INTRODUCTION
Solanum elaeagnifolium Cavanilles (Solanaceae) is widely distributed in the North of Mexico 1. The plant grows in scrubland, pasture and disturbed areas, accordingly is considered a weed for agricultural crops 1 and is highly toxic to livestock 2. S. elaeagnifolium fruit contains an enzymatic complex employed to artisan manufacture Asadero-cheese 3. The extracts are used in Mexican folk medicine and are known to contain solasodine, β-sitosterol, campesterol, stigmasterol and -5-avenasterol 4. The methanol extract contains glycosylated flavones with cytotoxic effect on MCF-7 and HPG-2 cell lines 5. Solanum has anti-inflammatory 6, anti-hepatotoxic 7, hypotensive 8, cytotoxic 9, antiviral 10 and antifungal 11 properties. Solanum contains steroidal saponins and glycoalkaloids 12 such as solasonine and solamargine 13. These compounds have inhibitory effects on human cancer cell lines such as HT-29, HCT-15 (colon), LNCaP, PC-3 (prostate), T47D, MDA-MB-231 (breast), PLC/PRF/5 and HepG2 (human hepatoma), and JTC-26 (cervical cancer cells) 14. The anticancer potential from Solanum is plausible 4 because species allied produce similar metabolites 15,16. Therefore, compounds isolated from S. elaeagnifolium potentially represent a great alternative to develop new anticancer agents 17. Artemia salina assay represent a reliable, reproducible, accurate, and economical method to selecting bioactive compounds from several sources 18. This experimental approach allows the selection of purified fractions with biological actions from plants (19,20. Therefore, in this study a bioguided fractionation of a methanolic extract from S. elaeagnifolium fruit was carried out to evaluate its cytotoxicity on human tumoral cell lines and anti-tumoral potential on breast tumor explants.
MATERIALS AND METHODS
Extraction S. elaeagnifolium fruit
The fruit S. elaeagnifolium Cavanilles was collected in Saltillo, Coahuila, Mexico during November 2014. The plant material was identified and a voucher specimen (026294) was deposited in the herbarium of the Faculty of Biological Sciences of the Autonomous University of Nuevo Leon 21. Ripe fruit was dried in an oven at 28°C and ground in a STAR Tisamatic-mill. The methanolic extraction was performed at 10% w/v under constant stirring for 2 h at 25°C (Hotplate Fisher Scientific), to obtain the crude methanolic extract of S. elaeagnifolium fruits (CMESeF). The CMESeF was filtered with Whatman #4 paper and clarified with Whatman G/ FA, then was concentrated on a rotary evaporator (Büchi R-120) at 40°C and dried with a lyophilizer (Labconco Freeze 5). Then, the yield Y (%) was calculated by gravimetric analysis (analytical balance Ohaus N1B110 Navigator) 22. The CMESeF was stored at -20°C in a freezer until use.
Bioguided fractionation of S. elaeagnifolium
Separation from CMESeF by liquid-liquid partition
The CMESeF (10%) dissolved in MeOH (w/v) was dispersed on Hexane (1:1) under constant stirring (Hotplate Fisher Scientific) with a flow rate of 6 mL/min. Then, the methanolic (MFSe) and hexane phase (HxFSe) were separated by a liquid-liquid partition. The HxFSe was dried in a rotary evaporator (Büchi R-120) at 30°C and determined the Y (%). The MFSe was dispersed into ethyl acetate (EtOAc) in a ratio 1:1 to obtain a homogeneous solution (MFSe/EtOAc). Later MFSe/EtOAc was separated in two phases by adding distilled H2O in a ratio 2:1 (MFSe/EtOAc:H2O) and stored for 24 h at -20°C, to achieve separation from MFSe and EtOAc to obtain EtOAcFSe. Then, the phases were dried on a rotary evaporator (Buchi R-120) at 30°C, and calculated the Y (%). The fractions obtained were tested in A. salina nauplii to select those with the highest effect and continue the purification process 23. The most active fractions were characterized by qualitative phytochemical tests and fractionated by vacuum liquid chromatography (VLC).
