Vitamin K2 promotes PI3K/AKT/HIF-1α-mediated glycolysis that leads to AMPK-dependent autophagic cell death in bladder cancer cells

Vitamin K2 has been shown to exert remarkable anticancer activity. However, the detailed mechanism remains unclear. Here, our study was the first to show that Vitamin K2 significantly promoted the glycolysis in bladder cancer cells by upregulating glucose consumption and lactate production, whereas inhibited TCA cycle by reducing the amounts of Acetyl-CoA. Moreover, suppression of PI3K/AKT and HIF-1α attenuated Vitamin K2-increased glucose consumption and lactate generation, indicating that Vitamin K2 promotes PI3K/AKT and HIF-1α-mediated glycolysis in bladder cancer cells. Importantly, upon glucose limitation, Vitamin K2-upregulated glycolysis markedly induced metabolic stress, along with AMPK activation and mTORC1 pathway suppression, which subsequently triggered AMPK-dependent autophagic cell death. Intriguingly, glucose supplementation profoundly abrogated AMPK activation and rescued bladder cancer cells from Vitamin K2-triggered autophagic cell death. Furthermore, both inhibition of PI3K/AKT/HIF-1α and attenuation of glycolysis significantly blocked Vitamin K2-induced AMPK activation and subsequently prevented autophagic cell death. Collectively, these findings reveal that Vitamin K2 could induce metabolic stress and trigger AMPK-dependent autophagic cell death in bladder cancer cells by PI3K/AKT/HIF-1α-mediated glycolysis promotion.


Vitamin K2 promotes glycolysis and inhibits TCA cycle in bladder cancer cells.
To evaluate the effect of Vitamin K2 on the glucose metabolism in bladder cancer cells, the glycolytic process was examined. In the present study, Vitamin K2 remarkably promoted the glucose consumption and lactate production in bladder cancer T24, EJ and J82 cells in a time or dose-dependent manner (Fig. 1A-D and Supplementary Fig. 1). Moreover, Western blot analysis showed that Vitamin K2 stepwise elevated the expression of some glycolytic proteins or enzymes, such as GLUT-1, Hexokinase II (HK2), PFKFB2, LDHA and PDHK1, in bladder cancer T24 (Fig. 1E) and EJ cells (Fig. 1F). These results indicate that Vitamin K2 promotes the glycolysis in bladder cancer cells. We next assessed the effect of Vitamin K2 on the tricarboxylic acid (TCA) cycle in bladder cancer cells. Comparing to the control group, Vitamin K2 remarkably decreased the amounts of Acetyl coenzyme A (Acetyl-CoA) in T24 cells (Fig. 1G). Furthermore, pyruvate dehydrogenase (PDH) was notably phosphorylated in Vitamin K2-treated T24 cells, suggesting that Vitamin K2 inactivates PDH (Fig. 1H). To determine the inhibitory effect of Vitamin K2 on TCA cycle in bladder cancer cells, the contents of key metabolites were detected. As shown in Fig. 1I,J, Vitamin K2 markedly reduced the amounts of metabolites in TCA cycle, compared with the control group, implying that Vitamin K2 suppresses the TCA cycle in bladder cancer cells. Taken together, our results reveal that Vitamin K2 renders bladder cancer cells more dependence on glycolysis than TCA cycle. Activation of PI3K/AKT and HIF-1α contributes to Vitamin K2-upregulated glycolysis in bladder cancer cells. We next investigate whether activation of PI3K/AKT and HIF-1α is implicated in Vitamin K2-upregulated glycolysis in bladder cancer cells. As shown in Fig. 2A, Vitamin K2 significantly induced the phosphorylation of PI3K/AKT, and increased HIF-1α expression in bladder cancer T24 cells in a time-dependent manner. Moreover, both YC-1 (a pharmacological inhibitor of HIF-1α) and HIF-1α siRNA significantly attenuated Vitamin K2-elevated glucose consumption and lactate production in T24 cells ( Fig. 2B-G). To validate the involvement of PI3K/AKT activation in Vitamin K2-induced glycolysis promotion, MK2206 (a specific inhibitor of AKT) and