ADAMTS8 inhibits glioma development in vitro and in vivo

Introduction

Glioma is the most prevalent primary brain tumour in human beings. According to the latest World Health Organization (WHO) CNS5 classification, glioma is divided into 4 families, including the adult diffuse gliomas, paediatric low-grade diffuse gliomas, paediatric high-grade diffuse gliomas and circumscribed astrocytic gliomas. Among these, glioblastoma (GBM), accounting for 57% of all the occurrences of gliomas, is one of the most malignant types that lead to high mortality and recurrence of solid tumour [32]. The low-grade gliomas (LGG) have better prognosis [1]. The survival rate in general for GBM patients is low even after the surgery combined with chemotherapy of radiotherapy [2].
In recent years, novel targets have been discovered in gliomas. CircNEIL3 was revealed to be enriched in glioma tissues and cells and its upregulation could promote the glioma progression and might be a potential therapeutic target [25]. Piwil1 was found upregulated in glioma stem-like cells and knockdown could suppress tumour progression, suggesting that Piwil1 might be a therapeutic target for GBM [15]. Long non-coding RNA (lncRNA) ST7-AS1 expression was found low in glioma tissues and cells and overexpression suppressed the glioma cell growth [30]. Similarly, YAP facilitated the glioma progression, suggesting that YAP might be a therapeutic target [38].
ADAMTS (a disintegrin and metalloproteinase with thrombospondin motifs) family members were detected to be in human brain tumours [14,17]. ADAMTS1 upregulation was discovered to be correlated with the malignant progression in low-grade gliomas [13]. Further research showed that knockout of ADAMTS1 could inhibit the tumour sphere formation and angiogenesis-like features in vitro in glioma [29]. ADAMTS5 was upregulated in GBM tissues and cells, and overexpression in GBM cells could promote the cell invasion through degradation of brevican [24]. So far, it has been reported that ADAMTS8 is downregulated in several tumours including colorectal cancer [16], oesophageal squamous cell carcinoma [33], breast cancer [35] and lung cancer [10]. Further studies reported that ADAMTS8 upregulation could inhibit the progression of lung cancer, breast cancer and colorectal cancer in vitro and in vivo [6, 16, 35, 37]. In addition, ADAMTS8 might act as a prognostic marker for metastasis in breast cancer patients in the lymph node-negative early stage [17]. ADAMTS8 has been reported to be reduced in brain tumours [11]. However, the role of ADAMTS8 in brain tumours is not confirmed yet. Therefore, whether ADAMTS8 could act as a tumour suppressor or not in glioma is what we try to explore in this study.

Material and methods

Ethical statement

This study did not involve any human specimens. The animal experiments were carried out strictly in accordance with the Animal Welfare Act of Tianjin First Central Hospital.

Online database analysis

Gepia online database was used to analyse the mRNA expression level of ADAMTS8 in GBM and LGG. The overall survival and disease-free survival analysis was performed in patients with GBM and LGG with 95% confidence interval. In addition, the Chinese Glioma Genome Atlas database (CGGA, http://www.cgga.org.cn/) was used to perform the overall survival analysis among patients with primary glioma or recurrent glioma. The correlations between ADAMTS8 and MMP2 or MMP9 were analysed.

Cell culture

The glioma cell lines T98G, U251 and U87 were purchased from Procell (Wuhan, China). HEB, the astrocyte cell line of human brain was bought from Mingzhou Bio (Ningbo, China). Cells were cultured in DMEM solution with 10% FBS (Evergreen, Zhejiang, China).

Cell transfection

The short hairpin RNAs (sh-RNA) against ADAMTS8 were synthesized by GenePharma (Suzhou, China) with the non-targeted sequences as the negative control (sh-NC). The overexpressed plasmid pc-ADAMTS8 was constructed based on pc-DNA3.1 plasmid. The glioma cells were transfected with shRNAs and pc-ADAMTS8 using the Lipo6000 Reagent (Beyotime, Shanghai, China). The pc-NC and pc-ADAMTS8 plasmids were stably transfected into the U87 cells and G418 was used to screen the stable transfected cells.

RT-qPCR

The total RNAs were extracted using Beyozol reagent (Beyotime). The ADAMTS8 mRNA expression level was detected using the SYBR PrimeScript RT-PCR Kit on 7500 RT-PCR system (Applied Biosystems, CA, USA), with GAPDH as the internal control. The primer sequences were listed in Table I.