Selection compounds by A. salina assay
This assay was carried out in a 96-well plate, as described by Meyer et al, with some modifications 24. A. salina cysts were hatched in artificial sea water at 37 g/L at 25°C with aeration and constant light source for 24 h. Concentrations of 50, 100, 250, 500 and 1000 μg/mL from CMESeF as well as each fraction recovered in the different purification steps were tested for 24 h at 25°C. A concentration-response curve with K2Cr2O7 at 0, 5, 10, 15, and 20 mg/mL, as well as a negative control (viability control) was performed. LC50 was determined with the percent of mortality M (%) by a linear regression analysis.
The M (%) was determined by the follow equation:
Qualitative phytochemical characterization
The sulphuric acid test was used to identify flavones, chalcones, and quinones. The Shinoda test was used to determine flavonoids while the Baljet test was applied to detect sesquiterpen lactones. The Dragendorff, Wager, and Mayer tests were used to determine alkaloids. The Liebermann-Burchard test was used to identify triterpenes and steroid compounds while Molisch test was utilized to detect carbohydrates. The sodium hydroxide test was used to determine coumarins 25. In this way, both CMESeF and the fractions were characterized on 12-well porcelain plates using a solution at 2000 mg/mL of each sample.
Chromatographic separation from CMESeF
The eluents employed in the separation of CMESeF (2000 μg/mL) by VLC were pre-selected on Merck 60 (1.6 x 5 cm) TLC plates. The eluent was selected with retention factor values (Rf) ≥ 0.1 between fractions 26. Fractions were observed at 254 and 365 nm with ultraviolet light. The eluents used were: EtOAc-MeOH-H2O (100:13.5:10), EtOAc-MeOH (90:10), EtOAc-MeOH (60:20) Hex-CHCl3-AcOH (45:45:10), CHCl3-MeOH-AcOH (47.5:47.5:5), CHCl3-An (4:1) CHCl3-MeOH (9:1), C2H5OC2H5-CHCl3 (1:4), CHCl3-EtOAc (1:1), Hex-AcOEt (3:1), where EtOAc= ethyl acetate, MeOH= methanol, H2O= distilled water, Hex= hexane, CHCl3= chloroform, AcOH= glacial acetic acid, C2H5OC2H5= ethyl ether, An= acetone.
On this way, the MFSe showed a greater effect on A. salina. In consequence was fractionated by VLC. For this, 1 g of MFSe was eluted within a glass column (Pyrex, 3 cm of diameter) prepacked with G silica gel (Fluka) 27. The elution was carried out by connecting the column to a negative pressure source of 20 psi (Felisa F-1500L vacuum pump), using a flow rate of 2 mL/min. The elution gradient used as mobile phase was 50 mL of CHCl3-MeOH (100:0 → 10:90). Ten fractions were obtained: FVLC1, FVLC2, FVLC3, FVLC4, FVLC5, FVLC6, FVLC7, FVLC8, FVLC9, and FVLC10. Fractions were dried and the Y (%) was determined as described above.
Evaluation toxicity on cellular lines and breast tumor explants
Viability assay
Vero, HeLa, and MCF-7 cell lines were donated by the Northeast Biomedical Research Center (CIBNOR-IMSS). Those fractions with higher activity on A. salina were tested on cell lines to assess viability by WST-1 technique. Briefly, 5 x 103 cells/ well from each cell line were cultured in DMEM/ F12 supplemented with penicillin-streptomycin (0.1%), glucose (25 mM) and inactivated fetal bovine serum (FBS, 10%) at 37°C, a humidified atmosphere and CO2 (5%) (NAPCO-6200 CO2 incubator). FVLC7 (10 and 100 μg/mL), two negative controls (10% ethanol v/v-3.8% DMSO v/v, and medium culture) and a positive control (Triton 1X). Then, WST-1 (10 μL), was added to each well and incubated during 2 h. Spectrophotometric reading was performed at 450 nm in a microplate reader using Gen 5 software and cellular morphology was observed by optic microscopy. Results were expressed as percent of the viability V (%).