LY294002 (a typical PI3K inhibitor) were respectively utilized. Our data showed that both MK2206 and LY294002 markedly alleviated Vitamin K2-increased glucose consumption and lactate production in T24 ( Fig. 2H-J,N-P) and EJ cells ( Supplementary Fig. 2). Additionally, Vitamin K2-upregulated glucose consumption and lactate production were significantly inhibited by respective treatment with AKT siRNA and PI3K siRNA ( Vitamin K2 induces metabolic stress in bladder cancer cells. We next investigate whether Vitamin K2-upregulated glycolysis results in metabolic stress, upon glucose limitation. Our results showed that glucose in T24 cell culture medium was almost depleted, and lactate generation attained the maximum after exposed to Vitamin K2 for 18 hours (Fig. 3A,B). To determine the metabolic stress resulted from Vitamin K2-elevated glycolysis in bladder cancer cells, we assessed the NADH:NAD + ratio, cyclic AMP content and c-MYC expression whose alteration are susceptible to glucose starvation 32,33 . As shown in Fig. 3C, both the ratio of NADH:NAD + and content of cyclic AMP were dramatically decreased in T24 cells after exposed to Vitamin K2 for 18 hours. Moreover, c-MYC protein level was also significantly reduced in T24 cells following treatment with Vitamin K2 for 18 hours (Fig. 3D). In addition, As shown in Fig. 3E, the ratios of AMP:ATP, ADP:ATP and GDP:GTP were markedly increased in T24 cells after exposed to Vitamin K2 for 18 hours. These results suggest that Vitamin K2 is able to induce metabolic stress, including glucose starvation and energy shortage, in bladder cancer cells, upon glucose limitation. www.nature.com/scientificreports www.nature.com/scientificreports/ Vitamin K2 triggers autophagic cell death in bladder cancer cells. To investigate whether Vitamin K2-induced metabolic stress subsequently triggers autophagic cell death, the autophagic and apoptotic effect of Vitamin K2 on bladder cancer cells were respectively evaluated. Our data showed that Vitamin K2 significantly upregulated the expression of P62, Beclin-1, LC3B II, and induced the cleavage of Caspase-9 and Caspase-3 in T24 (Fig. 4A) and EJ cells ( Supplementary Fig. 3B) in a time-dependent manner. Moreover, the autophagosomes and prominent punctate LC3 were observed in Vitamin K2-treated T24 cells ( Fig. 4B and Supplementary (G) T24 cells were exposed to 50 μM Vitamin K2 for 12 hours and the quantitative analysis of Acetyl-CoA amount was measured by LC-MS. (H) Western blot showed the effect of 50 μM Vitamin K2 on the phosphorylation of PDH in T24 cells. (I) and (J) T24 cells were exposed to 50 μM Vitamin K2 for 12 hours and the quantitative analysis of the metabolites content in TCA cycle was measured by LC-MS. Data are presented as the mean ± SD of at least three independent experiments. *P < 0.05, **P < 0.01 and ***P < 0.001. www.nature.com/scientificreports www.nature.com/scientificreports/ Fig. 3A). In addition, flow cytometry showed that the apoptotic rate was remarkably increased in T24 (Fig. 4C) and EJ cells ( Supplementary Fig. 3C) after extended exposure to Vitamin K2. To determine the autophagic cell death, we evaluated the effect of autophagy on apoptosis in Vitamin K2-treated bladder cancer cells, and chloroquine (CQ) and 3-Methyladenine (3-MA), two typical autophagic inhibitors, were respectively employed. As shown in Fig. 4D-G, both CQ and 3-MA markedly suppressed autophagy, and blocked apoptosis in Vitamin K2-treated T24 cells. Furthermore, Rapamycin, a specific autophagic agonist, significantly activated autophagy, and enhanced Vitamin K2-triggered apoptosis in T24 cells ( Supplementary Fig. 3D,E). These results indicate that Vitamin K2-triggered autophagy promotes the subsequent apoptosis, and ultimately results in autophagic cell death in bladder cancer cells, upon metabolic stress.
