Advanced
Phytochemical Analysis of the Phenolic Fat-Suppressing Substances in the Leaves of Lactuca raddeana in 3T3-L1 Adipocytes
Phytochemical Analysis of the Phenolic Fat-Suppressing Substances in the Leaves of Lactuca raddeana in 3T3-L1 Adipocytes
Natural Product Sciences. 2015. Mar, 21(1): 42-48
Copyright © 2015, The Korean Society of Pharmacognosy
  • Received : October 11, 2014
  • Accepted : November 24, 2014
  • Published : March 31, 2015
Download
PDF
e-PUB
PubReader
PPT
Export by style
Article
Author
Metrics
Cited by
TagCloud
About the Authors
Agung Nugroho
Department of Agro-industrial Technology, Faculty of Agriculture, Lambung Mangkurat University, Indonesia 70712, Indonesia
Jae Sue Choi
Department of Food and Nutrition, Pukyong University, Seoul 608-737, Korea
Hyo-Jin An
College of Oriental Medicine, Sangji University, Wonju 220-702, Korea
Hee-Juhn Park
Department of Pharmaceutical Engineering, Sangji University, Wonju 220-702, Korea
hjpark@sangji.ac.kr

Abstract
Lactuca raddeana (Compositae) is used to treat obesity and complications due to diabetes. The five phenolic compounds including chlorogenic acid, chicoric acid, luteolin 7- O -glucoside, luteolin 7- O -glucuronide, luteolin were qualitatively identified by LC-ESI-MS analysis. The contents were quantitatively determined by HPLC, under the condition of a Capcell Pak C18 column (5 μm, 250 mm × 4.6 mm i.d.) and a gradient elution of 0.05% trifluoroacetic acid (TFA) and 0.05% TFA in MeOH-H 2 O (60 : 40). The contents of chicoric acid (100.99 mg/g extract) and luteolin 7- O -glucoside (101. 69 mg/g extract) were high, while those of other three phenolic substances were very low. The 3T3-L1 adipocyte cells treated with chicoric acid and luteolin 7- O -glucuronide significantly suppressed the accumulation of fat, suggesting they are effective against obesity. Since high level of peroxynitrite (ONOO) causes cardiovascular disease in obese patients, its scavenging activity was also studied.
Keywords
Introduction
Lactuca raddeana (Compositae) is used to treat complications due to diabetes and obesity in addition to fever, depression, insomnia and cardiovascular diseases. Recently, it is being revealed that several polyphenols present in fruits, vegetables or cereals exhibit anti-obesity effects. 1 Obesity increases the risk for insulin tolerance and cardiovascular disease. 2 Continued differentiation and fat accumulation of adipose cells are closely associated with obesity. 3
It has been reported that many phenolic antioxidants attenuate obesity through apoptosis-inducing action in 3T3-L1 preadipocytes. 4 - 7 Studies have also shown a link between obesity and the the accumulation of reactive oxygen species (ROS) in adipose cells. 8 Pires et al . 9 reported that treatment of a peroxynitrite scavenger, Mn(III)tetrakis(4-benoic acid)porphyrin, reduced a weight gain and fat accumulation in adipose cells of mice on a high-fat diet. In addition, ONOO generation represents a crucial pathogenic mechanism in conditions such as stroke, myocardial infarction, chronic heart failure, diabetes, circulatory shock, chronic inflammatory diseases, cancer, and neurodegenerative disorders. 10
Qualitative and quantitative analysis on L. raddeana were pursued based on results from our preliminary study that shown an inhibitory effect in 3T3-L1 adipocytes. Through a UPLC-ESI-MS/MS analysis, phytochemical phenolic constituents (chlorogenic acid, chicoric acid, luteolin 7- O -glucoside, luteolin 7- O -glucuronide, and luteolin) were identified. Using the five compounds, quantitative analysis was also performed on the extracts and fractions by an HPLC method. In addition, the inhibitory effect on 3T3-L1 adipocyte cells and ONOO scavenging effect was investigated.
