Gene expression and survival analysis data were obtained from the Chinese Glioma Genome Atlas (http://www.cgga.org.cn; dataset ID: mRNAseq_693) (20,21) and The Cancer Genome Atlas (TCGA; http://cancergenome.nih.gov/) databases. Gene Ontology (GO) (22) and Kyoto Encyclopedia of Genes and Genomes (KEGG) (23) analyses were conducted by DAVID (https://david.ncifcrf.gov/gene2gene.jsp) (version 6.8) (24) using data downloaded from the CGGA database. The GO and KEGG analysis of genes were drawn using the ggplot2 package of R software (https://cran.r-project.org/web/packages/ggplot2/index.html).

The GBM specimens were collected from the Tianjin Medical University General Hospital (Tianjin, China). The inclusion criteria was a pathologically confirmed diagnosis of GBM. The age range of the patients was from 30 to 60 years and the sex ratio was 1:1. Informed consent for the use of these tissues and data in the present study was obtained from the patients, their immediate family members or their guardians. The study was approved by the Ethics Committee of Tianjin Medical University General Hospital (Tianjin, China).

Glioma tissues of clinical patients were fixed with 4% formaldehyde for 24 h at room temperature, paraffin-embedded at 54°C for 4 h and cut into 4-µm paraffin sections. Following heating at 60°C for 2 h, the paraffin sections were deparaffinized by xylene and rehydrated in graded ethanol. Following antigen extraction with 0.1% sodium citrate buffer (pH 6.0) at 95°C, the tissue sections were quenched of endogenous peroxidase activity with 3% H₂O₂·dH₂O and blocked with 5% goat serum at room temperature for 30 min. Next, the tissue sections were incubated with antibodies against AMOTL2 (1:200; cat. no. 36101; Signalway Antibody LLC) overnight at 4°C and then incubated with HRP-labeled goat anti-rabbit antibody (product no. ZB-2010; ZSGB-BIO, Inc.) at room temperature for 15 min. Following three washes by PBS for 5 min/wash, the slides were stained by diaminobenzidine (cat. no. DA1010; Beijing Solarbio Science & Technology Co., Ltd.) for 3 min at room temperature and counterstained by hematoxylin (cat. no. B3671; APeXBIO Technology LLC) for 10 min at room temperature. Finally, graded ethanol was used to dehydrate these tissues, which were then sealed using neutral balsam and visualized using a fluorescence microscope (cat. no. IX81; Olympus Corporation; magnification, ×20).

Human GBM cell lines U251 and glioblastoma of unknown origin U87MG were purchased from American Type Culture Collection. Cells were cultured at 37°C in a humidified 5% CO₂ atmosphere with Dulbecco's Modified Eagle's Medium (DMEM; Thermo Fisher Scientific, Inc.) containing 10% FBS (Thermo Fisher Scientific, Inc.). These cell lines underwent mycoplasma tests and were authenticated by Short Tandem Repeat (Beijing Microread Genetics Co., Ltd).

The lentiviral vectors with AMOTL2 and control shRNA were purchased from Shanghai GeneChem Co., Ltd. A 2nd generation system was used and the interim cell line used was 293T (Shanghai GeneChem Co., Ltd.). The packaging vector: Envelope was 3:1 and the duration of transfection was 48–72 h. The 4-µg lentivirus was transfected into the target cells (the MOI was ~1-3) when the cell density was 70–80%. After maintaining for 12–16 h, the medium with lentivirus was replaced to fresh DMEM containing 10% FBS. After infection, cells were selected using a 2 µg/ml puromycin solution (Thermo Fisher Scientific, Inc.) and maintained using a 1 µg/ml puromycin solution to create stable infection cell lines to conduct follow-up experiments. Subsequent experiments were performed 72 h after transfection.

