Oncogenic Ras is downregulated by ARHI and induces autophagy by Ras/AKT/mTOR pathway in glioblastoma

Human glioblastoma cell lines (LN229, T98G, U87, U251) and human brain astrocyte cell line NHA were obtained and authenticated from the China Infrastructure of Cell Line Resource at July 2017. All cell lines were recently authenticated by STR analysis and tested for mycoplasma contamination. LN229 (CRL-2611), T98G (CRL-1690), and U87 (HTB-14) were distributed by ATCC. CQ was purchased from Sigma Aldrich (St. Louis, MO, USA). 3-MA were purchased from MedChem Express (MCE, USA).

A total of 9 glioma and 3 normal brain samples were obtained from the First Affiliated Hospital of Harbin Medical University between 2012 and 2014. The patients’ clinical characteristics, such as age, gender, and WHO grade, were collected. For qRT-PCR and western blot analysis, tissues were immediately frozen in liquid nitrogen.

The ARHI plasmids and lentiviruses used for ARHI overexpression and the negative control plasmids and lentiviruses were purchased from Genechem Co., Ltd. (Shanghai, China). Before transfection, glioma cells were cultured in six-well plates at 5 × 10⁴ cells per well. Then, ARHI lentiviruses were introduced into glioma cells treated with 8 μg/ml polybrene (Genechem). Transfection effects were observed by a fluorescence microscope after 48 h. Puromycin was used to purify the infected cells.

Total RNA in cells and tissues was extracted using Trizol reagent, and the concentration of these RNA samples was measured using a spectrophotometer (Thermo Scientific™ NanoDrop 2000c). The RNA specimens were reverse transcribed into cDNA using a PrimeScript RT reagent kit (ToYoBo). The primer sequences for ARHI were forward 5′-CATAAGTTCCCCATCGTGC-3′; and reverse 5′-GAACAGCTCCTGCACATTCA-3′, and those for GAPDH used as a standard were forward 5′-ACCACAGTCCATGCCATCAC-3′; and reverse 5′-TCCACCACCCTGTTGCTGTA-3′.

The tissues and cells were lysed by RIPA buffer (Beyotime Institute of Biotechnology, Beijing, China) containing phenylmethylsulfonyl fluoride (PMSF) and a phosphatase inhibitor. The equal amounts of lysates were separated by 8–12.5% SDS-PAGE gels and transferred onto PVDF membranes. Antibody SQSTM1 (#23214), LC3B(#3868), Ras (#8955), Cleaved Caspase3 (#9662), mTOR (#2972), AKT (#9272) and Beta actin (#3700) were purchased from Cell Signaling Technology, ARHI antibody (ab107051) was purchased from abcam.

A density of 5000 cells were planted into 96-well plates for each well and mixed 10 μl MTT at 72 h. After 24 h of transfection with the ARHI and control plasmid, cells (500 cells/well) were seeded into a six-well plates and cultured for two weeks. Then, 0.1% crystal violet was used to stain clones, and cells were photographed using a ChemiDocTM MP system (Bio-Rad, USA).

Cells were plated on coverslips, and after treatment, cells were washed twice with phosphate-buffered saline (PBS). They were then fixed with 4% paraformaldehyde for 20 min, blocked with 5% goat serum at 37 °C for 30 min and then treated with 0.3% Triton X-100 for 10 min. Next, 50 μl TUNEL reaction mix (Wanleibio, WLA030a) was added to each sample, and cells were incubated for 60 min at 37 °C in the dark. They were then incubated with DAPI (Beyotime, C1005), and photographs were captured by using a FSX100 Bio Imaging Navigator system (Olympus, Japan).

GBM cells were seeded onto 6-well plates at a density of 1 × 10⁵ cells per well. After treatment, cells were harvested and fixed using 2.5% glutaraldehyde at 4 °C overnight. Then, the harvested cells were dehydrated with ethanol and acetone and fixed with 1% osmium tetroxide. Samples were observed and captured by using transmission electron microscope (Hitachi H-7650, Japan).

