Kukoamine A inhibits human glioblastoma cell growth and migration through apoptosis induction and epithelial-mesenchymal transition attenuation

Cortex lycii radicis is the dried root bark of Lycium chinense, a traditional Chinese herb used in multiple ailments. The crude extract of Cortex lycii radicis has growth inhibition effect on GBM cells. Kukoamine A (KuA) is a spermine alkaloid derived from it. KuA possesses antioxidant, anti-inflammatory activities, but its anticancer activity is unknown. In this study, the growth and migration inhibition effect of KuA on human GBM cells and the possible mechanism of its activity were investigated. After KuA treatment, proliferation and colony formation of GBM cells were decreased significantly; apoptotic cells were increased; the cell cycle was arrested G0/G1 phase; the migration and invasion were decreased, the growth of tumors initiated from GBM cells was inhibited significantly; the expressions of 5-Lipoxygenase (5-LOX) were decreased, apoptotic proteins, Bax and caspase-3 were increased, and antiapoptotic protein Bcl-2 was decreased significantly; The expressions of CCAAT/enhancer binding protein β (C/EBPβ), N-cadherin, vimentin, twist and snail+slug were decreased significantly, while the expression of E-cadherin was increased significantly in KuA treated GBM cells and tumor tissues. KuA inhibited human glioblastoma cell growth and migration in vitro and in vivo through apoptosis induction and epithelial-mesenchymal transition attenuation by downregulating expressions of 5-LOX and C/EBPβ.


KuA induced apoptotic cells detected by AO/EB staining and flow cytometry.
KuA induced apoptotic cells were detected by AO/EB staining and flow cytometry. By AO/EB staining, the apoptotic cells in untreated U251 and WJ1 cells were 3.5 ± 2.4% and 1.6 ± 1.3%, respectively. After treatment with 40, 60 and 80 μ g/ml of KuA for U251 cells and 10, 20 and 30 μ g/ml of KuA for WJ1 cells for 48 h, respectively, the apoptotic cells in treated U251 and WJ1 cells increased significantly in a dose dependent manner ( Fig. 2A-D, p < 0.01); Annexin-V-FITC/PI double staining assay and FCM analyses showed that apoptotic cells in the untreated U251 and WJ1 cells were 2.5 ± 1.9% and 5.4 ± 0.7%, respectively, After treatment with 40, 60 and 80 μ g/ml of KuA for U251 cells and 10, 20 and 30 μ g/ml of KuA for WJ1 cells 48 h, respectively, the apoptotic cells in the treated U251 and WJ1 cells increased significantly in a dose dependent manner increased significantly in a dose-dependent manner ( Fig. 2E-H, p < 0.01). These results indicate that KuA induced apoptosis of U251 and WJ1 cells.
KuA induced cell cycle arrest of GBM cells. Cell cycle analysis was performed in KuA treated U251 and WJ1 cells. After KuA treatment, the cell population in G 0 /G 1 phase was increased and decreased in S phase in a dose dependent manner in both U251 and WJ1 cells ( Fig. 3A-D). These results demonstrate that KuA treatment caused U251 and WJ1 cells cell cycle arrest in G 0 /G 1 phase.
KuA inhibited GBM cell migration. The distance moved by the untreated and KuA treated U251 and WJ1 monolayer cells was measured and the results were expressed as migration indexes, which represent the distance migrated by KuA-treated cells relative to the distance migrated by the untreated cells. The migration abilities of KuA-treated U251 and WJ1 cells were significantly decreased in a dose-dependent manner ( Fig. 4A-D, p < 0.01). Together, these results indicate that KuA inhibited migration of U251 and WJ1 cells.

