Oleanolic acid mediated the proliferation and invasion of U251 glioma cells and promoted their apoptosis through the IKK-β, MAPK3, and MAPK4 signaling pathway


 ObjectTo investigate the effects of Oleanolic acid (OA) on proliferation, apoptosis, migration, and invasion of human glioma cell U251, as well as IKK-β and MAPK signaling pathways.MethodsThe binding of OA to IKK-β and MAPK signaling pathway essential proteins IKK-β, MAPK3, and MAPK4 was analyzed by molecular docking technique. U251 cells were treated with different concentrations of OA. The proliferation and apoptosis rates of U251 cells were detected by CCK-8 assay, MTT assay, cell cloning assay, and AnnexinⅤ FITC/PI double staining assay. Transwell chamber assay was used to detect migration and invasion of U251 cells. Finally, Western blotting was used to detect the protein expression levels of IKK-β, MAPK3, and MAPK4 in U251 cells treated with OA.ResultsThe results of molecular docking showed that OA could stably bind to IKK-β, MAPK3, and MAPK4 proteins. OA could not only effectively inhibit the proliferation and induce apoptosis of U251 cells (P < 0.05), but also significantly inhibit the invasion of U251 cells (P < 0.05). Western blot assay confirmed that OA could dramatically inhibit the protein expression levels of IKK-β, MAPK3, and MAPK4 in U251 cells (P < 0.05).ConclusionsOA may inhibit the proliferation, migration, and invasion of glioma U251 cells by binding key molecules of the IKK-β signaling pathway and essential target proteins of MAPK3 and MAPK4 in the MAPK signaling pathway.


Background
Glioma is currently the most aggressive and fatal human intracranial primary malignant brain tumor [1] .
Because of the speci c and remote location of these tumors, removal of glioma is di cult, and thus glioma is associated with high morbidity and mortality rates [2] . The clinical treatment of glioma includes surgery, radiotherapy, and chemotherapy drugs, but the prognosis in patients remains unsatisfactory [3] and the 5-year survival rate of glioma is less than 10%. Therefore, identifying anti-tumor drugs with good curative effects, minor side effects, and high safety is critical. At present, targeted therapy has become a new research hotspot. Compared with traditional medicines, small molecule targeted drugs have increased e cacy in targeting tumors and show advantages such as speci city, accurate targeting, good tolerance, mild toxicity, and passing the blood-brain barrier [4] .
Oleanolic acid (OA) is a type of natural pentacyclic triterpenoid compound that is present in many plants such as Hedyotis o cinalis, Ligustrum lucidum, papaya, loquat leaves, olive, and others. OA has many pharmacological effects, such as protecting the liver, lowering lipids, lowering glucose, and enhancing immunity. Studies have shown that OA can induce apoptosis in a variety of tumor cells, such as lung cancer, pancreatic cancer [5] , cervical cancer [6] , and ovarian cancer cells [7] , and many other types of tumor cells that have anti-proliferation effect. Currently, OA is widely used in clinical medicine for the treatment of acute and chronic hepatitis and as an anti-cancer adjuvant drug [8] .
In this study, we investigated the effects of OA on proliferation, migration, and invasion of glioma cells and the IKK-β, MAPK3, and MAPK4 proteins, with the aim of providing experimental basis for the clinical treatment of glioma patients with OA.

Materials
Oleanolic acid (Lot: AF9071301), its molecular formula is C 30 H 48 O 3 , molecular weight is 456.70, purchased from Chengdu Alfa Biotechnology Co., Lid. Purity of the compound used in the present study was shown to be higher than 98% by HPLC; The human glioma cell line U251 was purchased from the Cell Bank of Shanghai Academy of Sciences; DMEM was purchased from Hyclone, USA; fetal bovine serum was from GIBCO. The CCK-8 kit, PBS, 0.25% trypsin digestion solution, trypsin EDTA digestion solution, penicillin-streptomycin, and electrophoresis buffer were all purchased from Solarbio. The RIPA tissue cell rapid lysate and BCA protein quanti cation kits were purchased from Beyotime. HRP-labeled secondary antibody and the Annexin V-FITC cell apoptosis detection kit were purchased from Biyuntian Biotechnology Co., Ltd. Formaldehyde was purchased from Sinopharm Chemical Reagent Co., Ltd.
Matrigel was purchased from Corning Company. IKKβ and MAPK4 antibodies were purchased from Proteintech; MAPK3 antibody was purchased from A nity; and GAPDH antibody was purchased from CST. The Type 311 CO2 constant temperature incubator was purchased from Thermo Company. The AMR-100 microplate reader was purchased from Hangzhou Aosheng. The DMIL LED inverted uorescence microscope was purchased from Leica. The Mini-PROTEAN 3 cell electrophoresis instrument was purchased from BIO-RAD. The Tanon-5200 imaging system was purchased from Tanon.
The CytoFLEX ow cytometer was purchased from Beckman Coulter. The Transwell Chamber model 3422 was purchased from COSTAR Corporation.