A total of 5 × 10⁴ cells were seeded onto 35-mm glass-bottomed dishes. After treatment, cells were harvested and fixed with 4% paraformaldehyde. Then, cells were treated with 0.1% Triton X-100 and blocked with 5% bovine serum albumin (BSA, BOSTER, AR0004). Cells were incubated with primary antibody at 4 °C overnight and then incubated with Alexa Flour 594 AffiniPure goat anti-rabbit IgG (ZSGB-BIO, ZF-0516) and DAPI (Beyotime, C1005).

After transfecting ARHI for 24 h, cells were seeded onto a six-well plate and fixed with 3-MA or CQ. After 48 h, cells were harvested and detected using an Annexin V-PE Apoptosis Detection Kit (BD Bioscience, 556,422). Then, samples were measured by an Accuri C5 flow cytometer (BD Bioscience).

Cells were transfected with GFP-RFP-LC3 lentivirus according to the manufacturer’s protocol. After treatment, the autophagosomes (yellow dots) and autolysosomes (red dots) were determined by augmented microscopy (BioTek Instruments, USA).

All 28-day-old (15-20 g) BALB/C nude mice were purchased from the Vital River Animal Center (Beijing, China) randomly divided into four groups and each group have 6 mice. The required sample sizes were calculated and tabulated with different levels of accuracies and marginal errors with 95% confidence level for estimating and for various effect sizes with 80% power for purpose of testing as well. LN229 cells (5 × 10⁶ ARHI-LN229 cells was subcutaneously injected into the right hips of each mouse. In drug treatment group, the mice were injected with CQ (30 mg/kg/d) for 4 weeks. Tumor volume was assessed and calculated (volume = (width)² × (length)/2) every 2 days. Mice were sacrificed on day 34, all mice had no panic, struggling, shouting. The procedure is as follows: After using Avertin (2, 2, 2-Tribromoethanol) anesthesia (the injection dose should reach lethal dose:> 0.5 mg/20 g, intraperitoneal injection), the heartbeat、breath and various reflexes of mice disappeared. No painful reaction was observed in the whole course of anesthesia. Then, mice were sacrificed by breaking neck for next step. After photography, tumor weight was measured. All experimental ethics and animal experiments were conformed to the European Parliament Directive (2010/63/EU) and were approved by the Institutional Animal Care and Use Committee at Harbin Medical University (No. HMUIRB-2008-06).

Formalin-fixed and using paraffin embedded samples were sliced into 5-μm-thick sections. Then, sample sections were immunostained using primary antibodies Ki67, Ras, cleaved caspase-3 and ARHI at 4 °C overnight. Then, applying secondary antibodies at 37 °C for 30 min. Next, samples were visualized according to manufacturer’s protocol and using a diaminobenzidine (DAB) substrate kit for 10 min. After intensive washing, samples were counterstained with hematoxylin, dehydrated and coverslipped. Obtaining pictures by using FSX100 Bio Imaging Navigator (Olympus, Japan).

Three independents experiments data are shown as the means ± standard deviations (SD). Two groups statistical analysis was performed using Student’s t test (two-tailed), *indicated a statistical significance of P-value< 0.05; **indicated a strong statistical significance of P-value< 0.01; *** indicated an even strong statistical significance of P-value< 0.001.

We used RT-PCR and western blot to detect the expression levels in glioma tissues and normal brain tissue (Fig. 1a-b, Additional files 1 and 2). The results showed that glioma tissues had lower ARHI expression, and higher grades of glioma corresponded to lower ARHI protein expression (T1-T3 WHO IV, T4-T6 WHO III, T7-T9 WHO II, N1-N3 normal brain tissues). We also measured ARHI expression in the human glioma cell lines LN229, T98G, U251, and U87 and human brain astrocyte cell line NHA (Fig. 1c-d). Normal human brain astrocyte cell line had higher expression levels of ARHI than the four glioma cell lines. To examine the effect of ARHI on the viability of human glioma cells, we chose the lower expression cell lines LN229 and T98G to investigate the effect of over-expressing ARHI on glioma cell clonogenicity and proliferation. In Fig. 1e-f, colony formation was significantly decreased after over-expression of ARHI. We established stably infected LN229-ARHI and T98G-ARHI cells, cell proliferation was measured by Ki67 immunofluorescence dyeing. The proliferative marker Ki67 was decreased after over-expression of ARHI (Fig. 1g-h). These results indicate that ARHI induces cytotoxicity and inhibits cell growth in glioma cells and the similar phenomena were found in glioma stem cells (Additional file 3).