KuA inhibited GBM cell invasion.
The cell numbers penetrated the transwell membrane of the untreated U251 and WJ1 cells were 264 ± 23.9 and 384 ± 12.2, respectively, after treatment with KuA for 24 h, the cell numbers penetrated the transwell membrane were decreased significantly in a dose-dependent manner, compared to the untreated GBM cells (Fig. 4E-H, p < 0.01). These results demonstrate that KuA suppressed invasion of U251 and WJ1 cells.
KuA suppressed GBM growth in vivo. The tumor growth inhibition effect of KuA on GBM generated from WJ1 is shown in Fig. 5. KuA slowed down GBM growth significantly (Fig. 5B, p < 0.01). The mean tumor weight of the control mice was 1.55 ± 1.12 g and those of the mice treated with 10, 20, and 40 mg/kg of KuA were 1.0 ± 0.83 g, 0.79 ± 0.63 g and 0.69 ± 0.28 g, respectively (Fig. 5C,D).Tumor inhibitory rates were 35.2% (p < 0.05), 48.8% (p < 0.01) and 55.3%, (p < 0.01), respectively (Fig. 5D). No difference of body weight increase between the control and treated animals were observed (Fig. 5A). Together, this result proved that KuA inhibited GBM cell growth in vivo.

KuA affected apoptosis and cell migration associated protein expressions in vitro and in vivo.
To explore the potential molecular mechanisms underlying the growth and migration inhibition effect of KuA on human GBM cells in vitro and in vivo, apoptosis and cell migration related protein expressions in KuA treated  U251cells, WJ1 cells and tumor tissues initiated from WJ1 cells were evaluated by Western blotting. Expressions of 5-LOX and Bcl-2 proteins were decreased significantly; expressions of Bax and active caspase-3 were increased significantly in KuA treated U251cells, WJ1 cells ( Fig. 6A-D, p < 0.01) and tumor tissues (Fig. 6I,J, p < 0.01); the expressions of C/EBPβ , N-cadherin, vimentin, twist and snail+ slug were decreased significantly, while the expression of E-cadherin was increased significantly in KuA treated U251cells, WJ1 cells ( Fig. 6E-H, p < 0.01) and tumor tissues (Fig. 6K,L, p < 0.01) compared with the controls. Together, these results suggest that KuA inhibited human GBM cell growth through apoptosis induction by decreasing the expressions of 5-Lipoxygenase and Bcl-2, increasing expressions of Bax and active caspaes-3; it suppressed GBM cell migration through epithelial-mesenchymal transition attenuation by downregulating the expressions of C/EBPβ , N-cadherin, vimentin, twist and snail+ slug, and upregulating the expression of E-cadherin.