Vitamin K2 triggers AMPK-dependent autophagic cell death in bladder cancer cells. Given that AMPK maintains energy homeostasis in response to energy shortage or metabolic stress, we next investigated whether AMPK was activated due to Vitamin K2-induced metabolic stress in bladder cancer cells. Our results showed that Vitamin K2 markedly increased the phosphorylation of AMPK, ACC1 (Acetyl-CoA-Carboxylase 1) and ULK1 (unc-51-like autophagy-activating kinase 1), whereas reduced the phosphorylation of mTORC1 (mammalian target of rapamycin complex 1) and P70S6K (one of the downstream of mTORC1), in T24 ( Fig. 5A) and EJ cells ( Supplementary Fig. 4A). These results suggest that Vitamin K2 could activate AMPK pathway and conversely suppress mTORC1 pathway in bladder cancer cells in response to metabolic stress. To investigate whether Vitamin K2-triggered autophagic cell death are AMPK dependent, the inhibition of AMPK was performed. Our data showed that both AMPK siRNA and Compound C (C.C, a pharmacological inhibitor of AMPK) were able to abrogate the activation of AMPK pathway and simultaneously restore mTORC1 pathway, which subsequently suppressed autophagic cell death in Vitamin K2-treated bladder cancer cells ( Fig. 5B-E). These results reveal that Vitamin K2 could trigger AMPK-dependent autophagic cell death in bladder cancer cells, upon metabolic stress. To determine whether Vitamin K2-induced metabolic stress is responsible for AMPK-dependent autophagic cell death in bladder cancer cells, glucose supplementation was performed to prevent metabolic stress. Intriguingly, glucose supplementation completely restored c-MYC expression, abrogated AMPK pathway activation and markedly rescued bladder cancer cells from autophagic cell death ( Fig. 5F-H and Supplementary Fig. 4B,C). Taken together, these results indicate that Vitamin K2 is able to trigger AMPK-dependent autophagic cell death in bladder cancer cells, upon metabolic stress.

Activation of PI3K/AKT and HIF-1α is crucial for Vitamin K2-triggered AMPK-dependent autophagic cell death in bladder cancer cells.
To investigate whether activation of PI3K/AKT and HIF-1α is involved in Vitamin K2-triggered AMPK-dependent autophagic cell death, the inhibition of PI3K/ AKT and HIF-1α were respectively performed. Our data showed that both HIF-1α siRNA and YC-1 significantly blocked the activation of AMPK pathway, and inhibited the autophagic cell death in Vitamin K2-treated T24 cells ( Fig. 6A-D), suggesting that Vitamin K2-increased HIF-1α expression is greatly responsible for AMPK-dependent autophagic cell death in bladder cancer cells. Moreover, both HIF-1α elevation and AMPK-dependent autophagic cell death induced by Vitamin K2 were remarkably suppressed by respective treatment with AKT siRNA (     www.nature.com/scientificreports www.nature.com/scientificreports/ K2 triggers AMPK-dependent autophagic cell death in bladder cancer cells via PI3K/AKT and HIF-1α signal pathways.

Attenuation of glycolysis prevents Vitamin K2-triggered AMPK dependent autophagic cell death in bladder cancer cells. To validate that AMPK-dependent autophagic cell death is due to
Vitamin K2-induced glycolysis promotion, 2-Deoxy-D-glucose (2-DG), Sodium dichloroacetate (DCA) and 3-Bromopyruvic acid (3-BP), three typical glycolytic inhibitors, were employed. As shown in Fig. 7A,B and Supplementary Fig. 6A,B, 2-DG, 3-BP and DCA respectively attenuated Vitamin K2-upregulated glycolysis, and reversed the decreased cell viability in T24 and EJ cells. Furthermore, Vitamin K2-triggered apoptosis in T24 and EJ cells was profoundly abolished by respective treatment with 2-DG and 3-BP (Fig. 7C). Concordantly, western blot showed that Vitamin K2-induced activation of AMPK, autophagy and apoptosis were significantly blocked by treatment with 2-DG, 3-BP and DCA, respectively (Fig. 7D-G and Supplementary Fig. 6C). These results indicate that Vitamin K2-upregulated glycolysis accounts for AMPK-dependent autophagic cell death in bladder cancer cells.