Molecular docking
IKK-β (PDB ID: 4KIK), MAPK3 (PDB ID: 2QNJ) and MAPK4 (PDB ID: 5ES1) data were downloaded from the RSCB PDB database (http://www.rcsb.org/pdb/home/home.do). The 3D structures of the target proteins were saved as a PDB format le. The mol2 format le of OA was downloaded from the ZINC website (http://zinc.docking.org/). The PyMOL software was used to remove water, phosphate, and other irrelevant ions from the protein. The target protein receptor and ligand were separated, and the ligand molecules were extracted. The pre-processed target protein and the original ligand were saved as PDB format les. AutoDockTools 1.5.6 software was used to set the size of Grid Box and box, search for the parameters of protein active site (active port), convert the active small molecules, pretreated target protein, and original ligand into pdbqt format le, run AutoDock Tools 1.5.6 software to perform Autogrid and Docking operations, and nally use Open Babel GUI and PyMOL software to analyze the docking results visually.

Cell culture
Human glioma U251 cells were cultured in DMEM high glucose medium containing 10% fetal bovine serum and 1% penicillin-streptomycin in a 5% CO 2 incubator at 37°C. Cells in logarithmic growth phase were used in experiments.

Cell viability assay
Cell viability of U251 cells was measured using a CCK-8 assay. Cells were plated in a 96-well plate with 3×103 cells/well in 100 µL per well and cultured overnight. Cells were treated with OA (0 2.5 10 20 50 100 µg/mL), with three wells for each treatment group, for 8 h. Next, 10 µL of CCK-8 solution was added to each well, and cells were incubated for 4 h. The OD value of each well was measured at 450 nm with an enzyme-linked immunoassay. The relative proliferation inhibition rate was calculated as (%) = [1-(OD value of each drug group-OD value of the blank group) / (OD value of the control group-OD value of the blank group)] × 100%.

MTT assay
Cells were plated in 96-well culture plates at 1-5×10 4 cells/mL and cultured overnight. Cells were then treated with OA (0 10 20 50 100 200 500 1000 µg/mL), with three wells for each treatment group, and cultured for an additional 24 h. Next, 20 µL of MTT solution (5 mg/mL) was added to each well, and cells were incubated for 4 h. The supernatant was then discarded. DMSO (150 µL) was added to each well and the plates were shaken for 10 min. Absorbance at 490 nm was measured with an enzyme plate analyzer. In addition, the OA group (100 µg/mL) was operated according to the above method, and the absorbance (OD value) at 490nm wavelength was measured after 0, 1, 2, 3, 4, 5, and 6 days of culture. The proliferation inhibition rate of each group was calculated.

Colony formation assay
The cell Colony formation assay were set up in control group and oleanolic acid group (100 µg/mL). Cells were inoculated in a 6 cm dish at 1×10 3 cells per dish and then cultured for 2 to 3 weeks. The culture was discontinued when the cells formed visible clones. The supernatant was discarded and the plates were washed with PBS; cells were xed with 4% paraformaldehyde for 15 min. The xed solution was discarded, and cells were stained with crystal violet staining solution for 10-30 min. The staining solution was removed and plates were air-dried. The number of clones was calculated using a microscope, and the rate of clone formation was determined as = (Number of clones/Number of inoculated cells) × 100%.