Discussion
Glioblastoma (GBM) is the most common malignant primary brain tumor with devastating proliferative and invasive characteristics; its aggressive growth pattern led GBM patients face a poor prognosis even after having received the best available treatment modalities [1][2][3][4]28,29 . However, the distinct biological behavior of aggressive proliferation and invasive growth of GBM might be the unexplored therapeutic targets for treatment of this  fatal tumor 5 . Thus, the discovery of new and specific chemotherapeutic agents inhibiting proliferation, migration and invasion of GBM cells are regarded as effective strategies to research and develop new drugs for GBM therapeutics [10][11][12][13][14]28 .
Kukoamine A (KuA) is spermine alkaloid derived from a traditional Chinese herb medicine, Cortex lycii radicis. It has hypotensive, hypoglycemic, antipyretic, antioxidant, anti-inflammatory, soybean lipoxygenase inhibition and neuroprotective activities 18,19 , but the anticancer activity of KuA and its underlying mechanism are unknown.
In this experimental study, human normal liver cells (LO2), rat glioma cells (C6), and human GBM cells were treated with KuA in vitro, the proliferation and cloning efficiency of human GBM cells were inhibited in a time-and dose-dependent manner, Little effect on human normal liver cells (LO2), were observed; human GBM cells are more sensitive to KuA than rat glioma cells (C6). After KuA treatment, apoptotic cells were increased dose-dependently; furthermore, the motility of KuA treated GBM cells (U251 and WJ1) were decreased significantly; In vivo experiment, KuA slowed down the tumor growth initiated from GBM cells (WJ1) and reduced the mean tumor weight significantly. These findings suggest that KuA might have potential growth and migration inhibition effect on human GBM cells in vitro and inhibits GBM growth in vivo.
To elucidate the molecular mechanisms of the growth and migration inhibition effects of KuA on human GBM cells in vitro and in vivo, the expressions of 5-Lipoxygenase, apoptotic proteins, CCAAT/enhancer binding protein β (C/EBPβ ) and EMT associated proteins in KuA treated GBM cells and tumor tissues were analyzed with Western blotting. 5-lipoxygenase (5-LOX) is an enzyme in charge of the metabolism of arachidonic acid to leukotrienes 27 . It plays an important role in carcinogenesis and tumor growth 30 . Increasing evidence indicated that 5-LOX is involved in the progression of different types of cancer, including glioblastoma. It promotes cancer cell survival, proliferation, migration, invasion, metastasis, and activation of anti-apoptotic signaling parhways 21,22,24,31 . 5-LOX overexpression was associated with poor prognosis of glioblastoma patients 21,22 . Inhibition of 5-LOX activity could reduce the proliferation activity of cancer cells 31 , and induce glioblastoma cell apoptotic death [24][25][26]32 .  Besides triggering apoptosis, 5-LOX was also involved in promotion of epithelial-mesenchymal transition (EMT) of cancer cells, abrogation of 5-LOX expression led to reduce featured molecular markers of EMT, including inactivation of E-cadherin and activation of snail, and cancer cell invasion 33 .
In this experimental study, KuA induced GBM cells apoptotic death dose-dependently, up-regulated the protein expressions of Bax and caspase-3 and down-regulated the expressions of Bcl-2 significantly in KuA treated glioblastoma cells and tumor tissues dose-dependently. It was suggested that KuA might have effects on activating intrinsic mitochondrial apoptotic pathway in glioblastoma cells, induces them caspase-dependent intrinsic apoptosis by 5-LOX inhibition.
Cancer cell migration and invasion resulted in the diffuse and invasive growth of glioblastoma, which makes it difficult to eradicate GBM for conventional therapeutics [3][4][5] . A mesenchymal phenotype is the biological character of GBM, GBM cells often obtain the ability to migrate, invade and metastasize through epithelial-to-mesenchymal transition (EMT) 1,34-36 .
CCAAT/enhancer binding protein β (C/EBPβ ) plays an important role in GBM cell growth, migration and invasion through EMT regulation, reduction of its expression inhibits the growth and invasion of glioblastoma cells 34,37,38 . In epithelial to mesenchymal transition, E-cadherin is a marker of epithelial phenotype, while, N-cadherin and vimentin are the markers for mesenchymal phenotype 39 ; Twist, slug and snail are the master regulators of the epithelial-mesenchymal transition 40,41 , they worked coordinately in regulation of GBM cell migration, invasion and metastasis through induction of EMT. Inhibition of the expressions of twist, slug and snail, the proliferation, migration, and invasion of glioblastoma cells were significantly suppressed [41][42][43] .
In present experiment, KuA inhibited GBM cell migration and invasion, and tumor growth dose-dependently. After KuA treatment, protein expressions of C/EBPβ , N-cadherin, vimentin, twist, slug and snail were downregulated significantly, while the expression of E-cadherin was upregulated significantly in KuA treated GBM cells and tumor tissues compared with the controls. It is suggested that EMT was be abated in GBM cells by KuA exposure, which might be associated with downregulating expression of C/EBPβ .
Based on the findings of this experimental study, it could be speculated that KuA treatment inhibited the expressions of 5-LOX and C/EBPβ in human GBM cells, which resulted in apoptosis and EMT attenuation of GBM cells, thereby inhibited human GBM cell growth and migration in vitro and in vivo.
Taken together, Kukoamine A has the potential to inhibit human glioblastoma cell growth and migration in vitro and in vivo through apoptosis induction and epithelial-mesenchymal transition attenuation mediated by downregulating expressions of 5-LOX and C/EBPβ (Fig. 7); it might serve as an effective candidate agent for the treatment and/or prevention of human glioblastoma, and deserve to be investigated further. (pH 4.7) was added to each well, cultivated for another 4 h, and then 100 μ L of DMSO was added to each well to dissolve the formazan. Absorbance was measured at 570 nm, the effect of KuA on the viabilities of on normal cells and GBM cells were calculated by: (OD 570nm of drug-treated samples)/(OD 570nm of none treated samples) × 100% 46 . Three independent experiments were performed.

