plant was identified by a botanist from Department of Botany, University of Karachi. A voucher specimen (No.G.H.No. 84346) was been deposited in the Herbarium, Department of Botany, University of Karachi.

Extraction and isolation

The berries were dried at room temperature and soaked in methanol for 72h (3times). The extract solution was filtered and concentrated under vacuum. The extract was suspended in water (0.5L) and partitioned with ethyl acetate to give the ethyl acetate-soluble part (111.0g). The ethyl acetate fraction was then fractionated with 4% Na₂CO₃ and 30% HCl to obtain acidic and basic fractions. The basic fraction was dried over Na₂SO₄ washed with water and evaporated the solvents under reduced pressure and obtained a neutral fraction. Both the acidic and neutral fractions were partioned with hexane to get hexane insoluble and soluble fractions.

The ethyl acetate acidic hexane insoluble part (10g) was subjected to a silica gel column chromatography, using a gradient solvent system of Pet. ether EtOAc CH₂Cl₂ MeOH with increasing polarity to give 20 fractions which were combined on the basis of TLC and obtained eight sub-fractions (HRF 1-8). The fraction 2 (1.02g) was fractionated into four sub-fractions through normal phase CC. Compound 1 was purified from the fraction HRF-2D using hexaneethyl acetate (55 vv) as solvent system. Fraction 4 (1.48g) was further subjected to column chromatography using Pet. etherEtOAc as solvent system (8.51.5vv) and obtained compound 2.Compound 3 was purified from fraction 4 using CH₂Cl₂ MeOH (91 vv) as solvent system through normal phase CC (Figure 1).

Figure 1: Structure of compounds 1, 2 and 3.

Sulforhodamine B assay

The growth inhibitory activities of compounds 1, 2, and 3 were evaluated against human non-small cell lung (NCI-H460) and breast (MCF-7) cancer cell lines by using sulforhodamine-B assay for the determination of IC₅₀ values. The concentration of the test agents that inhibits 50% of the cell growth was referred as IC₅₀. For this, the cells (10000cells100µL) from respective cell lines were incubated in 96 well plates for 24h at 37ºC in 5% humidified CO₂ incubator. The stock solutions of compounds 1, 2, and 3 (20 mM) in DMSO, and doxorubicin (1mM) in distilled water were prepared. A range of dilutions for 1 (1, 10, 25, 50 and100µM), 2 and 3 (1, 2.5, 5, 7.5 and 10 µM) were added (100µL) in respective wells. The pre-determined IC50 value of doxorubicin (5µM) was used as positive control (data for range of dilutions are not shown). After completion of 48h, ice cold TCA (50µL, 50%) was added and left at room temperature for 30min. The TCA was washed out with distilled water. After drying, SRB solution (100µL, 0.4% wtvol in 1% acetic acid) was added to each well and unbound stain was washed out after 30min with1% acetic acid. Tris-base (100 µL, 10 mM, pH 10.2) was added to each well to solubilize protein bound stain. The absorbance for test agents along with appropriate blanks (both test agent and control) was recorded at 545nm using a microplate reader (Synergy). This absorbance was used to calculate respective percent growth inhibitions of the test agents. The dose response curve graph was plotted between growth inhibitions (y-axis) vs concentrations of the respective test agents (x-axis) to determine IC₅₀ value.

DNA docking studies

To perform the DOCK calculations that predict the best orientations of the ligand in the binding site of the receptor we require preparing the receptor and ligand as inputs. UCSF Chimera is a visualizing tool used here to prepare the receptor and ligand with its Dock Prep tool option [31]. The structure of DNA PDB ID 1d29 was selected as receptor and downloaded from Protein Data Bank (PDB) [32]. The three compounds are taken as ligands to check if they are making any interactions with the DNA moiety. Hydrogen atoms and partial charges were added to both ligand and receptor. AM1-BCC charges were added to receptor, gasteiger charges were added to ligand and files were saved in mol2 format as program DOCK read file in Mol2 format. Molecular surface of receptor was prepared by DMS program. The DOCK accessory program sphgen was used to generate the spheres with a probe radius of 1.4 angstrom. It generated 12 clusters and the cluster 1 with maximum number of spheres i.e. 38 was selected for construction of box with the help of show box DOCK accessory program. The maximum 38 spheres in cluster 1 were retained for docking (Figure 2).

Figure 2: Surface diagram of DNA (PDB ID: 1d29) representing the cluster 1 with maximum numbers of violet colored spheres.