Transwell assays
The Transwell assays were set up in control group and oleanolic acid group (100 µg/mL). Transfected cells were cultured in serum-free medium for 24 h before the experiment. Matrigel mixed with serum-free medium (1:2; 80 µL) was added to a Transwell chamber, and the chamber was incubated at 37°C for 30 min to solidify. Cells (300 µL, 3 × 105 cells/mL) were added to the top chamber of the Transwell chamber and 700 µL of 10% FBS-containing complete culture medium was added to the bottom chambers. Cells were cultured for 24 h. Next, 4% formaldehyde solution was added to each well and cells were xed at room temperature for 10 min. The xing solution was removed and the cells were washed by PBS, followed by staining with 1 ml of 0.5% crystal violet solution for 30 min. Samples were washed three times with PBS and air-dried. The cells were observed and photographed under a microscope (at 400×), and the number of invaded cells in each group was counted.

Flow cytometry assay
The Flow cytometry assay were set up in control group and oleanolic acid group (100 µg/mL). Cells were harvested (5 × 10 4 to 1 × 10 5 cells) and resuspended in PBS. Next, 195 µL Annexin V-FITC binding solution was added to cells, followed by 10 µL Annexin V-FITC and 5 µL propidium iodide staining solution. The samples were mixed and incubated in the dark at room temperature for 15 min and then placed in an ice bath. A sample without Annexin V-FITC and PI was used as a negative control. The apoptosis rate of each group was detected by ow cytometry.

Western blotting assay
Cells were lysed with RIPA rapid lysis buffer and the BCA protein quantitative method was used to determine protein concentration. After denaturation, protein samples were separated by 10% SDS-PAGE gel electrophoresis and transferred to a PVDF membrane. The membrane was blocked with 5% skimmed milk powder at room temperature for 1 h and then incubated with the following primary antibodies overnight at 4°C: IKKβ antibody (1:600), MAPK3 antibody (1:1000), MAPK4 antibody (1:400), and GAPDH antibody (1:2000). After three washes in TBST for 5 min, the membrane was incubated with HRP-labeled secondary antibody (1:1000) at 37°C for 1 h and then washed with TBST three times. The membrane was developed using the ECL chemiluminescence method and the protein bands were visualized using a chemiluminescence imaging system. The protein bands were analyzed by the Tanon-5200 image analysis system with GAPDH as the internal reference.

RNA-Sequencing of Oleanolic acid
The control group and oleanolic acid treatment group were set up, and the total RNA of human glioma cell U251 was extracted respectively ™ RNA purity was detected by one / onec, and life Invitrogen qubit was used ® 3.0 uorescence quantizer was used for accurate quanti cation, and Agilent 4200 tapestation system was used to evaluate RNA integrity (RIN value). After the sample is quali ed, the library construction of the sample is carried out. Then the library concentration was accurately quanti ed by Kapa qPCR and the library fragment size was accurately detected by Agilent 4200 TapeStation. After the library is quali ed, the different libraries are pooled according to the effective concentration and the target off-machine data volume for Illumina PE150 sequencing.

GO and KEGG Enrichment Analyses
Gene Ontology (GO) is a comprehensive database describing gene functions, which can be divided into three parts: molecular function, biological process and cellular component. Kyoto Encyclopedia of Genes and Genomes (KEGG) is a public database resource for high-level functions and biological systems (such as cells, organisms, and ecosystems). It integrates genome, chemistry, and system function information. The categories of KEGG are represented by pathways. The clusterPro ler hypergeometric distribution algorithm was applied to the selected differential genes to analyze the GO function of the differential gene enrichment and the KEGG pathway enrichment analysis, and the enrichment results were visualized.

Statistical analysis
GraphPad Prism 5.0 software was used for image processing. The SPSS 21.0 software was used for statistical analysis. Measurement data are expressed as x ± s, and the comparison of multiple groups was performed by the one-way analysis of variance (ANOVA) method. P < 0.05 indicated statistical signi cance.