Experimental
Instruments and reagents − Agilent 1200 Technologies (Agilent Technology, USA) was HPLC used for qualitative analysis by UPLC/Q-TOF-ESI-MS/MS and Synapt G2 (Waters, USA) was a MS detector used as a mode of QTOF MS. The HPLC system used for quantitative analysis consisted of two Prostar 210 pumps, Prostar 325 UV-Vis detector, and a Shiseido Capcell PAK C18 column (5 μm, 4.6 mm × 250 mm, Japan) equipped with its MetaTherm controller. Data were processed using Varian Star Workstation. HPLC solvents (HPLC grade of H 2 O, MeOH, CH 3 CN) were purchased from J.T.Baker (Phillisburg, NJ, USA). Three standard compounds used for both analysis and bioassay were chlorogenic acid (Lot# 104K0722, ≥ 95% by titration), chicoric acid (Lot# 060M1184V, ≥ 95% by HPLC), and luteolin (Lot# 11K4085, ≥ 95% by HPLC) purchased from Sigma-Aldrich (St. Louis, MO, USA), and other two standards, luteolin 7- O -glucuronide (> 98% purity) and luteolin 7- O -glucoside (> 98% purity), are the compounds preserved in our laboratory.
Plant material Lactuca raddeana (Compositae) was collected during August in 2012, on a mountain area in Wonju, Korea. This plant was identified by Prof. Sang-Cheol Lim (Department of Horticulture and Landscape, Sangji University). A voucher specimen (natchem#-49) was deposited in the Laboratory of Natural Products Chemistry, Department of Pharmaceutical Engineering, Sangji University. The leaves of collected plants were dried in a shadow area, crushed, and prepared for extraction.
Extraction and fractionation − The leaves of L. raddeana (100 g) were extracted with 3 L of 70% MeOH three times under reflux. The extracted solution was filtered, evaporated under reduced pressure, and freezedried to give a 70% MeOH extract (aq. MeOH extract, 21.4 g). The extract was partitioned between aqueous phase and CHCl 3 phase, and then the latter phase was further concentrated in vacuo to give a CHCl 3 fraction (7.42 g). In the same way, the aqueous phase was successively fractionated with EtOAc and BuOH, respectively, to give an EtOAc fraction (1.68 g) and BuOH fraction (3.10 g). The aqueous extract was prepared as the same way with aq. MeOH extract: 30.6 g aqueous extract was obtained from 100 g of plant material.
Conditions of UPLC and MS − The conditions of UPLC and MS on qualitative identification are described. A column used in this experiment was Acquity@BEH C18 (1.7 μm, 2.1 × 50 mm). The two solvents, 0.05% trifluoroacetic acid (TFA) and 0.05% TFA MeOH-CH 3 CN (60 : 40), were used as the A- and B solvents, respectively. Gradient elution was programmed at the flow rate of 0.3 ml/min ad follows: 0 min (100% A), 4 min (85% A), 20 min (60% A), 30 min (0% A), 35 min (100 A). LC system was coupled to a Q-TOF MS equipped with ESI source. In a TOF MS analysis, mass spectra were measured in a positive ion mode. The condition for ESI source was as follows: ESI capillary (3.0 kV), sampling cone (40 V), temp. source (120°), desolvation (300°), cone gas (100 L/h), desolvation gas (600 L/h). A UV wavelength was fixed at 254 nm for detection.
Preparation of sample solution − A freeze-dried aq. MeOH extract were used for preparation of sample solutions. Five concentrations (1,000, 1,500, 2,000, 2,500, and 3,000 μg/ml) were prepared according to ICH (International Conference on Harmonization) guidelines and used for the repeatability test. Of the five concentrations, the 1,500 μg/ml solution was used for intra-day- and inter-day variability tests. Sample solutions were filtered through a 0.50-μm syringe filter prior to injection to HPLC.
Condition for HPLC quantification − To prepare standard solutions, five standard compounds were completely dissolved in MeOH by vortexing. These solutions were filtered through a 0.50-μm syringe filter prior to injection onto an HPLC system. The two solvents, 0.05% TFA-H 2 O solution and 0.05% TFA-MeOH/CH 3 CN (60 : 40), were the A and B solvents, respectively, for the mobile phase. Gradient elution was programmed as follows: (A)/(B) = 85/15 (0 min) → (A)/(B) = 35/65 (35 min) → (A)/(B) = 0/100 (47 min; hold for 6 min to wash the column) → (A)/(B) = 85/15 (54 min; hold for 6 min to equilibrate the column condition). Chromatograms were recorded during a period of 0 - 37 min. The column was maintained at 40 ℃, and a wavelength 254 nm was chosen because it was more sensitive than 280 and 360 nm for the simultaneous detection of caffeoylquinic acid, caffeoyltartric acid and luteolin glycosides.