AMOTL2 plasmid was created using Shanghai GeneChem Co., Ltd., and the overexpressed plasmids of AMOTL2 were purchased from Addgene, Inc. The 723 ng/µl plasmid was transfected into target cells using Lipofectamine 3000 (Thermo Fisher Scientific, Inc.) at room temperature for 72 h and subsequent experiments were conducted.

Cell samples were lysed in RIPA lysis and extraction buffer (cat. no. 89901; Thermo Fisher Scientific, Inc.) containing protease inhibitor (cat. no. 36978; Thermo Fisher Scientific, Inc.). After evaluating the concentration using a BCA protein assay kit, protein samples were denatured by heating at 100°C and mixed with loading buffer. The protein samples were separated on 10% SDS-PAGE gels at 10 µg per lane and electrophoretically transferred to a PVDF membrane (EMD Millipore). Following blocking with 5% skimmed milk for 1 h at 37°C, incubation with primary antibodies overnight at 4°C and incubation with secondary antibodies for 1 h at room temperature, the membrane was exposed through G:BOX (Syngene Europe) using immobilon western chemiluminescent HRP substrate (cat. no. WBKLS0500; EMD Millipore) to achieve the target proteins. The primary antibodies used in this study were as follows: Anti-AMOTL2 rabbit monoclonal antibody (1:1,000; product no. 43130; Cell Signaling Technology, Inc.), anti-β-catenin rabbit monoclonal antibody (1:5,000; product code ab32572; Abcam), anti-cyclin D1 rabbit monoclonal antibody (1:1,000; product no. 2922; Cell Signaling Technology, Inc.), anti-c-Myc rabbit monoclonal antibody (1:1,000; product no. 18583; Cell Signaling Technology, Inc.), anti-GAPDH mouse monoclonal antibody (1:2,000; cat. no. 40493; Signalway Antibody LLC), anti-histone H3 Rabbit monoclonal antibody (1:1,000; product no. 4499; Cell Signaling Technology, Inc.). The secondary antibodies used in this study were as follows: Goat anti-rabbit IgG H&L (HRP) (1:5,000; product code ab6721) and goat anti-mouse IgG H&L (HRP) (1:5,000; product code ab6789; both from Abcam).

Total RNA from cells was extracted by TRIzol (cat. no. 15596018; Thermo Fisher Scientific, Inc.) and reverse-transcribed using GoTaq^® Reverse Transcription system (cat. no. A3500; Promega Corporation), according to the manufacturer's instructions. Moreover, 2X SYBR Green qPCR Master mix (low ROX) was purchased from Bimake, and the reaction conditions used were according to the manufacturer's instructions (denaturation: 95°C for 15 sec; annealing/extension: 60°C for 30–60 sec; 40 cycles). Comparative quantification was performed using the 2^−ΔΔCq method with GAPDH as the endogenous control (25). All primers were synthesized by Beijing Tianyi Huiyuan Bioscience & Technology Inc. The primer sequences used were as follows: GAPDH forward, 5′-GGAGCGAGATCCCTCCAAAAT-3′ and reverse, 5′-GGCTGTTGTCATACTTCTCATGG-3′; AMOTL2 forward, 5′-TGGAGAAGACCATGCGGAAC-3′ and reverse, 5′-CTTCTCTTGCTCCTGCTGCT-3′.

Cells were generated by seeding 500 cells/well in 6-well plates and subsequently incubating for 10–14 days at 37°C. Cells were fixed with 4% paraformaldehyde for 10 min at room temperature and stained with 0.5% crystal violet blue for 10 min at room temperature. The cell colonies were observed and images were captured by camera and colonies with >10 cells were counted using a fluorescence microscope (cat. no. IX81; Olympus Corporation; magnification, ×10).

A CCK-8 assay was conducted using a CCK-8 kit (cat. no. PA5-84814; Thermo Fisher Scientific, Inc.) at 1–5/6 days. GBM cells were seeded at 3×10³ cells/well in 96-well plates and cultured for 24 h at 37°C. The CCK-8 mixture was configured and added to each well, and the cells were incubated for 1 h at 37°C. Finally, the optical density value was measured using a microplate reader (BioTek Instruments, Inc.) at 450 nm.

