Article Cite This: J. Agric. Food Chem. 2018, 66, 1039−1046
pubs.acs.org/JAFC
Analysis of the Effects of δ‑Tocopherol on RAW264.7 and K562 Cells Based on 1H NMR Metabonomics Yang Lu, Hui Li, and Yue Geng* Key Laboratory of Food Nutrition and Safety of SDNU, Provincial Key Laboratory of Animal Resistant Biology, College of Life Science, Shandong Normal University, Jinan 250014, China S Supporting Information *
ABSTRACT: δ-Tocopherol (δ-TOH) is a form of vitamin E with higher bioactivity. In this study, we studied the bioactivity of δ-TOH using the IC50 of δ-TOH on RAW264.7 (80 μM) and K562 (110 μM) cells. We compared the differential metabolites from the cell lines with and without δ-TOH treatment by 1H NMR metabonomics analysis. It was found that δ-TOH affected the protein biosynthesis, betaine metabolism, and urea cycle in various ways in both cell lines. Metabolic levels of the cell lines were changed after treatment with δ-TOH as differential metabolites were produced. The betaine level in RAW264.7 cells was reduced significantly, while the L-lactic acid level in K562 cells was significantly enhanced. The metabolic changes might contribute to the switch of the respiration pattern from aerobic respiration to anaerobic respiration in K562 cells. These results are helpful in further understanding the subtoxicity of δ-TOH. KEYWORDS: 1H NMR, metabonomics, δ-TOH, RAW264.7, K562
1. INTRODUCTION Vitamin E, as an important nutrient, is necessary for maintaining the metabolism and physiological functions of humans. The best way to supplement vitamin E is to ingest it from food such as almonds, hazelnuts, soybeans, avocado, and so on.1 Vitamin E, a fat-soluble vitamin, contains tocopherols (TOH)2 and tocotrienols (TT), including their α-, β-, γ-, and δforms. The classification is based on the different numbers and positions of methyl groups on the chroman ring. The bioactivity of α-TOH is the largest of the different forms of vitamin E, and the bioactivity of γ- and δ-forms are only 10% and 1%, respectively.3 However, the antioxidant capacity of δTOH is stronger than the antioxidant capacity of α-TOH.4 The RAW264.7 cell line has usually been used as the common inflammatory model in research studies, and the K562 cell line has acted as the tumor phenotype, as it can be grown as suspension cell cultures in order to conduct research. Parker et al. reported that the RAW264.7 cells were incubated with αTOH, γ-TOH, and δ-TOH: Cell viability for α-TOH was not affected but for γ-TOH was low and lower still for δ-TOH.5 Therefore, we decided to study the subtoxicity of δ-TOH on RAW264.7 and K562 cells. As has been established, vitamin E plays a role in many types of physiological functions, such as antioxidation,6 antiaging,7 antiallergic, treating frostbite, protecting blood vessel and plasma membranes,8 and even signal transduction and gene expression.9 Vitamin E was reported to alleviate obesity and its metabolic complications through regulating many signaling pathways, such as the WNT signaling pathway, the Janus Kinase (JAK)-signal transducer and activator of transcription (Stat) signaling pathway, and the phosphatidylinositol 3′-kinase (PI3K)/Akt/mammalian target of rapamycin (mTOR) pathway, all of which are involved in the regulation of the bioactivity of tumors. 10,11 Recently, researchers found that γ-TOH had anticancer effects by suppressing aerobic glycolysis.11 It was reported that different © 2018 American Chemical Society
forms of vitamin E had distinct metabolic pathways in vivo and in vitro.1 According to data from the World Health Organization,12 the Recommended Nutrient Intakes (RNIs) of α-TOH is 10 mg for normal adults per day12 (145), and the Tolerable Upper Intake Level (UL) is 1000 mg12 (148). Results of a δ-TOH treatment experiment indicated that the IC50 concentration was 55 μM, 47 μM, and 23 μM for preneoplastic, neoplastic, and highly malignant mouse mammary epithelial cells.13 It was reported that a certain dose of vitamin E prevented the mice liver primary cells from the toxicity of silver nanoparticles.14 An earlier study had demonstrated that vitamin E mitigated leukopenia caused by some certain cancer chemotherapy drugs, suggesting that vitamin E might be effective in reducing the side effects of cancer chemotherapy.15 Furthermore, it was reported that the biological effects of δTOH on RAW264.7 cells were greater than of those of αTOH.3 However, there is lack of studies on the metabonomics of cells treated with δ-TOH. The field of metabolism has become more and more important in many aspects, such as cancer treatment16 and research on traditional Chinese medicine, depending on the basis of the molecule changed. With the development of metabonomics, the metabolic differences between normal cells and cancer cells have been increasingly characterized.17−19 Recently, it was reported that hematopoietic stem cell transplantation leads to lethal therapy-related myelodysplasia syndrome or acute myeloid leukemia. This process includes regulation in metabolic pathways involving alanine and aspartate metabolism, glyoxylate and dicarboxylate metabolism, phenylalanine metabolism, the citrate acid cycle, and aminoReceived: Revised: Accepted: Published: 1039