Validation on the HPLC method − Validation experiments were performed in terms of linearity, sensitivity, precision and accuracy. Serial-dilution was conducted to prepare 3.13 - 100.0 μg/ml concentrations of chlorogenic acid, chicoric acid, and luteolin 7-glucuronide, and 1.56 - 50.0 μg/ml of luteolin 7-glucoside and luteolin. Peak areas ( y ) were plotted against the concentrations ( x axis), and the linearity of regression equations was assessed by R 2 (correlation coefficient) values. Sensitivity was evaluated by LOD (limit-of-detection) and LOQ (limit-of-quantification) values which were determined by the signal-tonoise (S/N) method. Intra-day and inter-day variability tests were performed to evaluate precision of the method. Intra-day variability was completed within 24 h, while inter-day variability was conducted on four different days by injecting the same solutions five times a day. Relative standard deviations (RSDs) were obtained by injecting the same solution five times, and were considered a measure of precision. Accuracy was evaluated from the mean recovery rates (%) of standards from the spiked extract versus non-spiked solution extract sample.
Cell Culture and adipocyte differentiation − 3T3-L1 mouse embryo fibroblasts were obtained from the American Type Culture Collection (Rockville, MD, USA). Cells were grown in DMEM plus 10% calf serum and plated for final differentiation in DMEM plus 10% FBS with 100 units/ml of penicillin-streptomycin solution at 37 ℃, in 5% CO 2 , at 95% humidity until confluence. Two days after confluence (Day 0), the cells were stimulated to differentiate with differentiation inducers (1 μM dexamethasone, 500 μM 3-isobutyl-1-methylxanthine, and 1 μg/mL insulin, MDI) that were added to DMEM containing 10% FBS for two days (Day 2). Preadipocytes were then cultured in DMEM, 10% FBS supplemented with 1 μg/mL insulin for another two days (Day 4), followed by culturing with 10% FBS/DMEM medium for additional two days (Day 6), at which time more than 90% of cells were mature adipocytes with accumulated fat droplets. On Day 2, the samples of the aq. MeOH extract, fractions (CHCl 3 -, EtOAc- and BuOH fractions) and chicoric acid and luteolin 7- O -glucoside were prepared in a differentiation medium at three concentrations (50 μg/ml, 100 μg/ml, and 200 μg/ml).
Cell cytotoxicity assay − Cell viability was measured with a CellTiter 96 Aqueous One Solution Cell Proliferation Assay Kit (Promega Corporation, Madison, USA) according to the manufacture’s instruction. Briefly, the cells (5 × 103 per 96 well) were incubated at 37 ℃ in 5% CO 2 and 95% air with samples. After 48 h for 3T3-L1 preadipocytes, 20 μL of MTS [3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2(4-sulfophenyl)-2H-tetrazolium, inner salt] solution was added to each well and incubated for 4 h. Absorbance at 490 nm was measured using a VERSA maxmicroplate reader (Molecular Devices, Sunnyvale, CA, USA) to determine the formazan concentration, which is proportional to the number of live cells.
Oil red O staining − Intracellular fat accumulation was measured using Oil Red O. The Oil Red O working solution was prepared as described by Ramirez-Zacarias et al . 11 The 3T3-L1 cells were washed twice with phosphate-buffered saline (PBS) and were then fixed in 10% formaldehyde in PBS for 1 h. After washing with 60% isopropanol, the cells were stained with Oil Red O solution for 30 min at room temperature. The cells were washed with water four times to remove the unbound dye. The stained cells were observed with an Olympus IX71 Research Inverted Phase microscope (Olympus Co., Tokyo, Japan). Following the microscopic observation, 100% isopropanol was added to extract the excess staining dye from the cells.