Glioma cells were seeded at 4×10⁵ cells/well in 6-well plates and incubated for 12 h. A total of 4 vertical straight scratches were drawn in each well using a 10-µl sterile pipette tip. Plates were washed using PBS and incubated with serum-free medium. Images of the scratches were captured at 0, 12 and 24 h by a fluorescence microscope (cat. no. IX81; Olympus Corporation; magnification, ×4) to evaluate the rate of wound healing. Gap width analysis was performed with ImageJ 1.43 software (National Institutes of Health) and the maximum confluence percentage reached 93%.

A total of 2×10⁴ cells were seeded in the upper chambers of 24-well plates (pore size: 3 µm; Corning, Inc.), incubated with 200 µl DMEM without FBS in the upper chambers, and 500 µl DMEM with 10% FBS was added at the bottom of the 24-well plates. Following incubation for 24 h at 37°C, cells were fixed with 4% paraformaldehyde for 10 min and stained with 0.5% crystal violet blue for 5 min at room temperature. Images of the Transwell were captured by a fluorescence microscope (cat. no. IX81; Olympus Corporation; magnification, ×10). For the Matrigel^® invasion assay, 5×10⁴ cells were added to the bottom membrane of the upper chamber, which was coated with 100 µl Matrigel^® (1:5 dilution with DMEM without FBS) for 1 h at 37°C, and the aforementioned steps as in the migration assay were conducted.

Cells were lysed in 400 µl IP lysis buffer (cat. no. 87787; Thermo Fisher Scientific, Inc.), supplemented with complete protease inhibitor cocktail (CAS no. 329-98-6; Beijing Solarbio Science & Technology Co., Ltd.). The lysates were incubated with 40 µl Protein A Agarose (cat. no. AS046; ABclonal Biotech Co., Ltd.) and appropriate antibodies were placed in a rotating incubator overnight at 4°C. Then, the cells were centrifuged at 5,939 × g at 4°C for 5 min, to remove the supernatant and the precipitation was washed three times with PBS. After boiling with 5X SDS-PAGE loading buffer (cat. no. Top2225; Biotopped Life Sciences, Inc.) for 10 min at 100°C, western blotting was conducted to further analyze the immunoprecipitated samples. The antibodies used in Co-IP assay in this study were as follows: Anti-AMOTL2 rabbit monoclonal antibody (1:200; product no. 43130; Cell Signaling Technology, Inc.), anti-β-catenin rabbit monoclonal antibody (1:200; product code ab32572; Abcam).

Glioma cells were seeded on cover slides in a 24-well plate chamber with 500 µl medium containing 10% FBS incubated for 24 h at 37°C. Cells were then fixed with 4% paraformaldehyde for 30 min, permeabilized by PBS with 0.1% Triton X-100 (PBST) for 5 min, and blocked with 1% BSA in PBST at room temperature. Immunostaining was conducted using the primary antibodies against β-catenin (1:50; cat. no. ab32572; Abcam) overnight at 4°C and goat anti-rabbit secondary antibodies conjugated with Alexa Fluor^® 594 (1:200; cat. no. A11037; Thermo Fisher Scientific, Inc.) for 1 h at room temperature. Images were captured using a confocal microscope (×20).

β-Catenin pathway inhibitor (XAV-939; cat. no. S1180) was purchased from Selleck Chemicals. The AMOTL2-knockdown U87MG and U251 cells were treated with 20 µM of XAV939 for 24 h.

All quantitative data are expressed as the mean ± SD. Statistical analysis was performed with GraphPad Prism 8 software (GraphPad Software, Inc.). Unpaired Student's t-test was used to determine the statistical significance of the data between two experimental groups, while one-way ANOVA was used to analyze the significance of multiple group comparisons. Tukey's post hoc test was performed after ANOVA. All statistical analyses were two-sided, and different cut-off values, (P<0.05, P<0.01, P<0.001), were considered to indicate statistically significant differences. All experiments were repeated at least in triplicate.