Vitamin K2 triggers AMPK-dependent autophagic cell death in vivo. To determine whether
Vitamin K2 is able to induce AMPK-dependent autophagic cell death in vivo, bladder cancer xenografts were established in nude mice. As shown in Fig. 8A, comparing with the sustained growth of tumor in vehicle group, treatment with Vitamin K2 at dose of 10 and 20 mg/kg gradually reduced the tumor sizes. Moreover, during drug administration, the weights of nude mice treated by Vitamin K2 were almost intact, compared with the significant decrease in vehicle group (Fig. 8B). Importantly, treatment with Vitamin K2 markedly extended the mice survival, compared with 100% death of vehicle group after administration for 58 days (Fig. 8C). To investigate whether the inhibitory growth effect of Vitamin K2 on bladder cancer cells in vivo is due to the induction of AMPK-dependent autophagic cell death, we excised the tumors from the mice, sectioned for immunohistochemistry (IHC) and lysed for western blot analysis. As shown in Fig. 8D, Vitamin K2 dose-dependently upregulated the expression of P62, Beclin-1 and LC3B II and activated Caspase-3 in tumor sections. Besides, the increased expression of GLUT-1, HIF-1α, p-AKT and p-AMPK were also detected in Vitamin K2-treated tumor group (Fig. 8E). Similar results were obtained from western blot analysis (Fig. 8F). Taken together, these results suggest that Vitamin K2 might promote the glycolysis and trigger AMPK-dependent autophagic cell death in bladder cancer cells in vivo.

Discussion
This study was the first to show that Vitamin K2 could remarkably promote the glycolysis, whereas suppress TCA cycle in bladder cancer cells. Importantly, under glucose limited conditions, Vitamin K2-upregulated glycolysis could significantly induce metabolic stress in accompany with AMPK activation and mTORC1 pathway suppression, which subsequently triggers AMPK-dependent autophagic cell death in bladder cancer cells. In addition, following the inhibition of PI3K/AKT and HIF-1α or attenuation of glycolysis, the above phenotypes that Vitamin K2 induced metabolic stress and triggered AMPK-dependent autophagic cell death were markedly blocked, indicating that Vitamin K2 induces metabolic stress and triggers AMPK-dependent autophagic cell death via PI3K/AKT/HIF-1α activation and glycolysis promotion. In summary, the major novelty in this study is that Vitamin K2 is able to induce autophagic cell death in bladder cancer cells by establishing the metabolic signal axis: PI3K/AKT/HIF-1α dependent glycolysis promotion-metabolic stress-AMPK/ULK1 activation-mTORC1 pathway suppression (Fig. 8G).
Aerobic glycolysis is commonly hyper-activated in many cancer cells. The upregulated glycolysis in cancer cells profoundly provides rapid ATP generation and metabolic intermediates for the production of proteins, lipids and nucleotides, which meets the requirement for cell growth and metastasis. Traditional anticancer drugs are often reported to curb the growth of cancer cells via glycolysis blockade [34][35][36][37] . Different from the above anticancer mechanisms, our study showed that Vitamin K2 markedly promoted the glycolysis in bladder cancer cells, which induced metabolic stress and subsequently triggered AMPK-dependent autophagic cell death, upon glucose limitation. 2-DG, 3-BP and DCA are three typical glycolytic inhibitors that have been shown to exhibit excellent inhibitory growth effect on cancer cells by blocking the glycolytic pathway [38][39][40] . In this study, the above three glycolytic inhibitors were utilized to verify the hypothesis that Vitamin K2 induces metabolic stress and triggers AMPK-dependent autophagic cell death via glycolysis promotion. As expected, 2-DG, 3BP and DCA-induced glycolysis attenuation significantly prevented metabolic stress and rescued bladder cancer cells from Vitamin K2-triggered AMPK-dependent autophagic cell death, implying that Vitamin K2-increased glycolysis contributes to metabolic stress and AMPK-dependent autophagic cell death in bladder cancer cells.
Recent studies have revealed that hyper-activation of PI3K/AKT and HIF-1α in cancer cells plays a central role in glycolysis promotion. In our study, inhibition of PI3K/AKT and HIF-1α notably attenuated Vitamin K2-upregulated glycolysis, indicating that Vitamin K2 promotes glycolysis in bladder cancer cells via PI3K/ AKT and HIF-1α signal pathways. Accumulating evidences suggest that Reactive oxygen species (ROS) is able to activate PI3K/AKT pathway by inactivating some protein phosphatase, such as protein tyrosine phosphatase 1B (PTP1B), protein phosphatase 2 A (PP2A) and phosphatase and tensin homologue (PTEN) lipid phosphatase 41 . In this study, we also attempt to investigate the involvement of ROS in Vitamin K2-activated PI3K/ AKT pathway. As shown in Supplementary Fig. 7, N-acetyl cysteine (NAC, a ROS scavenger) not only alleviated Vitamin K2-induced AKT activation and glycolysis promotion, but also significantly suppressed the subsequent AMPK-dependent autophagic cell death. These results reveal that Vitamin K2-induced ROS generation activates AKT and subsequently promotes the glycolysis, resulting in AMPK-dependent autophagic cell death in bladder cancer cells.