Reagents
Colony formation assay. GBM cells ( U251 and WJ1) were seeded in 6 well plates at a density of 150 cells/well and 300 cells/well, respectively, After incubation overnight, different concentrations of KuA (5, 10 and 20 μ g/ml) was added, and incubated for 12 days. The cells were fixed with methanol for 15 min and stained with 0.1% crystal violet for 10-20 min, the cell aggregates with > 50 cells were scored as a colony. Colony forming efficiency was expressed as: Colony forming efficiency = (colony number of drug-treated cells/cell population) × 100% 47 . Three independent experiments were performed. AO/EB staining apoptotic cells. Both U251and WJ1 were seeded in 96-well plates (2000/well), and treated with different concentrations of KuA (40, 60 and 80 μ g/ml for U251; 10, 20 and 30 μ g/mL for WJ1) for 48 h respectively. The untreated and treated cells were stained with AO/EB dye mix (100 μ g/mL acridine orange and 100 μ g/mL ethidium bromide) for 2 min to detect apoptotic cells 48 . The stained cells were visualized immediately, images were captured using a Leica DMI400B inverted fluorescence microscope linked to a DFC340FX camera, more than 500 cells were counted for each sample. Three independent experiments were performed. Wound healing assay. GBM cells (4 × 10 5 cells/ml) were seeded in 12-well plates, incubated overnight and reached confluent monolayers for wounding. Wounds were made with a 10 μ L pipette tip, various concentrations of KuA was added to monolayer GBM cells. Photographs were taken immediately (time 0) and 24 h after wounding for GBM cells under microscope (Nikon ECLIPSE Ti-U, Japan). The distances migrated by the monolayer cell to close the wounded area during this time course were measured using Image-Pro Plus 6.0 software. Results were expressed as a migration index-that is, the distance migrated by KuA treated cells relative to the distance migrated by the untreated cells 49 . Three independent experiments were performed. Cell invasion assay. Cell invasion assay was performed in 24-transwell chambers (Corning, Costar, USA), with 8 μ m pores 49 . GBM cells were treated with different concentrations of KuA for 24 h, and harvested. 4 × 10 4 cells in 200 μ L serum-free medium were plated on the upper chamber, the lower chamber contained 500 μ L of complete medium containing 15% FBS as a chemo-attractant. After 24 h incubation, cells on the upper surface of the membrane were scrubbed, and the cells that penetrated through the filter were fixed with methanol, stained with 0.1% crystal violet. Images were taken under microscope (Nikon ECLIPSE Ti-U, Japan), the penetrated cells in 5 non-overlapping random fields per well were counted. Three independent experiments were performed.

GBM growth inhibition test in vivo.
Animal care and experiments were conducted according to the guidelines approved by the Institutional Animal Care and Use Committee of Sichuan University. 32 five-week-old male nude mice (BALB/C-nu/nu) were inoculated with 2 × 10 6 human GBM cells (WJ1). One week later, the tumor bearing mice were randomized into 4 groups and each group containing 8 mice, three test group mice were administrated with 10, 20 and 40 mg/kg of KuA i.p., 5 times weekly for 4 weeks, respectively; the control mice were injected with the same volume of solvent; During the experimental process, mice were weighted, and tumor volumes were measured every two days, tumor volumes were calculated using a standard formula (length × width 2 × 0.5) 50 . At the end of experiment, the mice were sacrificed by carbon dioxide asphyxiation and dissected; the tumors were taken out and weighed. The tumor inhibitory rates were calculated using the formula: tumor inhibitory rate (%) = (mean tumor weight of the control group − mean tumor weight of the treated group) ÷ mean tumor weight of the control group × 100%. All animals were maintained under standard conditions according to the guidelines of the Institutional Animal Care and Use Committee of Sichuan University.

Statistical analysis.
All values are expressed as means ± SD (standard deviation) and analyzed using the SPSS software (version 19.0) and analysis of variance (ANOVA), followed by Dunnett's test for pairwise comparison. Statistical significance was defined as p < 0.05 for all tests.