 Following the construction of box, the DOCK accessory program grid was used to pre-compute the energy interaction points with grid spacing of 0.3 angstrom resolution within the box [33]. The number of grid points on x, y, and z axis was 109, 85, and 108 Å, respectively with box dimension 32.375, 25.076, 32.008Å and center on 12.814, 23.837, and 4.833 Å (Figure 3). Once the grid box was set we performed the docking by treating the ligand as flexible. All the parameters were set as default value given in standard protocols file. The maximum number of orientation was set as 1000 and maximum number of conformations was kept 10 for each compound.
Electrophoretic mobility of plasmid pBR322 DNA
Electrophoretic mobility of pBR322 plasmid DNA was performed to assess the direct interaction of compounds with DNA. Briefly, agarose gel (1%) was prepared in 1x TAE buffer (70mL) by heating for 1min followed by its casting in gel tray at room conditions. pBR322 DNA (500ng, 5µLwell) was mixed either with PBS (control) or with compounds 1, 2, 3 (2.5µM), and doxorubicin (250 nM) and placed at 37ºC for 30min. Loading buffer (3µL, bromophenol blue, xylene cyanol and glycerol in a ratio of 11120) was applied to the above reaction mixture. This sample was loaded in the wells of the agarose gel and electrophoresis was performed in TAE buffer at 70 V for 1.5h. The gel was dipped in ethidium bromide solution (5μgµL) for 20min and washed with tap water. The mobility pattern of the circular DNA was observed under UV light and photographed.

Figure 3: Grid box setting.

Statistical analysis
Duncan’s multiple range test using SPSS 17 program. The data was analyzed by using one-way ANOVA with p<0.05 level was considered as significant followed by.

Results and Discussion

Compound 1 was purified as white powder having the molecular formula C₁₁H₈O₂ ([M⁺] mz 172.05). The IR spectrum showed absorption peaks at 3486cm^-1.The 1H-NMR shows four doublets of the aromatic proton at δ 8.32 (d, J = 8.5 Hz, 1H), δ 7.96 (d, J = 7.9 Hz, 1H), δ 7.78 (d, J = 8.5 Hz, 1H), δ 7.59 (m, 1H), δ 7.41 (m, 1H) and δ 7.12 (d, J = 9 Hz, 1H). Two singlets at δ 10.792 and 13.14 were appeared for aldehyde proton and hydroxyl proton resp.Compound 2 was isolated as white amorphous powder. The molecular ion [M⁺] peak was appeared at mz 456.4 in EI-MS spectrum matching to C₃₀H₄₈O₃. Other characteristic peaks were appeared at mz 248, 203 and 133. In ultra-violet spectrum, an absorption band was appeared at 213 nm. The infra-red spectrum of compound 2 shows absorption peak in the area of 2940.9cm^-1 for symmetric vibrations of CH₂cm^-1 group and at 1696.1cm^-1 appears a characteristic band of carbonyl group (C=O). A broad peak was also appeared in the area of 3446.7 cm^-1 for hydroxyl group. The ¹H-NMR spectrum of compound 2 revealed seven singlets of tertiary methyl protons at δ 0.66, 0.67, 0.79, 0.79, 0.81, 0.86 and 1.03. A doublet of doublet of C-18 proton at δ 2.71 and a triplet of one vinyl proton of C-12 at δ 3.09 indicates an olea-12-ene skeleton. One broad singlet peak at δ 4.8 showed the signal of O-H group. The above mentioned spectral data of compound 2 were in complete agreement with those reported in literature for oleanolic acid [34].
Compound 3 was obtained as white amorphous solid and revealed to have the molecular formula C₃₆H₅₂O₅ ([M]⁺ at mz 564.3814, calculated 564.3815) by HREI-MS. The fragment ion peaks at mz 472.3 [M-C₆H₅O]⁺ showed the loss of phenol moiety. Two fragment ion peaks at mz 454.4 [M-H₂O]⁺ and at mz 437.3 [M-2H₂O]⁺ indicate that compound 3 contained two hydroxyl groups.The UV spectrum showed absorption band at 230 and 298 nm. In infra-red spectrum, a very intensive absorption peak in the area of 2932 cm^-1 for symmetric vibrations of CH₂ group and at 1699 cm^-1 appears a characteristic band of carbonyl group (C=O). A broad peak was also appeared in the area of 3289 cm^-1 for hydroxyl group. The ¹H NMR spectra of compound 3 showed the presence of five methyl signals at δ 0.75, 0.99, 1.05, 1.09, and 1.17 two methyl doublets that appeared at δ 0.92 and 1.01. The ¹H NMR spectra of compound 3 also showed two oxymethine protons resonating at δ 3.32 and 3.37 and a olefinic proton at δ 5.21. The presence of five methyl singlets and two methyl doublets suggested that compound 3 belongs to ursane type triterpenoid having two secondary hydroxyl groups and a tri substituted double bond between C-12C-13. In aromatic range, two doublet of doublet appeared at δ 6.97 and 7.48. The spectrum also displayed a multiplet at δ 7.22 and a doublet at δ 7.34.  The ¹³C-NMR values for all the protons and carbons were assigned on the basis of HMQC and HMBC correlations and were given in Table 1. The appearance of an ester carbonyl group resonating at δ 180.1 in the ¹³C NMR spectral data of 3 suggested the presence of an ester functional group and its location was identified at C-28 by the key COSY and HMBC correlations as shown in Figure 4. 