Western blot

Total protein was extracted using RIPA from cells and then the SDS-PAGE method was applied to separate the proteins. The proteins were transferred to PVDF membranes. The primary antibodies against ADAMTS8 (bs-5859R, 1 : 500, Bioss, China), E-cadherin (bs-1016R, 1 : 500, Bioss), Vimentin (bs-8533R, 1 : 500, Bioss), Snail (bs-21598R, 1 : 500, Bioss), MMP2 (ab181286, 1 : 1000, Abcam, USA), MMP9 (ab76003, 1 : 1000, Abcam, USA) and GAPDH (bs-41373R, 1 : 1000, Bioss) were used to incubate the membranes overnight at 4°C. HRP-labelled Goat Anti-Rabbit IgG (H+L) antibody from Beyotime was used as the secondary antibody, which was diluted at 1 : 2000 and used to incubate the membranes for an hour at 37°C. The ECL kit was added on the membranes and the blot bands were then visualized on the western blot imager (Peiqing, Shanghai, China). The relative expression was analysed using Image J (NCBI).

CCK8 analysis/h5> Cells after transfection were selected and 2000 cells/100 µl were seeded onto each well of the 96-well plates for viability test. The CCK8 kit from Beyotime was applied. Then, 10 µl CCK8 solution was added in 3 replicates at 0, 24, 48 and 72 h respectively. After incubation for 1 h, the plates were detected on a microplate reader at 450 nm for OD values.

Apoptosis assays

Cells after transfection were collected and seeded into 24-well plates and then the apoptosis kit was used as per the manufacturer’s instructions (Annexin V/PE, Biolegend Co., Changsha, China). The apoptosis rates in each group were detected on the lab flow cytometer (BD Biosciences, CA, USA).

Transwell experiments

Transwell experiments Cells after transfection were selected from each group and then seeded into the serum-free upper chambers of Transwell for further culture. The upper chambers used in invasion assays were pre-coated with Matrigel (BD Biosciences). The lower chambers were filled with 10% fetal bovine serum (FBS). The cells that migrated and invaded were collected and fixed using 1% paraformaldehyde. Thereafter, the haematoxylin was used to stain the fixed cells. Finally, the cell numbers from each group were observed under microscope and cell images were taken from five different fields.

Animal experiments

Six Balb/c nude mice at the aged of 4 weeks old were purchased from Vital River (Beijing, China). The stable transfected U251 cells were injected subcutaneously on the back of the mice. The tumour volumes were estimated on Day 7, 13, 18, 24, 30, 35, 40 and 45. The mice were sacrificed on Day 45 and tumour tissues were separated from the mice. Tumours from each group were weighed. The tumour images were taken on a smart cell phone.

Statistical analysis

GraphPad 8 was used for data analysis and figure generation (Prism, CA, USA). All groups in assays were examined in triplicate. The Kruskal-Wallis test was applied in multiple groups. Unpaired t test was used within two groups. Two-Way ANOVA analysis was applied in analysis of cell viability.

Results

ADAMTS8 is associated with higher survival in primary glioma patients

According to the Gepia database, the mRNA expression of ADAMTS8 was suppressed in GBM and LGG tumour tissues (Fig. 1A). The survival analysis on Gepia further showed that the glioma patients with a higher ADAMTS8 expression might be correlated with higher overall survival and higher disease-free survival rates, which suggests that ADAMTS8 might be a prognosis biomarker in glioma (Fig. 1B, C). The survival analysis based on CGGA database further showed that a higher expression of ADAMTS8 was associated with higher overall survival in primary glioma patients (Fig. 1D). However, the correlation between the ADAMTS8 expression and overall survival rate was not significant in recurrent glioma patients (Fig. 1E). In vitro, the ADAMTS8 mRNA and protein expression levels were lower in glioma cell lines than the normal HEB cells (Fig. 1F-H).

The upregulation of ADAMTS8 inhibits the viability and promotes the apoptosis of the glioma cells

The U87 and T98G cells were transfected with the shRNAs against ADAMTS8 and the overexpressed plasmids. Then RT-qPCR and western blot assays were used to validate the transfection efficiency. Results showed that ADAMTS8 mRNA and protein levels were enhanced in pc-ADAMTS8 groups in both cell lines (Suppl. Fig. 1A, B, E-G). Similarly, in the sh-ADAMTS8 groups of both cell lines, the mRNA and protein levels of ADAMTS8 were inhibited (Suppl. Fig. 1C, D, H-J). Further, we performed the CCK8 and apoptosis experiments to examine the effect of ADAMTS8 in glioma cell lines. It was validated that ADAMTS8 upregulation could reduce the viability but facilitate the apoptosis of U251 and T98G cells (Fig. 2A, C). On the contrary, the knockdown of ADAMTS8 in U251 and T98G cells could increase the viability and suppress the apoptosis of U251 and T98G cells (Fig. 2B, D).