October 12, 2017 January 8, 2018 January 9, 2018 January 9, 2018 DOI: 10.1021/acs.jafc.7b04667 J. Agric. Food Chem. 2018, 66, 1039−1046
Article
Journal of Agricultural and Food Chemistry acyl-tRNA biosynthesis.20 Meanwhile, there was relatively more research on metabonomics of RAW264.7 cells than there was on metabonomics of K562 cells. Most of the existing studies are focused on the effects of α-TOH and γ-TOH on macrophages.8,21,22 The other reason for studying these two cell lines is that the RAW264.7 cell is an adherent cell, whereas the K562 cell is the suspension cell. Hence, we decided to search the key metabolites and pathways involved in the effect of δ-TOH on RAW264.7 and K562 cells. In this study, we investigated the effect of δ-TOH on RAW264.7 and K562 cells by 1H NMR metabonomics. Our study may provide useful information on the subtoxicity and biological capacity of δ-TOH.
to break up cells. Sonication was performed in the iced water bath at 1 min sonication/break alternations for a total of 9 min. The aqueous phase was separated from the organic phase by centrifugation at 12,000 rpm and 4 °C for 30 min. After centrifugation, the clear aqueous supernatant was collected. The same extraction process was repeated three times. The supernatant liquid was combined and stored at −80 °C. 2.5. Pretreatment of Samples for NMR Spectroscopy. The aqueous supernatant samples were evaporated at 60 °C, and the solid compound was dissolved in 900 μL of D2O (pH 7.4, containing 0.1% TSP). The 24 samples were centrifuged at 12,000 rpm and 4 °C for 15 min and lyophilized. The lyophilized samples were redissolved in 700 μL of D2O and centrifuged under the same conditions above. From the aqueous supernatant, 600 μL was taken and mixed with 50 μL of phosphate buffer solution involving D2O (pH 7.4, containing 0.1% TSP). A supernatant in the amount of 550 μL was collected into a 5 mm NMR tube for analysis. 2.6. Nuclear Magnetic Resonance Spectroscopy Analysis. All 1 H NMR spectra were obtained by a superconductor shielding fourier transform nuclear magnetic resonance spectrometer equipped with 13 C and 1H double resonance optimization of a 5 mm CPTCI three trans detector CryoProbeTM AVANCE 600 III (Bruker) under the following conditions: 600.104 MHz for proton resonance frequency, zg30 for pulse sequence, 12,019.230 Hz for spectral width, number of repetitions was 256, number of dummy scans at 298 K was 2, 1 s of relaxation delay, 12 μs of pulse length, and Topspin 3.2 were used to process the spectral data. 2.7. Data Analysis. The NMR spectra were processed with MestReNova 6.11 (Mestrelab Research, Santiago de Compostela, Spain). All spectra were added with an exponential window function to the spectra and to process the manual baseline correction. Taking TSP as the standard, the spectra were manually cut off from the water peak and normalized. Data of the integrated peak were exported into numbered groups in Excel files. Next, the data were imported into SIMCA-P+12.0 software (Umetrics Inc., Umea, Sweden). The data were analyzed using principal components analysis (PCA), partial least-squares-discriminant analysis (PLS-DA),24 and orthogonal partial least-squares-discriminant analysis (OPLS-DA). The reliability of the PLS and OPLS-DA models was verified by permutation testing and CV-ANOVA.25 Differential metabolites were identified based on variable importance in projection (VIP) and loading weights of primary predictive component. The chemical shifts (δ) were selected with the standard of their values of VIP ≥ 1. Then, the values of chemical shift were imported into the Biological Magnetic Resonance Bank (BMRB), and the list of possible materials was compiled. Based on the list, we identified the substances through matching the locations and patterns of peaks in the human metabolome database (HMDB).24 In order to quantify the different up and down regulations of metabolites, we normalized the peak areas of identified metabolites by the sum of the metabolites of the integral area multiplied by 1000 (relative value). After distinguishing certain metabolites, we enriched pathways through the Kyoto Encyclopedia of Genes and Genomes (KEGG) and MetaboAnalyst (MetPA). The metabolites were imported into KEGG to get a list of their KEGG ID numbers. Then, the list was imported into MetPA to match the metabolic pathways with two methods. One method is the pathway analysis, whose pathway impact characterization of horizontal could coordinate graphs by the topological analysis of the importance of the metabolic pathways of the computed value, while the ordinated −log P provided significant metabolic pathway enrichment analysis. According to P < 0.05, we identified the possible metabolic pathways in the two cell lines. Another method to analyze the functional enrichment analysis of humans and mammalians is called metabolite set enrichment analysis (MSEA). We chose Over Representation Analysis (ORA) to perform analysis when a list of compound names was provided. Statistical analyses were used by SPSS version 22.0, and the results were shown by mean ± SEM. Different analysis and comparisons between groups were analyzed by variance analysis and the Duncan test.