Figure 4: HMBC correlation of compound 3.

Table 1: NMR spectral data of compound 3 (1H-NMR 500 Hz, 13C 125 MHz).

A search in literature found that the spectral characteristics of 3 were consistent to the reported literature values of Corosolic acid except the aromatic spectral data that confirmed by the key COSY and HMBC correlations as shown in Figure 3. The correlation between protons and carbon of compound 3 are summarized in table 1.
Figures 5 and 6 depicts compound 1 was not effective against non-small cell lung cancer cell line (NCI-H460) while it displayed growth inhibition against breast cancer cells (MCF-7) with IC₅₀ value of ~43 µM. In addition, compound 2 and 3 exhibited growth inhibition against both non-small cell lung and breast cancer cells. In lung cancer cell line (NCI-H460), HI-6 exhibited IC₅₀ value of ~2.8 µM which was ~2x more potent than that of HI-7. However, in case of breast cancer cells (MCF-7), both were equipotent with IC50 value of ~3 µM (Figure 7).

Figure 5: Growth inhibitory effects of HI-2, HI-6, and HI-7 on non-small cell lung cancer (NCI-H460) cell line.

Figure 6: Growth inhibitory effects of comp. 1, 2 and 3against breast cancer (MCF-7) cell line.

Figure 7: Growth inhibition in the presence of test agents against human non-small cell lung (NCI-H460) and breast (MCF-7) cancer cell lines. Non-small cell lung cancer cells (NCI-H460)

After successful docking runs, the best docked conformation for each compound was viewed in UCSF chimera by using View Dock tool option. All three compounds were docked in the minor groove of DNA model (Figure 8). 

Figure 8: All three compounds docked in the minor groove of DNA.

Figure 9: The docking of doxorubicin in the minor groove of DNA.

Table 2:Given are the grid scores of the top ranked conformation of doxorubicin and all three compounds. The details of hydrogen bonding mediated by doxorubicin and DNA.

Structure analysis of each compound was done by Structural analysis tool. We did not find any hydrogen bond interactions between DNA and our top ranked conformations of three compounds. They formed weak interactions such as van dar waal and hydrophobic interactions with the DNA. We also docked the Doxorubicin by following the same protocols. It produces its anticancer activity through intercalation mode. The doxorubicin was also docked in the minor groove but at the different position (Figure 9). The docking of doxorubicin in minor grooves suggests that minor groove binding is the dominant pre-intercalation step (Lei, 2012 #17). The rigid scores of all compounds and doxorubicin are given in Table 2. The high negative value of doxorubicin indicates that it is a strong DNA binder than our compounds. The docking outcomes also correlate with the gel electrophoresis run where doxorubicin is showing high affinity as compared to our three compounds. 

The energy score of compound 3 has a relatively high negative value than compounds 1 and 2. In case of doxorubicin while performing the structure analysis with the help of UCSF chimera, three hydrogen bonds were found. The purine bases guanine (DG14.B and DG16.B) of strand B are involved in mediating the hydrogen bonding shown in Figure 10. The details of the hydrogen bond mediators, donors, acceptor and distances of doxorubicin and DNA are given in table 2.

Figure 10: The three hydrogen bond interactions formed between doxorubicin and DNA.

Figure 11: Effect of HI-2, HI-6, HI-7 and doxorubicin on plasmid pBR322 DNA. Control (Lane-1), 2.5 μM of compound 1 (Lane-2), compound 2 (Lane-3), compound 3 (Lane-4), and doxorubicin (250 nM, Lane-5).

Figure 11 depicts that in control (lane-1), two bands of similar shape were observed and no change was observed in the presence of 1, 2, 3 at 2.5μM (lanes -2, -3 and -4, respectively). However, at 250nM, doxorubicin (lane-5) exhibited smear formation suggesting its interaction and damaging effect on DNA. From the DNA docking and electrophoretic mobility experiments, it is noted that anticancer activity of these three compounds is not related to DNA intercalation. However, weak interactions of these compounds in the minor groove of DNA suggest that these may modify the DNA function at the level of gene expression [35]. Therefore, further studies at transcriptional level should be performed to determine the mechanism of their anticancer activity, whereas the docking result of doxorubicin suggests clearly that it is a strong DNA binder rendering it as an anticancer agent.

Conclusions

It is concluded that the new compound 3 was isolated from the sea buck thorn. It displayed potent anticancer activity against lung cancer than that of known compounds 1 and 2 with no structural DNA damage. Their weak interactions with minor groove of DNA suggest that they may modify gene expression to induce anticancer activity. Lahore, Pakistan. The plant materials leaf, stems and seeds were properly cleaned and kept in room temperature.

Keywords

Anti-cancer, Corosolic acid, DNA, Hippophae rhamnoides, Pentacyclic triterpene