Evaluation from antitumor effect
The antitumor effect was evaluated using Alamar blue bioassay on breast tumor explants obtained from a patient in remission following ethics codes of the Helsinki Declaration 28. Tumor was cultured in trypticase-yeast-iron (TYI) culture medium. The TYI medium was supplemented with inactivated FBS (10%), vitamins (2.26%), sodium selenite insulin transferrin (1%), D-glucose (1 M), L-glutamine (200 mM) and sodium pyruvate (100 mM). Then, the tumor was cut in 2 explants (Brendel Viton, Viton Inc, Tucson Arizona, USA) using a Krebs-Henseleit buffer and each explant subdivided in explants with a uniform diameter (UDE). The assay was carried out with 2 UDE´s/well in a 6-well plate pre-filled with TYI medium at 37°C, humidified atmosphere, CO2 (5 %) and shaking at 25 rpm. FVLC7 (10 mg/mL) and a negative control (TYI medium) were evaluated. Supernatant from each treatment was placed in a new well-containing DMEM/F12 (200 μL) and Alamar blue (1:10), then, incubation realized for 3-4 h. Fluorescence values were read using a multimode microplate reader (Synergy BioTek HT) at 530 nm excitation/590 nm emission wavelengths. The viability percentage regarding control was calculated using the software AbD Serotec.
Characterization by RP-HPLC-MS
FVLC7 was characterized by Reversed-Phase High Performance Liquid Chromatography coupled to Mass Spectroscopy (RP-HPLC-MS). An electrospray ion detector (ESI-MS) and an ion trap (IT Varian 500-MS Mass Spectrometer, USA) source were used. Samples were analyzed in full scan mode, and all experiments were examined in negative mode [M-H]-1 using a range 50-2000 m/z. Samples were injected in mass spectrometer at a flow rate of 50 mL/min for 7 min. Nitrogen was used as nebulizing gas and helium as damping gas. The parameters utilized for ion source were: electrospray voltage (5.0 kV), capillary voltage (90.0V) and temperature (350°C). Data were analyzed using MS Workstation software V 6.9.
Statistical analysis
All data were reported as a mean value ± standard error (EE). The LC50 data from A. salina obtained with CMESeF (N= 24), MFSe (N= 9), EtOAcFSe (N= 3), HxFSe (N= 9) fractions isolated by VLC (N= 3), positive control (N= 9), and negative control were analyzed by ANOVA and Dunn test. A. salina data was compared against K2Cr2O7, while cell lines data with Triton 1X. Breast tumor explants results were analyzed using Student’s T-test (N= 3) and compared with the negative control. All statistical analysis were per-formed in GraphPad Prism 5© software. Statistical significance was accepted with values of p ≤ 0.05. The assays with A. salina, cell lines, and breast tumor explants performed in triplicate.
RESULTS
Bioguided fractionation of S. elaeagnifolium
Separation from CMESeF by liquid-liquid partition
A yield of 10.90 ± 2.82% of CMESeF was extracted from S. elaeagnifolium fruit while with the liquid-liquid partition the yield was of 2.340, 13.519 and 61.522% of the fractions HxFSe, EtOAcFSe and MFSe, respectively.