2. MATERIALS AND METHODS 2.1. Reagents and Instruments. The following reagents were used: δ-TOH (≥98%) (Tauto Biotech, Shanghai, China); D2O containing 0.1% TSP, DMEM culture medium, and RPMI 1640 culture medium (Sigma-Aldrich, St. Louis, MO); dimethyl sulfoxide (DMSO) and 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) (Solarbio, Beijing, China); penicillin, streptomycin, and L-glutamine (M&C Gene Technology Ltd., Bejing, China); and fetal calf serum (FBS) (Zhejiang Tianhang Biotechnology Co., Ltd., Hangzhou, China). The following instruments were used: Stat Fax-2000 Microplate Reader (Awareness Technology, Palm City, FL); LabServ CO2 incubator (Thermo Fisher, Waltham, MA); Epsilon 2-4 LSCplus freeze-dryer (Christ, Osterode, Germany); Centrifuge 5804R (Eppendorf, Hamburg, Germany); SB-1000 (Eyela, Tokyo, Japan); VC 130PB (Sonic Material Inc., Newtown, CT); and AVANCE III 600 MHz (Bruker Biospin, Zurich, Switzerland). 2.2. Cell Culture. RAW264.7 cells were purchased from the Chinese Type Culture Collection (CTCC, Shanghai, China) and incubated in DMEM supplement with 100 units/mL of penicillin, 100 μg/mL of streptomycin, 10 μL/mL of L-glutamine, and 10% FBS. This cell line was macrophage-originated from the ascites in the leukemia virus-induced tumor Abelson murine. The K562 cells were purchased from the Chinese Type Culture Collection (CTCC, Shanghai, China) and cultured in RPMI 1640 supplemented with 100 units/mL of penicillin, 100 μg/mL of streptomycin, 10 μL/mL of L-glutamine, and 10% FBS. The cells were cultured at 37 °C with 5% CO2 and at 95% humidity. The K562 cell is a cell line derived from a 53-year-old female chronic myelogenous leukemia patient in blast crisis. δ-TOH was dissolved in ethanol before being added to culture medium (the final concentration of ethanol was 0.5%). 2.3. Cell Viability Assay. To test the effects of δ-TOH on cell viability, the MTT experiment could be utilized. Cells in suspension were placed in 96-well plates at 1 × 105 cells/well and incubated for 8 h at 37 °C. Then, the old culture medium was removed, and fresh culture medium containing δ-TOH with different concentrations (20, 40, 50, 80, and 100 μM) was added. The cells were incubated for 48 h at 37 °C. Then 20 μL of MTT was added to cells in each well of the 96-well plates and incubated for 4 h at 37 °C. The culture medium in each well was then removed, and 150 μL/well of DMSO was added followed by incubation for 20 min with shaking at room temperature. The plates were centrifuged with 4 °C and 1,000 rpm for 10 min before removing medium for K562 cells. The absorbance of the cell suspension was measured at 492 nm with the Microplate Reader. 2.4. Cell Extraction. Based on the results from the MTT experiment, the half-maximal inhibitory concentrations (IC50) of δTOH were 80 μM for RAW264.7 cells and 110 μM for K562 cells. The cells were cultured in the culture medium with a certain amount of δ-TOH for 48 h at 37 °C and washed with PBS three times, and the addition of 4 mL of iced methanol was utilized for fixation. The cells were removed from the culture dishes with a cell scraper,23 suspended in the iced methanol, and stored at −4 °C. Then, the cell suspensions in methanol were then mixed with ultrapure water and trichloromethane at a ratio of 2.85:4:4 (V:V:V, ultrapure water:methanol:trichloromethane). After vortexing, the cell suspensions were sonicated 1040
DOI: 10.1021/acs.jafc.7b04667 J. Agric. Food Chem. 2018, 66, 1039−1046
Article
Journal of Agricultural and Food Chemistry
3. RESULTS 3.1. Effect of δ-TOH on Cell Viability. To determine the IC50 of δ-TOH acting on RAW264.7 and K562 cells, we added different concentrations of δ-TOH adding to the cell culture.3 We found the distinguishing effects of δ-TOH on these two cell lines after 48 h of culturing, as shown in Figure 1a. Although
Figure 2. 1H NMR spectra show metabolites in (a) RAW264.7 cells (n = 6) and (b) K562 cells (n = 6).