Scientific RepoRtS |
(2020) 10:7714 | https://doi.org/10.1038/s41598-020-64880-x www.nature.com/scientificreports www.nature.com/scientificreports/ Upon glucose limitation, the increased glycolysis will inevitably result in metabolic stress, which subsequently leads to cancer cell death. In our study, Vitamin K2-upregulated glycolysis significantly induced metabolic stress, including glucose starvation and energy shortage, and triggered AMPK-dependent autophagic cell www.nature.com/scientificreports www.nature.com/scientificreports/ death in bladder cancer cells. To validate whether Vitamin K2-induced metabolic stress results in the subsequent AMPK-dependent autophagic cell death, the glucose supplementation was performed. As shown in Fig. 5F-H, glucose supplementation not only restored c-MYC expression, but also rescued bladder cancer cells from Vitamin K2-triggered AMPK-dependent autophagic cell death, suggesting that metabolic stress induced by Vitamin K2 indeed triggers AMPK-dependent autophagic cell death in bladder cancer cells, under glucose limited condition.
AMPK is an energy sensing protein which is rapidly activated upon metabolic stress. Once activated, AMPK could repress the anabolism and simultaneously stimulate the catabolic process including autophagy to maintain intracellular energy homeostasis 16,42 . In line with the previous studies, our results showed that Vitamin K2 remarkably activated AMPK pathway, and triggered AMPK-dependent autophagy, in response to metabolic stress. Furthermore, apoptosis induced by Vitamin K2 was profoundly suppressed by respective treatment with CQ and 3MA, two typical autophagic inhibitors, suggesting that Vitamin K2-induced AMPK-dependent autophagy promotes the subsequent apoptosis in bladder cancer cells, upon metabolic stress. The mTORC1 pathway plays a pivotal role in the growth and survival of cancer cells by sensing intracellular nutrients levels. Upon metabolic stress, AMPK negatively regulates mTORC1 through phosphorylating TSC2 or Raptor (a mTORC1-positive regulatory subunit), resulting in suppression of mTORC1 pathway. In our study, inhibition of AMPK by AMPK siRNA or Compound C significantly reversed Vitamin K2-inhibited mTORC1, implying that AMPK activation contributes to mTORC1 pathway suppression in Vitamin K2-treated bladder cancer cells, upon metabolic stress.
Our previous studies indicated that Vitamin K2 is able to induce Mitochondria-related apoptosis in bladder cancer cells via ROS-JNK/P38 MAPK signal pathways. In that study, bladder cancer cells were directly exposed to Vitamin K2 for 24 hour, and a great amount of intracellular ROS generation was induced, which results in JNK/ P38 MAPK activation and mitochondria-associated apoptotic cell death. By contrast, in this study, bladder cancer cells were time-dependently treated with Vitamin K2. At early stage, Vitamin K2 could promote glycolytic process in bladder cancer cells, which might prevent intracellular ROS accumulation. However, under glucose limited condition, the increased glycolysis inevitably resulted in metabolic stress, which augments ROS accumulation due to lack of glucose for sustained glycolysis. Therefore, we speculate that Vitamin K2-induced metabolic stress might trigger autophagy and apoptosis in bladder cancer cells by AMPK-ULK1 or ROS-JNK/P38 MAPK, two distinct signal pathways.
In conclusion, our study regarded Vitamin K2-upregulated glycolysis as a bridge to connect activation of PI3K/AKT and HIF-1α with metabolic stress, which triggers AMPK-dependent autophagic cell death in bladder cancer cells. Our study laid a solid foundation for unveiling the precise anticancer mechanism of Vitamin K2. In addition, the anticancer function of Vitamin K2 through glucose metabolic regulation might shed insight into starving cancer cells in the real tumor microenvironment. Glucose consumption and lactate production assays. Cells were seeded at a density of approximately 1×10 6 cells per well, and treated with Vitamin K2 for the indicated time. The glucose consumption and lactate production were respectively assessed using the glucose and lactate content detecting kits (E1010, Applygen) (A019-2, Nanjing Jiancheng) according to the manufacturer's protocols, and measured by microplate readers.