Selection of isolated fractions with biological activity (A. salina assay)
The positive control (K2Cr2O7) induced toxic effect in A. salina with high potency (LC50= 13.90 ± 0.29 mg/mL). In contrast, the CMESeF and MFSe had the same effect than the control although with less potency since 40-fold higher concentrations (LC50= 567.0 ± 12.1 mg/mL and 565.1 ± 49.9 mg/ mL, respectively) were used, whereas with EtOAcFSe and HxFSe fractions 72-fold higher concentrations (LC50 > 1000 mg/mL) were required. Similar results were obtained with fractions separated by VLC since LC50 of FVLC2, FVLC4, and FVLC9 were 518 ± 26.0, 840.1 ± 86.6, and 450 ± 0.0 μg/mL respectively, while the LC50 for FVLC3, FVLC5, FVLC6, FVLC8, and FVLC10 was > 1000 μg/mL. FVLC1 exhibited a toxic effect similar to the positive control but with 7-fold higher concentration (LC50= 123.5 ± 0.2 mg/mL). Interestingly, FVLC7 (LC50= 16.0 ± 0.0 mg/mL) had the same toxic effect than the control positive (Figure 1). The negative control had no toxic effect.
Qualitative phytochemical characterization
Sesquiterpene lactones, alkaloids and coumarins were identified in the CMESeF, MFSe and EtOAcFSe but not in HxFSe. In FVLC1 and FVLC9, sesquiterpene lactones and alkaloids but not coumarins were detected while in FVLC2, FVLC8 and FVLC10, alkaloids and coumarins, but not sesquiterpene lactones, were found. In FVLC3 and FVLC6, sesquiterpene lactones, alkaloids, coumarins, and lactones were identified. In FVLC4, sesquiterpene lactones and alkaloids were found. In FVLC5 and FVLC7, sesquiterpene lactones, alkaloids, and coumarins were identified.
Chromatographic separation from CMESeF
Chromatographic profile showed that only four eluents did separate the compounds of CMESeF. With CHCl3:MeOH:AcOH (47.5:47.5:5) three fractions with different Rf´s (Rf1= 0.68, Rf2= 0.80 and Rf3= 0.90) were separated. With EtOAc:MeOH:H2O (100:13.5:10) four fractions with various Rf´s (Rf1= 0.03, Rf2= 0.05, Rf3= 0.13 and Rf4= 0.23) were obtained. With EtOAc:MeOH (60:20) and CHCl3:MeOH eluent (9:1), a fraction (Rf1= 0.08) and three fractions (Rf1= 0.08, Rf2= 0.13 and Rf3= 0.50) were separated. The fractionation by VLC showed higher yield (%) of FVLC5 (36.941%) than the rest of the isolated fractions of MFSe. These results were calculated considering the yield as 100% of the sample used for this fraction (21.652 g) (Table 1).
FVLC7 (100 μg/mL) reduced the viability of all cell lines at 39 ± 1.67, 15.05 ± 0.09 and 66.10 ± 4.44% of Vero, HeLa, and MCF-7 cells. Similar results were observed microscopically since in all cases, FVLC7 decreased cell density (Figure 2).
Characterization by RP-HPLC-MS
Mass spectra obtained from FVLC7 fraction detected traces of quinic acid (1) (ESI m/z 191 [M-H]-1), chlorogenic acid (2) (ESI m/z 353 [M-H]-1), dicaffeoylquinic acid (3) (ESI m/z 515 [M-H]-1). The signal (ESI m/z 395 [M-H]-1) might correspond to an alkaloid (ESI m/z 395 [M-H]-1) (Figure 3).
DISCUSSION
Yield of CMESeF is higher than reported in S. elaeagnifolium leaf under same extraction conditions 29. This difference in Y could be explained by the number of compounds with the same chemical nature found in the fruit or leaves, because they interact through a dipole moment with methanol, which allows the intermolecular forces that keep together are overcome to extract them 30. The eluent with the best profile fractionation was CHCl3:MeOH:AcOH (47.5:47.5:5) 31. However, using MFSe we found a better yield probably due to substances with solvated dipole moments 32. Assays with A. salina revealed that LC50 of CMESeF is similar those reported for crude methanolic extracts from other solanaceae in ranges of 107.2 to ≥ 1000 mg/mL 33,34. The positive control had a LC50= 13.90 ± 0.29 mg/mL comparable with those obtained in other works 35. In order to determine the relationship between the effect presented by CMESeF and the fractions we follow the next: elevated lethality= 0.1-100 mg/mL, moderate lethality= 100-300 mg/mL, low lethality=300-640mg/mL and minimum lethality=≥640mg/mL 37. On this way, CMESeF exhibited a low lethality while EtOAcFSe y HxFSe exhibited a minimum lethality similarly to several researches 36.