OA shows binding activity with IKK-β, MAPK3, and MAPK4
The interaction model of OA with IKK-β, MAPK3, and MAPK4 was examined through molecular docking. The docking results are shown in Table 1 and Fig. 1. Studies have shown that a lower binding energy required for the binding of the ligand and the receptor is associated with a more stable binding conformation and higher binding possibility. A binding energy less than 0 indicates that the ligand and receptor can spontaneously bind. We found that the binding energy of OA with IKK-β, MAPK3, and MAPK4 was − 5 KJ/mol through hydrogen bonds with ASP-484, TYR-134, ILE-62, LYS-193, GLU-81, and ASP-187 in IKK-β, MAPK3, and MAPK4 kinases, indicating that OA has an excellent binding activity with the three target proteins. µg/mL induced 100% inhibition. (Fig. 2). Comparing the different treatment groups with the control group, the difference in inhibition rate was statistically signi cant (P < 0.05). These results suggested that OA inhibits the proliferation of U251 glioma cells, and it has a concentration-dependent effect at 2.5-50µg/mL.
The MTT assay results showed that OA inhibited U251 glioma cells proliferation. Concentrations of 500 µg/mL induced 91% inhibition. Signi cant inhibition of proliferation was seen at concentrations at low as 10 µg/mL. (Fig. 3A). Comparing the different treatment groups with the control group, the difference in inhibition rate was statistically signi cant (P < 0.05). The results indicate that oleanolic acid can signi cantly inhibits the proliferation of U251 cells, and its inhibition rate is positively correlated with the concentration of 10µg/mL ~ 500µg/mL, and there is a phenomenon of inhibition and saturation afterwards. After treatment with 100µg/mL oleanolic acid for 3 days, the proliferation ability of U251 cells decreased signi cantly, and after the 4th day, it had almost no effect on the proliferation ability of U251 cells (Fig. 3B).

Cell clone assay to detect cell proliferation
We next performed cell clone assays to evaluate the effect of OA on the colony formation ability of U251 glioma cells. The number of cell clones (41 ± 2) in the OA group was signi cantly reduced compared with the number in the control group (124 ± 7) (P < 0.005) (Fig. 4). These results con rmed that OA treatment signi cantly decreased the colony formation ability of U251 glioma cells.

Cell invasion was detected by Transwell chamber assay
Transwell experiments showed that the number of invaded cells in the OA treatment group (34 ± 4) was signi cantly reduced compared with the control group (74 ± 5) (P < 0.005) (Fig. 5). These results indicated that OA substantially inhibits the invasion ability of human glioma U251 cells.

OA promotes apoptosis of human glioma U251 cells
We next evaluated apoptosis using Annexin V-FITC/PI double staining and ow cytometry. The results showed that the apoptotic rate of U251 cells after OA treatment was 25.27 ± 1.11%, while the apoptotic rate of the control group was 4.39 ± 0.22% (Fig. 6). The apoptotic rate increased signi cantly after OA treatment, and the difference was statistically signi cant (P < 0.01). These results showed that OA promotes the apoptosis of human glioma U251 cells.

Signi cant difference gene analysis
In order to understand the potential mechanism of oleanolic acid against glioma, we performed RNA-seq analysis on human glioma cells U251 treated with oleanolic acid and cells without oleanolic acid treatment as controls. In RNA-seq analysis, a total of 446 signi cantly differentially expressed genes were detected. The Volcano Plot chart can be very intuitive to view the overall distribution of genes that are differentially expressed between two samples. According to the set threshold: |log2(FoldChange)|> 1 and padj < 0.05, we screen the differential genes between samples to obtain FoldChange and padj corresponding to the up-regulated and down-regulated genes. Among them, there are 96 up-regulated genes and 350 down-regulated genes that are differentially expressed. The results are shown in Fig. 8A. These genes are mainly involved in processes such as in ammation, metabolism, immunity, and regulation of cell growth. The hierarchical clustering analysis of genes was carried out, and the hierarchical clustering results were displayed by heat maps. The clustering results are shown in Fig. 8B. The clustering heat map between samples is drawn according to the fpkm value of different genes.

GO and KEGG Enrichment Analyses
In order to study the functions of oleanolic acid signi cantly different genes, clusterPro ler hypergeometric distribution algorithm was used to analyze the GO function and KEGG pathway enrichment analysis of the different genes. The results of GO and KEGG enrichment of differential genes are visualized, as shown in Fig. 9. In the GO functional enrichment category, it is about cellular biological processes, defense response defense response, cellular response to interferon, basement membrane, endoplasmic reticulum, extracellular matrix, cytokine activity, cytokine There are many genes with signi cant differences in the activity of receptor binding, cytokine receptor binding, transcription factor activity, and transcription activator activity. The KEGG pathway is enriched, with a total of 236 signal pathways enriched, including TNF signaling pathway, NOD-like receptor signaling pathway, and AGE-RAGE signaling pathway in diabetic complications, etc. TNF-α can up-regulate the protein expression of ERK1/2 in glioma cells. MAPK extracellular signal-regulated kinase MEK-ERK inhibitor can signi cantly attenuate the invasion of U87 glioma cells mediated by TNF-α stimulation. Inhibition of ERK1/2 [9] . The combination of TBK-1, a member of the IKB kinase family, and IKKε promotes autophagy and in ammation by activating important factors related to cancer development, including Akt and TNF-Rrelated factor 2TRAF2, and inhibits tumor suppressor factors. Its inhibitors promote the survival and growth of glioblastoma cells through anti-in ammatory, autophagy and apoptosis pathways [10] . It can be seen from KEGG results that oleanolic acid may activate MAPK pathway through TNF pathway, thus inhibiting the proliferation of human neural U251 cells.