The upregulation of ADAMTS8 inhibits the mobility and EMT of the glioma cells

The transwell assays were performed to examine the migration and invasion of U251 and T98G after the upregulation or downregulation of ADAMTS8. The findings were that the upregulation of ADAMTS8 could suppress the migration and invasion of cells (Fig. 3A, B, E, F). In contrast, the knockdown of ADAMTS8 was discovered to enhance the cell mobility (Fig. 3C, D, G, H). Furthermore, the EMT biomarkers were detected in each group by western blot. E-cadherin was promoted in both cell lines after the upregulation of ADAMTS8 while the Vimentin and snail were inhibited (Fig. 3I, Suppl. Fig. 2A, C, E, G, I, K). On the other hand, the knockdown of ADAMTS8 in both cell lines were found to inhibit E-cadherin and promote snail and Vimentin (Fig. 3I, Suppl. Fig. 2B, D, F, H, J). In addition, ADAMTS8 was negatively correlated with MMP2 and MMP9 in glioma tissues according to CGGA database (Suppl. Fig. 3).Further, we confirmed in glioma cells that ADAMTS8 upregulation could inhibit the secretion of MMP2 and MMP9 and its downregulation could promote the secretion (Fig. 3J, Suppl. Fig. 2M, N).

The upregulation of ADAMTS8 suppresses the glioma growth in vivo

After the injection of stable transfected cells into the nude mice, results confirmed that the overexpression of ADAMTS8 inhibited the tumour growth in nude mice (Fig. 4A-C). In addition, MMP2 and MMP9 protein expression levels in tumour tissues were inhibited in the pc-ADAMTS8 group (Fig. 4D-F).

Discussion

Glioma is a considerably complicated disease and so far, there have been various genes reported to be involved in the modulation of the pathogenesis. It was recently discovered that MXRA8, matrix remodelling-associated protein 8 expression was higher in glioma and higher expression in glioma was correlated with lower survival rates; downregulation of MXRA8 could inhibit the glioma cell growth through regulating the ferroptosis and immune microenvironment [34]. HMGA1 was revealed to be upregulated in glioma and modulate the cell survival, invasion and migration via PI3K/Akt/c-Jun pathway [26]. ARPC1B, actin-related protein 2/3 complex subunit 1B, was revealed to facilitate the invasion, migration and the EMT in glioma cells and promote intracranial tumour growth [18]. Further, ARPC1B was discovered to regulate the tumour microenvironment in the co-culture model of macrophage and glioma cells [18]. Previously, ADAMTS8 was reported to be downregulated in brain tumours [11]. Apart from this, there is no further study on the role of ADAMTS8 in glioma. In this study, we disclosed that ADAMTS8 was inhibited in glioma, and higher ADAMTS8 expression in glioma patients was correlated with better overall survival and disease-free survival, particularly in primary glioma patients, suggesting that clinically, ADAMTS8 expression might act as a prognostic biomarker for glioma patients. Therefore, we studied the regulatory function of ADAMTS8 in glioma cells and confirmed that ADAMTS8 overexpression could suppress the cell proliferation, invasion and migration and promote the cell apoptosis. Further, we also validated the suppressive role of ADAMTS8 in glioma growth in nude mice.
Glioma malignancy is correlated with the tumour infiltration and tumour microenvironment. The cell invasion enables glioma cells to escape from the original tumour site, resulting in metastasis [5]. The degradation of extracellular matrix is of importance in tumour infiltration and development in microenvironment [4,5,7,27,28]. The dense extracellular matrix is a natural obstacle for the infiltrating cells and in order to go through, proteases are secreted, including the matrix metallopeptidases (MMPs) and proteases like a disintegrin and metalloproteinases (ADAMs) [12]. MMP2 and MMP9 play important roles in degrading the extracellular matrix components and thereby effecting on the migration and invasion of glioma cells [8,9,12,19,22,23,31]. In addition, MMP2 and MMP9 were revealed to be associated with higher recurrence in glioma patients [39]. In this study, we reported that ADAMTS8 was not only negatively correlated with MMP2 and MMP9 in glioma tissues, but also capable of inhibiting MMP2 and MMP9 secretion in glioma cells and tumours in xenograft models, suggesting that ADAMTS8 might play its suppressive role in glioma through inhibiting secretion of MMP2 and MMP9.
On the other hand, epithelial-mesenchymal transition (EMT), featuring the loss of epithelial characteristics like cell-cell adhesion and gain of mesenchymal traits including the increase of invasion and migration, is widely involved in malignancy of various cancers [36]. Apart from the epithelial protein marker E-cadherin, mesenchymal marker Vimentin, transcriptional factors like snail are correlated with the EMT-like features in glioma [21]. In addition, MMPs like MMP2 and MMP9 are actively involved with the EMT regulation [3,28]. In this study, we found that ADAMTS8 could inhibit the EMT-like features in glioma, perhaps via suppression of MMP2 and MMP9.

Conclusions

Taken together, this study reports that higher ADAMTS8 expression in glioma patients was correlated with better prognosis and ADAMTS8 might inhibit the glioma development via regulating MMP2 and MMP9.

Supplementary figures are available on journal’s website.

Disclosure

The authors report no conflict of interest.

References