These results revealed that MFSe and CMESeF contain the main bioactive compounds probably due to its high polarity. Compounds detected in S. elaeagnifolium such as flavonoids, alkaloids, steroids, saponins, triterpenes and tannins have been reported in Solanum plants 37. Therefore, the chemical composition found in CMESeF could be related to conditions of collection, storage capacity, synthesis (in each organ), geographical region, time season, and phenological age of plant 37. It has been proposed that alkaloids and coumarins are responsible for A. salina toxicity of CMESeF, because there are a close relationship between toxicity and alkaloids in other plants 38. The absence of these compounds by qualitative phytochemical analysis in the HxFSe was due to the polarity. The fractionation by VLC increased the Y (%) of CMESeF when the proportion of methanol was rised during the separation process, whereby a larger amount was obtained with CHCl3:MeOH (60:40). Therefore, the Y (%) each fraction depends on the equilibrium between mobile and stationary phases, as well as their distribution coefficient and relative retention 39.
FVLC5, FVLC8, and FVLC10, had no toxic effect on A. salina40, whereas FVLC7 (100 mg/ mL) had toxic effect rather than the rest of the fractions tested 36. This fraction also decreased the viability of Vero, HeLa, and MCF-7 cells. However, HeLa cell line was more susceptible than Vero and MCF-7 cells. Furthermore, FVLC7 produced morphological changes in the size of HeLa and MCF-7 cells 41. Here, we speculate that the fraction permeates into the cells and trigger molecular interactions associated with the respiratory and metabolic activity in the cell membrane or cytoplasm 42. Interestingly, the Vero cells did not present morphological changes after treatment with FVLC7 which is an advantage because it is a healthy cell line 43. FVLC7 (10 mg/mL) induced a slight toxicity on the breast tumor explants, this effect could be increased with a higher dose 44,45. We believe that this response may be due that the explants have more than one cellular interaction, since a complex net of different components from extracellular matrix that together with tumor microenvironment protects neoplastic cells from the cytotoxic agents 46. A difference on the effect of FVLC7 in the viability of the explants compared with cell lines is that both have different anatomical and chemical composition, and that one or more inducers or enzyme inhibitors could participate in the transformation of Alamar blue 47. The bioactive compounds identified in S. elaeagnifolium are similar to those reported in other Solanum species, in this regard, species such as Solanum tuberosum L. contain quinic, dicafeoylquinic and chlorogenic acids 48. Solanum melongena L. also contain chlorogenic acid 49. Although the chemical structure of the alkaloid found in S. elaeagnifolium is unknown, it may be mediating the toxicity of S. elaeagnifolium since compounds wich contain an alkaloid structure reduce the growth of several cancer cell lines 49. The identification of this alkaloid structure guarantees future investigations.
CONCLUSIONS
In this study we develop methods to separation of bioactive compounds from S. elaeagnifolium. Of this way, the FVLC7 reduced the viability of HeLa and MCF-7 rather than Vero cells indicating a selective response of this fraction. The reduction of the viability of breast tumor explants suggests that FLVC7 could represent a new source of antitumor agents. Chemical analysis of this fraction has revealed that quinic, chlorogenic and dicaffeoylquinic acids may be mediating the toxic effect of FVLC7 in HeLa and MCF-7 cells and in breast explants.