Peroxynitrite-scavenging activity − An assay method described by Kooy et al . 12 was used to measure the peroxynitrite-scavenging activity of the aq. extract from the leaves of L. raddeana . The principle of this method is to monitor the intensity of highly fluorescent rhodamine formed from non-fluorescent DHR 123 under the presence of ONOO . Rhodamine buffer (pH 7.4) was consisted of 50 mM sodium phosphate dibasic, 50 mM sodium phosphate monobasic, 90mM sodium chloride, 5mM potassium chloride, and 100 μM DTPA. The final concentration of DHR 123 was 5 μM. The samples were dissolved in 10% DMSO (concentration: 5 μg/mL). The final fluorescent intensity was measured with or without the treatment of 10 μM ONOO in 0.3N NaOH. The fluorescence intensity of oxidized DHR 123 was measured at the excitation and emission of 480 nm and 530 nm using microplate fluorescence reader FL 500 (Bio-Tek Instruments Inc., Winooski, VT, USA). Through the detection of the oxidation of DHR-123, peroxynitrite-scavenging activity was calculated by subtracting the background fluorescence from the final fluorescent intensity. L-penicillamine with a great peroxynitrite-scavenging activity was used as a positive control.
Results and Discussion
Qualitative analysis by a UPLC-ESI-MS method − Five compounds of L. raddeana were identified on the basis of pseudomolecular and fragment ions ( m/z ) on the mass spectra measured by a UPLC-ESI-MS experiment. The chromatogram and mass spectral data of pseudomolecular and fragment ions measured by this experiment were shown in Fig. 1 and Table 1 , respectively. Mass spectra of compounds 1 and 2 exhibited a caffeoyl moiety at m/z 163.03, and those of 3 , 4 and 5 displayed a luteolin moiety at m/z 287.05. The five compounds were identified as chlorogenic acid ( 1 ), chicoric acid ( 2 ), luteolin 7-glucoside ( 3 ), luteolin 7-glucuronide ( 4 ), luteolin ( 5 ). The structures were shown in Fig. 2 . The identification of compounds 1 - 5 were confirmed by comparisons of retention times ( t R s) on the chromatogram and each mass spectrum with standard compounds.
PPT Slide
Lager Image
UPLC chromatogram of the MeOH extract of L. raddeana.
UPLC-ESI-MS data of compounds1 - 5identified fromL. raddeana
PPT Slide
Lager Image
UPLC-ESI-MS data of compounds 1 - 5 identified from L. raddeana
PPT Slide
Lager Image
Structure of compounds 1 - 5 identified from L. raddeana.
Optimization and validation of HPLC method − Four parameters, mobile phase composition, gradient elution, UV wavelength and column temperature, were optimized to establish a more reliable method. The A and B solvents, 0.05%-trifluoroacetic acid (TFA)/H 2 O and 0.05%-TFA in MeOH-CH 3 CN (60 : 40), respectively, were chosen because they were more selective, environmentally-friendly and economic. The A and B solvents were adjusted with 0.05% TFA concentration so that the acidified solution could inhibit release of protons from phenolic substances. Solvents showed more selective peaks than acid-free solvent. Gradient elution was employed to cover a wide range of peaks within a shorter time. This method produced a more selective and a more repetitive chroma- togram at a fixed temperature of 40 ℃ than at room temperature.
The optimized HPLC method was validated in terms of linearity, sensitivity, precision and accuracy. The R 2 , LOD and LOQ values are shown in Table 2 . Linearity was verified from the R 2 values > 0.9996. Sensitivity was validated from the LOD- (< 0.89 μg/ml) and LOQ values (< 2.17 μg/ml). Experimental results obtained from the intra-day- and inter-day variability and from the recovery test were shown in Table 3 . The RSD values in the intraday variability were shown from 0.48 - 2.42%, suggesting that it is sufficiently precise. And the RSD values of the inter-day variability were between 0.69 - 4.16%, suggesting that this method has sufficient stability. Accuracy was also established with recovery rates of 96.01 - 102.75%.
Linearities and limits of detection and quantification (LOD and LOQ) of the analytes
PPT Slide
Lager Image
ay, peak area at 254 nm; x, concentration of the standard (μg/mL); bR2, correlation coefficient for 6 data points in the calibration curves (n = 4); cLOD, limit of detection (S/N = 3); dLOQ, limit of quantification (S/N = 10).
Precision and recovery data of the analytes
PPT Slide
Lager Image
Relative standard deviation (RSD) values were calculated for both retention time (tR) and peak area of three experiments. Recovery tests were performed in the 70% MeOH extract spiked with each standard compound.
PPT Slide
Lager Image
HPLC chromatograms of extract and fractions of L. raddeana.