Discussion
Glioma is the most common malignant tumor among all intracranial malignancies [11] and is a complex tumor lesion that is characterized by invasiveness and a high recurrence rate. Despite the availability of treatments, such as early surgical resection, radiation, or chemotherapy, the prognosis of patients with glioma is poor. In addition to the poor therapeutic effect of chemotherapy, early radiotherapy, and late adjuvant therapy on malignant glioma, these treatments also have toxic side effects. Therefore, identifying new therapeutic agents to improve the prognosis of patients with malignant glioma is critical [12,13] . To overcome the limitations of the existing treatment methods for malignant glioma, many new treatment methods such as molecular targeted therapy, immunotherapy, gene therapy, stem cell therapy, and nanotechnology have begun to appear in preclinical and clinical research. In recent years, traditional Chinese medicine has shown signi cant value in the treatment of human glioma. Bi [14] found that Paris polyphylla saponin reduced the viability of U251 cells by inhibiting the expression of ARA1 and ARA3, leading to inhibited phosphorylation of Akt and p44/42 MAPK and induced apoptosis and cell cycle arrest. Liu [15] showed that berberine inhibited the proliferation of glioma cells by interfering with wild-type and mutant p53.
OA is an effective anti-tumor drug that inhibits glioma cells with a variety of malignant phenotypes.
Studies have shown that OA and its derivatives have no cytotoxicity to normal human cells [16] and it inhibits tumor development in various malignant tumors through multiple methods and pathways. In vivo experiments showed that OA signi cantly reduced the size and quality of tumors, and in vitro experiments revealed that OA has a signi cant inhibitory effect on cell migration, cell viability and proliferation ability [17,18] . OA has a sensitizing effect on radiotherapy of C6 cell-bearing rats [19] . The pathway by which OA inhibits cell processes varies in different tumor cells, and its anti-tumor effect is related to the inhibition of intracellular signaling pathways such as STAT3, VEGF, MAPK, ERK, JNK, Akt, mTOR, and NF-κB pathways. In this study, molecular docking technology showed that the binding energies of OA and IKK-β, MAPK3, and MAPK4 proteins were all less than − 5 kcal /mol, indicating that OA had a good binding activity with the three proteins. We speculate that OA may inhibit glioma proliferation by downregulating IKK-β, MAPK3, and MAPK4 proteins.
To further study the effect of OA on glioma, we performed in vitro experiments. CCK-8 and MTT assays revealed that the proliferation inhibition rate of cells increased in response to OA in a dose-dependent manner. Colony formation assays showed that the number of U251 cells decreased signi cantly after OA treatment. Together these results indicated that OA inhibits the proliferation of U251 cells. Invasion is a critical biological feature of malignant tumors. It is the process by which tumor cells penetrate the blood vessel wall and enter the circulatory system. Transwell assays showed that the number of invaded U251 cells treated with OA was signi cantly reduced compared with the control group, suggesting that OA inhibits the invasion of U251 cells. Apoptosis is a process in which cells undergo programmed death through gene regulation. Many tumor suppressor genes inhibit tumor development through inducing apoptosis, and apoptosis induction is one of the main mechanisms of current tumor treatments. Flow cytometry showed that the apoptosis rate of cells treated with OA was signi cantly increased, indicating OA has a promoting effect on cell apoptosis. Together these results show that OA inhibits the proliferation and invasion of glioma U251 cells and promotes glioma cell apoptosis, thereby exhibiting anti-cancer effects in glioma.