3.3. Multivariate Statistical Analysis. Then, the integral areas of chemical shifts in Excel files were applied to the SIMCA-P software for multivariate statistical analysis. It was found that two cell lines were separated by PCA and PLS-DA compared with each control group. The score plot and loading plot of these cell lines showed that the cells treated with δTOH were able to be distinguishable from their controls. The data dimension was first reduced with the PCA multivariate model. The majority of the information contained in the data-rich area was extracted for the observation of the natural distribution and the relationship between the control group and the δ-TOH group (Figures S2A and S2B). It could be seen from the first and second principal components (PC1:88.38%; PC2:9.16%) that these two groups could be distinguished, which showed that the effects of δ-TOH on the metabolic profile changed in RAW264.7 cells. Also, the results of K562 cells suggested that it was separated into two principal components (PC1:61.01%; PC2:20.45%). As an unsupervised analysis method, PCA only reflects the original state of the data. However, many factors in the environment and some systematic errors might affect the results of the experiment in the realistic situation. In order to eliminate such interfering factors and obtain more accurate results, the original data were further analyzed using the supervised PLS-DA analysis and OPLS-DA analysis. Three indicators, R2X, R2Y, and Q2, were used to evaluate the PLS-DA model. R2X and R2Y represent the percentage of the X and Y matrix, while Q2 indicates the predictive ability of the model. For each of the above three values, the closer to 1, the more credible the model is considered. Figure 3a shows that the PLS-DA model fitted well in RAW264.7 cells (R2X = 0.975,
Figure 1. The effect of different concentrations of δ-TOH on the viability of RAW264.7 and K562 cells. (a) The cell viability was measured by MTT experiment (n = 6) (Duncan test, P < 0.01), (b) regression equation of (a), and (c) effects of ethanol concentration on two cell lines.
the decreasing trends occurred in both cell lines, the IC50 of these were visibly different (Figure 1b). According to the regression equation of cell viability, we determined the IC50 of these two cell lines, which were 80 μM for RAW264.7 cells and 110 μM for K562 cells, respectively. As seen in Figure 1c, the same concentration of ethanol used to dissolve δ-TOH had no significant effect on these two cells. These results showed that there is a significant difference in tolerance and sensitivity between the RAW264.7 and K562 cells to treatment of δ-TOH. 3.2. NMR Spectroscopy. One-dimensional (1D) 1H NMR spectra of aqueous extracts of two cell lines are shown in Figures 2a and 2b, respectively. The superimposed spectra of the control group and δ-TOH treatment on cells (δ-TOH group) are shown in Figure S1. 1041
DOI: 10.1021/acs.jafc.7b04667 J. Agric. Food Chem. 2018, 66, 1039−1046
Article
Journal of Agricultural and Food Chemistry
Figure 3. The multivariate statistical analysis measured in RAW264.7 and K562 cells. The PLS-DA score scatterplot of (a) RAW264.7 cells and (b) K562 cells. The validation model of (c) RAW264.7 cells and (d) K562 cells.
R2Y = 0.981, Q2 = 0.972). Also, the PLS-DA model of the K562 cells fitted better with the 0.813 of R2X, 0.964 of R2Y, and 0.848 of Q2 (Figure 3b). The external model validation method (permutation experiment) was used to evaluate the validity of the model. We found that for the regression lines generated cell lines have relatively large slopes. Any randomly generated R2 and Q2 values were smaller than the values at the right ends of the two lines (the gap between the right end values was small), indicating that the model was effective to perform further analyses. Figures 3c and 3d showed that the PLS-DA models of RAW264.7 and K562 cells based on the permutation experiment do not overfit. In the OPLS-DA models, the score scatter figure of RAW264.7 cells showed that the δ-TOH group can be distinguished from the control group to a high extent and that the intergroup differentiation is low for RAW 264.7 cells, compared with the unsupervised principal component analysis (Figure S2C). Figure S2D showed a similar situation of separation of K562 cells could be found through the OPLS-DA score scatter figure. Combining with VIP (≥1), differential metabolites were identified, including 9 metabolites for RAW264.7 cells and 10 metabolites for K562 cells (Tables 1 and 2). The P-value of CV-ANOVA (