To assess the differences in AMOTL2 expression between different glioma grades, data on the expression of AMOTL2 in glioma from TCGA and CGGA databases were analyzed. Data from TCGA database revealed that AMOTL2 expression was significantly decreased in WHO IV grade glioma, as compared with WHO II and III grade glioma (Fig. 1A). Similarly, data from the CGGA database indicated that AMOTL2 expression was significantly decreased in GBM, as compared with low-grade glioma (LGG; Fig. 1A). Notably, the overall survival analysis of these data indicated that glioma patients with a high AMOTL2 expression had a high survival rate (Fig. 1B). Furthermore, data from the CGGA database on isocitrate dehydrogenase (IDH) indicated that the AMOTL2 expression level of the IDH mutant (MUT) was significantly higher than that of the IDH wild-type (WT) (Fig. 1A). Consistently, the results of survival analysis revealed that patients with IDH MUT had a longer survival rate than patients with IDH WT (Fig. 1B). Next, IHC of clinical specimens from patients with glioma was performed to further examine the expression level of AMOTL2. The results indicated that the expression of AMOTL2 was decreased as the malignancy grades increased (Fig. 1C). In conclusion, it was revealed that AMOTL2 was clearly decreased in GBM, and may act as an inhibitor of glioma to prolong the survival time of patients with glioma.

To identify the effect of the differential expression of AMOTL2 in glioma progression, the classic glioma cell lines U87MG and U251 were selected for subsequent tests. First, lentiviral vectors with AMOTL2 shRNA and overexpressed plasmids of AMOTL2 were used to transfect the U87MG and U251 cells to induce AMOTL2-knockdown or -overexpression. Western blotting and RT-qPCR were used to examine the efficiency of the infection of shRNAs and plasmids in U87MG and U251 cell lines (Figs. 2A and B, and S1A and B). Colony formation assays revealed that the downregulation of AMOTL2 significantly promoted colony formation, and AMOTL2-overexpression inhibited colony formation both in U87MG and U251 cells (Figs. 2C and S1C). Consistently, as revealed in Fig. 2D, the result of CCK-8 assays also demonstrated that AMOTL2 overexpression could inhibit the proliferation of U87MG and U251 cells.

It is well-known that the migrating and invasive abilities of tumor cells cause tumor metastasis and poor prognosis. In the present study, following the incubation of cells for 24 h in 6-well plates, 4 vertical straight scratches were drawn. Next, the wound-healing rate was detected at 0, 12 and 24 h to examine the migrating ability of cells, which was revealed to be significantly inhibited by AMOTL2 overexpression and increased following AMOTL2 silencing (Figs. 2E and S1F). Next, a Transwell assay with or without Matrigel^® was conducted to further detect the migrating and invasive abilities of glioma cells. The results revealed that the knockdown of AMOTL2 clearly promoted the migration and invasion of U87MG and U251 cells. On the contrary, the overexpression of AMOTL2 significantly reduced the cell migratory and invasive abilities of the two cell lines (Figs. 2F, and S1D and E). In conclusion, AMOTL2 was revealed to inhibit proliferation, migration and invasion in glioma cells.

To further explore the mechanism of AMOTL2, functional and pathway enrichment analyses were performed using data from the CGGA database by ‘ggplot2’ R language package. The GO analysis results revealed that changes in AMOTL2 biological processes (BP) were significantly enriched in ‘DNA-templated transcription’ and ‘DNA-templated regulation of transcription’ (Fig. 3A). Changes in molecular function (MF) were mainly enriched in ‘protein and DNA binding’ (Fig. 3B). The results of cellular component (CC) analysis revealed that AMOTL2 was primarily enriched in the ‘nucleus’, ‘cytoplasm’ and ‘nucleoplasm’ (Fig. 3C). Furthermore, the results of KEGG pathway analysis revealed that the changes of AMOTL2 were mainly enriched in the ‘Wnt’, ‘cell cycle’ and ‘FoxO signaling pathways’ (Fig. 3D).