Increasing evidence has shown that autophagy plays an important role in anticancer therapies [12]. ARHI has been reported to induce autophagy in many cancer cells, including ovarian cancer and breast cancer. Therefore, we questioned whether over-expression of ARHI can induce autophagy in glioma cells. After over-expressing ARHI, we used transmission electron microscopy to detect autophagic vacuoles. We found massive autophagosomes in the cell cytoplasm of the ARHI group compared with the negative control group (Fig. 2a-b), indicating that ARHI can induce autophagy in glioma cells. Abundant cytoplasmic protein LC3-I is an that is cleaved and lapidated and forming LC3-II, translocating to and associating with the autophagosome in a punctate pattern. Autophagy thus enable the cell to eliminate and recycle proteins or organelles to sustain metabolism and can be recognized in part by formation of LC3-II punctate. We using LC3B immunofluorescence stain to detect LC3B puncta (Fig. 2c-d) and over-expressing ARHI can dramatically increase LC3B puncta formation. SQSTM1 is widely used as a marker for autophagic degradation, as it is binding to LC3 and selectively degraded in autolysosomes. A smooth autophagy flux is associated with decreased levels of SQSTM1 [13]. Similarly, we found ARHI significantly induced autophagy with increased levels of LC3-II and decreased levels of SQSTM1. This western blot result showed that ARHI can enhance autophagy flux in glioma cells (Fig. 2e-f). In patients derived glioma cells, we got the same results (Additional file 4).

To investigate whether ARHI-induced glioma cell growth arrest depends on autophagy, we used shATG5 plasmids to infect LN229-ARHI and T98G-ARHI cells. ARHI expression significantly inhibited glioma cell growth within 72 h but exhibited little growth-inhibiting effect on ARHI-shATG5 cells (Fig. 3a-b), suggesting that ARHI induces autophagy-mediated glioma cell death. As ARHI is a member of the Ras superfamily, and it encodes a 26 KDa protein that has 50–60% homology to Ras [8]. Ras can regulate autophagy by indirectly enhancing AKT and down regulating mTOR phosphorylation levels. A previous study reported that ARHI induces autophagy in ovarian cancer cells by inhibiting Ras/MAP signaling [13]. Therefore, we questioned whether ARHI induces autophagy in glioma cells by regulating Ras. To further investigate the molecular mechanism of ARHI-associated autophagy in glioma, we used western blotting to detect Ras after over-expressing ARHI in glioma cells. We found that ARHI can decrease Ras expression and inbibit the activation of AKT and mTOR, then, promoting autophagy in glioma cells.

Recent reports have indicated that CQ can strengthen the autophagy inducer’s toxicity in cancer cells mainly because CQ can reduce/increase the apoptosis regulatory protein Bcl-2/cleaved caspase3, thus promoting apoptosis in cancer cells [14]. ARHI as an autophagy inducer, and it is unknown whether ARHI has synergistic action with CQ. Autophagic flux are monitored by GFP-RFP-LC3 fluorescence assays. The GFP signal can be digested in the lysosomes’ acidic conditions. Contrary to GFP, RFP can exist in acidic condition stably. Therefore, the stage of without lysosome fusion was indicated by the colocalization fluorescence of GFP and RFP, which suggests autophagosomes shown as yellow dots. In contrast, the RFP signal indicates autolysosomes shown as red dots [15, 16]. GFP-RFP-LC3 fluorescence assays showed that more yellow dots were aggregated in ARHI-LN229 cells after 48 h of exposure to CQ (3 μM). Studies have shown that 3-methyladenine (3-MA) can inhibit the formation of beclin1-PtdIns 3KC3 compounds, thus indirectly suppressing autophagosome formation. Thus, 3-MA can block the early stage of autophagy [17, 18]. In the 3-MA group (2 mM), the little yellow fluorescence was observed, and red dots were aggregated in ARHI-LN229 cells (Fig. 4a-b). This result indicated that CQ can inhibit the late stage of autophagy and that blockage of the late stage of autophagy is associated with increased levels of substrate proteins, including toxic proteins. 3-MA inhibited the early stage of autophagy but did not induce this effect. As we know, transmission electron microscopy (TEM) is the golden standard for detecting autophagosomes. In Fig. 4c, we use TEM for further observing autophagosomes. In ARHI-lN229 cells, initial autophagic vacuoles (AVi) were predominated and we found after adding CQ in ARHI-lN229 cells, more late autophagy vacuoles (AVd) were accumulated in the cytoplasm. Also, as the accumulation of AVd, and the accumulation of AVd may promote apoptosis in ARHI-lN229 cells. The morphological changes in CQ-treated ARHI-LN229 cells were examined with a TUNEL assay, and more apoptotic cells were detected in the CQ-treated ARHI-LN229 cells (Fig. 4d).