Contents of phenolic substances − Contents of five phenolic substances in the extracts and fractions were shown in Table 4 . In plants, caffeic acid esters mainly occurred in the form of caffeoylquinic acids or caffeoytartartic acid, where their typical compounds are chlorogenic acid and caffeoyltartaric acid, respectively. The content of chicoric acid (100.99 mg/g) was considerably higher than that of chlorogenic acid (6.69mg/g), suggesting that the former compound is the main caffeoylquinic acid in L. raddeana . Luteolin 7- O -glucuronide was a main substance among the two luteolin glycosides and their aglycone (luteolin). These five phenolic substances were quantitatively higher in the EtOAc- and BuOH fractions than CHCl 3 fraction. Further, the analysis was also performed to compare aq. MeOH and H 2 O which solvent yielded the highest extracts. The aq. MeOH extract yielded higher contents than the aq. extract.
Content of analytes in the lyophilized extract and fractions ofL. raddeana(mg/g)
PPT Slide
Lager Image
Values in the parentheses are content of analytes in the dried plant materials (mg/g).
Inhibition on fat accumulation in 3T3-L1 adipocytes and peroxynitrite formation − To determine the cytotoxicity of test samples, 3T3-L1 cells were treated with various concentrations (1.56 - 200 μg/mL), and the cell viability was measured by the MTS assay. As shown in Fig. 4 , treatment with 1.56 - 200 μg/mL of samples did not have significant cytotoxic effects on 3T3-L1 cells. Further, we measured the effect on adipocyte differentiation. We used a differentiation mixture (MDI) to induce the differentiation of 3T3-L1 cells. The 3T3-L1 cells were treated with samples (50, 100, and 200 μg/mL; aq. MeOH extract, fractions, chicoric acid and luteolin 7- O -glucuronide) during differentiation. Six days later, cells were stained with Oil Red O. As shown in Fig. 5 , the aq. MeOH extract and EtOAc and BuOH fractions (200 μg/ml concentrations) effectively suppressed fat accumulation in 3T3-L1 cells compared to the control. The two substances (chicoric acid and luteolin 7- O -glucuronide) significantly reduced fat accumulation in cells at 200 μg/ml, which suggests that the effect of the aq. MeOH extract is mainly attributed to the two substances. These effects were observed in dose-dependent manners (data not shown). The pharmacological actions of chicoric acid such as inducing apoptosis, 13 protecting against stress, 14 and antiviral 15 have been reported. Luteolin 7- O -glucuronide has anti-gastritis, 16 antidepressant, 17 and antimutagenic effects. 18 As shown in Table 5 and 6 , ONOO -scavenging activities of the extract of L. raddeana , chicoric acid, and luteolin 7- O -glucuronide were potent. The IC 50 values of L. raddeana , chicoric acid, and luteolin 7- O -glucuronide were 1.15 μg/ml, 0.76 μM, and 3.13 μM, respectively.
PPT Slide
Lager Image
Effect of aq. MeOH extract of L. raddeana, its fractions and its substances (luteolin 7-O-glucuronide and chicoric acid) on cell viability in 3T3-L1 cells. 3T3-L1 cells were treated with samples at various concentrations (1.56 - 200 μg/mL) for 48 h. Cell viability was determined by the MTS assay. Postconfluent 3T3-L1 cells were differentiated in the absence or in the presence of the aq. MeOH extract of L. raddeana and its fractions (1.56 - 200 μg/mL) and chicoric acid, luteolin 7-O-glucuronide (1.56 - 200 μg/mL) for 6 days.
PPT Slide
Lager Image
Effect of Aq. MeOH extract of L. raddeana, its fractions and its substances (luteolin 7-O-glucuronide and chicoric acid) on lipidaccumulation of 3T3-L1 adipocyte differentiation. Postconfluent 3T3-L1 cells were differentiated in the absence or presence of theextracts and fractions of L. raddeana (1.56 - 200 μg/mL), chicoric acid, and luteolin 7-O-glucuronide (1.56 - 200 μg/mL) for 6 days. (a) Fat droplets were observed by Oil Red O staining.
Peroxynitrite-scavenging activity of the aq. MeOH extract
PPT Slide
Lager Image
*Value represents mean ± S.D. (n = 2).