Based on the results of GO and KEGG analysis, whether AMOTL2 regulates the Wnt/β-catenin signaling pathway via protein binding was next examined. Co-IP assay between AMOTL2 and β-catenin was performed, and the results revealed that AMOTL2 and β-catenin proteins could mutually combine (Fig. 4A). It was consistent with our hypothesis that AMOTL2 probably plays its role in tumor suppression by directly binding to β-catenin. Subsequently, western blotting demonstrated that AMOTL2-knockdown could increase the expression of certain downstream genes of the Wnt/β-catenin signaling pathway, such as c-Myc and cyclin D1, to activate Wnt signaling; however, no change was observed in the expression of β-catenin (Fig. 4B). This indicated that the function of AMOTL2 in the regulation of Wnt pathway genes activity occurs other than via direct action on the β-catenin destruction complex. It has been reported that, in addition to the β-catenin destruction complex, the activation of the Wnt pathway was also regulated by the distribution of β-catenin in the nucleus and cytoplasm (26). Therefore, to further identify the mechanism of AMOTL2 in the Wnt signaling pathway, the nuclear and cytoplasmic localization of β-catenin in control and AMOTL2-silenced cells were next examined.

First, western blotting of nuclear and cytoplasmic proteins in AMOTL2-silenced and control cells was performed. AMOTL2 was revealed to significantly increase the nuclear distribution and decrease the cytoplasmic distribution of β-catenin proteins (Fig. 4B). Next, immunofluorescence assays were performed to further illustrate the effect of AMOTL2 on β-catenin protein localization. AMOTL2 downregulation was revealed to clearly promote the nuclear transfer of β-catenin protein, both in U87MG and U251 cells (Fig. 4C).

In conclusion, these results indicated that AMOTL2 could prevent β-catenin from translocating to the nucleus by directly binding to the β-catenin protein.

It is unclear whether the regulatory function of AMOTL2 over proliferation, migration and invasion of glioma is realized through the Wnt/β-catenin signaling pathway. Therefore, the AMOTL2-knockdown U87MG and U251 cells were treated with Wnt/β-catenin pathway inhibitor XAV939 with 20 µM for 24 h. The efficiency of the Wnt/β-catenin pathway inhibition was identified by western blotting, and the results showed that the treatment group exhibited a clearly lower β-catenin protein expression, as compared with the control (Figs. 5A and S2A). Colony formation assays revealed that the AMOTL2-silencing-induced increased colony formation ability decreased following XAV939 treatment (Figs. 5B and S2B). Notably, it was revealed by CCK-8 assay that XAV939 reduced the proliferation of AMOTL2-knockdown cells (Figs. 5C and S2C). These results indicated that the Wnt pathway inhibitor XAV939 could reverse the AMOTL2-silencing-induced increase in the proliferative ability of glioma.

Subsequently, rescue assays of migration and invasion were conducted. It was shown by the wound healing assay results that XAV939 could suppress the increased wound-healing rate of AMOTL2-silenced cells (Figs. 5D and S2D). The Transwell assays with or without Matrigel^® indicated that the AMOTL2-downregulation-induced increase of migration and invasion of glioma were significantly reduced following XAV939 treatment (Figs. 5E and S2E). These results indicated that Wnt pathway inhibitor XAV939 could reverse the AMOTL2-silencing-induced increase in the migratory and invasive abilities of glioma.

In conclusion, these results revealed that the Wnt/β-catenin signaling pathway inhibitor XAV939 could reverse the AMOTL2-knockdown-induced proliferation, migration and invasion of glioma. Collectively, AMOTL2 could regulate the proliferation, migration and invasion of glioma through the Wnt/β-catenin signaling pathway.