Reagents
Liquid chromatography-mass spectrometry (LC-MS) assays. T24 cells were seeded in 10cm-cell culture dishes and treated with 50 μM Vitamin K2 for the indicated time. For cellular metabolite extracting, T24 cells were washed with PBS and fixed with the metabolite extraction buffer consisting of methanol, acetonitrile, and water (2:2:1). Cells were then scraped and transferred into 1.5 ml centrifuge tubes and vortexed for 30 seconds. Samples were centrifuged at 14,000 × g and the supernatants were harvested for analysis. The metabolites were analyzed by LC-MS using a Agilent 1290 Infinity LC system coupled to a 5500 QTRAP mass spectrometer (AB SCIEX). Briefly, 4 μl of sample was injected on a column and heated at 45°C. The flow rate was set to 300 μl/ min. The gradient mobile phase was composed by water with 15 mM CH3CO2NH4 (A) and acetonitrile (B). The program with a linear gradient beginning at 90% (B) and decreasing to 40% (B) for 18 min was used and followed by washing and re-equilibration. The total running time is 23 min. The metabolites were analyzed by a 5500 QTRAP mass spectrometer under the anion mode conditions. The ion pairs were detected in the mode of MRM. The data processing was performed using Multiquant software.
Immunofluorescence. Cells were seeded in 6-well plates, and grown on the coverslips. After treatment with 50 μM Vitamin K2 for 14 hours, cells were washed with PBS and fixed with 4% paraformaldehyde, then incubated with LC3 primary antibody at 4°C for overnight, and secondary antibodies for 2 hours. The intracellular LC3 expression was observed under the fluorescence microscope (Olympus, Japan).
Electron microscopy. Cells were treated with 50 μM Vitamin K2 for 14 hours, washed with PBS, trypsinized, and collected. The cell pellets were fixed with 4% glutaraldehyde, embedded in the paraffin, and sectioned. The autophagosomes were observed under the transmission electron microscope (Olympus, Japan).
Cell viability analysis. Cells were seeded in 96-well plates at a density of approximately 1×10 5 cells per well, and treated by 50 μM Vitamin K2 combined with or without the correlative inhibitors for 18 hours. The cell viability was evaluated by the MTS assays in accordance with the manufacturer's protocol (G3580, Promega).
Apoptosis analysis. Cells were seeded in 12-well plates at a density of approximately 1×10 6 cells per well, and treated by 50 μM Vitamin K2 combined with or without the correlative inhibitors for 18 hours. To determine cell apoptosis, cells were collected, washed with PBS, and suspended in binding buffer. Cells were then stained with Annexin V-FITC and propidium iodide (PI) according to the instructions of an Annexin-V-FITC/PI apoptosis detecting kits (556547, BD Biosciences). The apoptotic cells were analyzed by the flow cytometer (Beckman coulter FC500).
Animal experiments. Female BALB/c nude mice (4-5 weeks old) were provided by experimental animal center (Tongji Medical college of Huazhong University of Science and Technology). All animal studies were strictly conducted in compliance with guidelines approved by both the Science and Technology Department of Hubei province and the Animal Experimentation Ethics Committee of Huazhong University of Science and Technology. Bladder cancer EJ cells (1×10 7 ) were subcutaneously injected into the right flank of nude mice. When the tumor sizes reached approximately 100 mm 3 , the tumor-bearing nude mice were intra-tumorally treated with 10 and 20 mg/kg Vitamin K2 each day, while intra-tumor injection of the equivalent volume of PBS as the vehicle group. For evaluating the changes of tumor sizes, the sliding caliper was used and the tumor sizes were measured by the formula: length × width 2 /2. For detecting the expression of the relevant proteins, mice were sacrificed and the tumors were excised, and lysed for western blot or sectioned for immunohistochemistry (IHC). Specific explanation for the absence of image of adequate length (Fig. 8F): In vivo experiment of western blot, the gels containing the separated protein were cropped, then transferred in PVDF, incubated with antibodies, and exposed in the films. Therefore, no strict fuller-length of gels was provided in our Supplementary information file, only the original cropped films with labels were displayed. Statistical analysis. The results were presented as the Mean ± SD for representative experiments with at least three time independent biological repeats, and analyzed using the GraphPad Prism 6 software. The statistical significance was measured using Student's t-test. The statistical significance in the figures was set at *p < 0.05, **p < 0.01, ***p < 0.001, ns: no significance.