Peroxynitrite-scavenging activity of chicoric acid (1), luteolin 7-O-glucuronide (4) and penicillamine (positive control)
PPT Slide
Lager Image
*Value represents mean ± S.D. (n = 2).
In conclusion, it is suggested that the leaves of L. raddeana , which are used to treat diabetic complications including obesity and cardiovascular disorders, contained high amounts of chicoric acid and luteolin 7- O -glucuronide. These two substances are regarded as the active substances in L. raddeana which are responsible for suppressing fat accumulation in 3T3-L1 adipocytes. Therefore, the leaves of L. raddeana could be used to prevent or treat obesity.
Acknowledgements
This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (#20100021039).
References
Rayalam S. , Della-Fera M. A. , Baile C. A. 2008 J. Nutr. Biochem. 19 717 - 726    DOI : 10.1016/j.jnutbio.2007.12.007
Steinberger J. , Daniels S. R. 2003 Circulation 107 1448 - 1453    DOI : 10.1161/01.CIR.0000060923.07573.F2
Huang C. C. , Huang W. C. , Hou C. W. , Chi Y. W. , Huang H. Y. 2014 Int. J. Mol. Sci. 15 8280 - 8292    DOI : 10.3390/ijms15058280
Wu B. T. , Hung P. F. , Chen H. C. , Huang R. N. , Chang H. H. , Kao Y. H. 2005 J. Agric. Food Chem. 53 5695 - 5701    DOI : 10.1021/jf050045p
Hsu C. L. , Yen G. C. 2006 Mol. Nutr. Food Res. 50 1072 - 1079    DOI : 10.1002/mnfr.200600040
Hsu C. L. , Lo W. H. , Yen G. C. 2007 J. Agric. Food Chem. 55 7359 - 7365    DOI : 10.1021/jf071223c
Yang J. Y. , Della-Fera M. A. , Hartzell D. L. , Nelson-Dooley C. , Hausman D. B. , Baile C. A. 2006 Obesity 14 1691 - 1699    DOI : 10.1038/oby.2006.194
Furukawa S. , Fujita T. , Shimabukuro M. , Iwaki M. , Yamada Y. , Nakajima Y. , Nakayama O. , Makishima M. , Matsuda M. , Shimomura I. 2004 J. Clin. Invest. 114 1752 - 1761    DOI : 10.1172/JCI21625
Pires K. M. , Ilkun O. , Valente M. , Boudina S. 2014 Obesity 22 178 - 187    DOI : 10.1002/oby.20465
Pacher P , Beckman J. S. , Liaudet L. 2007 Physiol Rev. 87 315 - 424    DOI : 10.1152/physrev.00029.2006
Ramirez-Zacarias J. L. , Castro-Munozledo F. , Kuri-Harcuch W. 1992 Histochemistry 97 493 - 497    DOI : 10.1007/BF00316069
Kooy N. W. , Royall J. A. , Ischiropoulos H. , Beckman J. S. 1994 Free Radic. Biol. Med. 16 149 - 156    DOI : 10.1016/0891-5849(94)90138-4
Xiao H. , Wang J. , Yuan L. , Xiao C. , Wang Y. , Liu X. 2013 J. Agric.Food Chem. 61 1509 - 1520    DOI : 10.1021/jf3050268
Kour K. , Bani S. 2011 Pharmacol. Biochem. Behav. 99 342 - 348    DOI : 10.1016/j.pbb.2011.05.008
Reinke R. A. , Lee D. J. , McDougall B. R. , King P. J. , Victoria J. , Mao Y. , Lei X. , Reinecke M. G. , Robinson W. E. 2004 Virology 326 203 - 219    DOI : 10.1016/j.virol.2004.06.005
Min Y. S. , Bai K. L. , Yim S. H. , Lee Y. J. , Song H. J. , Kim J. H. , Ham I. , Whang W. K. , Sohn U. D. 2006 Arch. Pharm. Res. 29 484 - 489    DOI : 10.1007/BF02969421
Vilela F. C. , Padilha-Mde M. , Alves-da-Silva G. , Soncini R. , Giusti-Paiva A. 2010 J. Med. Food 13 219 - 222    DOI : 10.1089/jmf.2008.0303
Nagy M. , Krizková L. , Mucaji P. , Kontseková Z. , Sersen F. , Krajcovic J. 2009 Molecules 14 509 - 518    DOI : 10.3390/molecules14010509