Inhibition of the late stage of autophagy can cause the accumulation of autophagic vacuoles and robust apoptosis [18‐20]. In our research, we found that using CQ to inhibit late stage of autophagy can cause accumulation of autophagic vacuoles, increased the cytotoxicity of ARHI and raising apoptosis level in glioma cell. An MTT assay was used to determine the effect of different autophagy inhibitors, CQ and 3-MA, on the viability of the human glioma LN229 cells and ARHI-LN229 cells. As shown in Fig. 5a-b, the results indicated that 3-MA (2 mM) attenuated ARHI-induced autophagy-mediated glioma cell death, while CQ (3 μM) significantly enhanced the cytotoxicity of ARHI. In the western blot results, we found that either CQ or 3-MA could block the autophagic flux with increased levels of SQSTM1 (Fig. 5c-d). However, in the 3-MA groups, the reduced level of LC3-II and a reduced number of autophagic vacuoles indicated that 3-MA blocked ARHI-induced autophagy. Contrary to 3-MA, CQ treatment led to an increase in the level of LC3-II and the accumulation of autophagic vacuoles induced by ARHI. These results suggested that CQ enhanced the cytotoxicity induced by ARHI, partly because of the accumulation of autophagic vacuoles [21‐23]. The accumulation of autophagic vacuoles can promote cancer cell death and apoptosis [24‐28]. Thus, we questioned whether CQ enhanced the cytotoxicity of ARHI in glioma cells partly by promoting apoptosis. Furthermore, we evaluated cleaved caspase-3 and Bcl-2 protein expression levels and the percentage of apoptotic cells. The results showed that both the expression of apoptosis-related cleaved caspase-3 protein and the percentage of Annexin V-positive cells were increased in the CQ-treated ARHI-LN229 group (Fig. 5e-h). However, after treatment with 3-MA, the apoptotic levels in ARHI-LN229 cells were decreased. Taken together, CQ can promote ARHI-LN229 cell death partly by increasing its apoptosis levels.

Next, we assessed the efficacy of CQ-associated ARHI therapy in vivo. Either autophagosomes or apoptosis levels were increased after CQ-associated ARHI therapy and tumor growth was significantly suppressed. As shown in Fig. 6a, after CQ-associated ARHI treatment, the suppression of tumor growth was observed by gross inspection. By the way, body weight showed no significant difference among the four groups over four weeks (Fig. 6c). The therapeutic effect was assessed by tumor weight and volume. CQ-associated ARHI significantly inhibited tumor growth, which resulted in decreased tumor weight and volume in this group (Fig. 6b, d). In addition, elevated levels of cleaved caspase-3, LC3B and decreased levels of Ki67 and Ras were observed in the CQ-associated ARHI treatment group compared with the control group (Fig. 6e, Additional file 5). Taken together, these data support the hypothesis that CQ-associated ARHI exhibits anti-glioma properties in vivo. At last, we use TEM to detect autophagy markers. Compared with control, in ARHI group, more initial autophagy vacuoles (AVi) were observed in cytoplasm and in the group of CQ-associated ARHI, more AVd were accumulated in cytoplasm. All these indicated ARHI can also induce autophagy in vivo and that there were more autophagic vacuoles (almost all late autophagic vacuoles) accumulated in the CQ-associated group (Fig. 6f).