The IKK and MAPK pathways play an essential role in the proliferation, migration, and invasion of glioma cells. The overexpression of IKK-β is closely related to the occurrence of a variety of malignant tumors, and inhibition of IKK-β in tumor cells results in speci c anti-tumor effects [20][21][22] . Recent studies found a close relationship between malignant tumors and chronic in ammation. IKK-β activation helps promote the occurrence of cancer and in ammatory lesions, and in ammation can inhibit cell apoptosis through the IKK-β pathway, thus enabling the development malignant tumors. The speci c inhibition of IKK-β provides a new strategy and approach for the prevention and targeted treatment of in ammationassociated tumors. Activation of the IKK/NF-κB pathway induces tumor cells to proliferate and migrate and inhibits apoptosis [23] . Wang et al. showed that TGFβ1 regulates TNF-κB activity induced by TNF-α in glioma cells through the PP1A/IKK-β pathway and promotes the progression and development of gliomas [24] .
The MAPK pathway participates in the regulation of cancer cells through roles in proliferation, differentiation, and apoptosis, and other critical biological responses [25,26] . Recent studies showed that the MAPK signaling pathway has a series of active metastases in glioma and other malignant tumor tissues, and MAPK directly mediates the anti-tumor effects of a variety of drugs. One study showed that miR-326 inhibits the malignant biological behavior of glioma cells through the MAPK signaling pathway [27] . MAPK3 and MAPK4 are members of the MAPK pathway. Yang et al. showed that miR-127-3p inhibits the proliferation of U251 glioma cells by down-regulating the target gene MAPK4 [28] . Some researchers also reported that a new MAPK3 and MAPK4 double-molecule inhibitor PCC0208017 reduced glioma cell migration, glioma cell invasion, and angiogenesis and may be a promising lead compound for the treatment of glioma [29] . Studies have con rmed that CYP17A1 gene silencing signi cantly inhibits the invasiveness of glioma cells, and CYP17A1 gene silencing promotes the apoptosis of glioma cells [30] . Recent studies have shown that PTGS2 can not only promote the proliferation rate of glioma cells, but also improve radiation tolerance [31] . The studies described above showed that the CYP17A1/IKKβ/PTGS2/MAPK3/MAPK4 signaling pathway plays an essential role in the development of glioma. In this study, our western blot analysis showed that the protein expression levels of CYP17A1, IKK-β, PTGS2, MAPK3, and MAPK4 were suppressed in glioma cells treated with OA. We speculate that OA may inhibit IKK-β, MAPK3, and MAPK4 protein expression by directly binding to the proteins. In this study, in vitro experiments observed that OA has inhibitory effects on the proliferation, cloning and invasion of glioma cells, and can promote their apoptosis. Furthermore, through transcriptome sequencing, it was found that the signi cantly differentially expressed genes are mainly involved in processes such as in ammation, metabolism, immunity, and regulation of cell growth. In the GO function enrichment analysis, information about cellular biological processes, defense response, cellular response to type interferon, basement membrane, endoplasmic reticulum, extracellular matrix, cytokine activity, cytokine receptor binding, transcription factor activity and transcription activator activity, etc. There are many genes with signi cant differences in activity. It can be seen from KEGG results that oleanolic acid may activate MAPK pathway through TNF pathway, thus inhibiting the proliferation of human neural U251 cells.

Conclusions
In summary, we found that OA inhibits the proliferation and invasion of U251 glioma cells and promotes cell apoptosis. We further showed that OA inhibited the IKK-β/MAPK3/MAPK4 signaling pathway in glioma cells. These ndings suggest that OA may have a good application prospect in the clinical treatment of glioma and provides ideas for the development and study of tumor drugs. The treatment of human glioma with oleanolic acid still needs to be further studied in vivo. Further in-depth research can lay a theoretical foundation for nding new molecular targeted therapeutic drugs at the gene level. Page 16/20 CCK-8 assay to detect the inhibitory effect of OA on the proliferation of U251 glioma cells. Data represent mean SD of three different experiments. *** P < 0.05 compared with control group.  Cell cloning to detect the proliferation inhibitory effect of OA on human glioma U251 cell. Data represent mean SD of four different experiments. *** P < 0.005 compared with control group. Transwell chamber method to detect the effect of OA on the invasion of human glioma U251 cells. Data represent mean SD of six different experiments. *** P < 0.005 compared with control group.