Glioma is one of the most malignant and aggressive tumors among the primary central nervous system tumors(20). Although great effort has been made in the treatment of glioma, the prognosis of patients with glioma has not effectively improved, likely due to the lack of understanding of the mechanism of glioma(4).
The present study performed bioinformatics analysis using expression data from the GEO, TCGA and GTEx databases. The results indicated that the mRNA transcription level of CNN3 in LGG and GBM was significantly increased compared with that in human normal brain tissues, which suggested that dysregulation of CNN3 expression may be involved in the pathogenesis and development of glioma. KEGG enrichment analysis of CNN3 showed the same trend. In addition, according to the survival curves generated using the data from TCGA and CGGA databases, the prognosis of patients in the CNN3 high-expression group was significantly worse than that of patients in the CNN3 low-expression group (P < 0.05).
Subsequently, to verify whether the mRNA transcription level of CNN3 was higher in glioma, as obtained by bioinformatics analysis, four commonly used glioma cell lines (U251, LN229, T98G and A172) and NHAs were selected for RT-qPCR assays. The results showed that CNN3 was significantly overexpressed in the above glioma cell lines compared with NHAs. In addition, the expression levels of CNN3 were highest in U251 and T98G cells. Thus, the U251 and T98G cell lines were selected for subsequent experiments. Western blotting was also employed to detect the protein level of CNN3 in the four aforementioned glioma cell lines and NHAs. The results also showed that CNN3 was highly expressed at the protein level in all the glioma cell lines. Furthermore, 30 pairs of glioma and adjacent normal tissues were selected from patients with glioma for RT-qPCR and IHC assays, and the experimental results were consistent with those obtained in cell lines. Thus, it was concluded that CNN3 was highly expressed in glioma at both the mRNA and protein level.
In different malignant tumors, the expression level and functional role of CNN3 are different. For example, in NSCLC, CNN3 exhibits low expression and acts as a tumor suppressor gene, which inhibits the proliferation, migration and invasion of NSCLC cells by affecting the PI3K/AKT signaling pathway (17). By contrast, in cervical cancer, CNN3 is highly expressed, and by affecting the mRNA transcription level of RPLP1, it increases the proliferation and invasion abilities of cervical cancer cells, thereby shortening the survival time of patients with cancer (15). In osteosarcoma, CNN3 was reported to play an oncogenic role in tumor development by activating the ERK1/2 and p38 signaling pathways, and was associated with poor prognosis in patients with osteosarcoma (21). CNN3 is also associated with drug resistance. For example, (16)found that, in colorectal cancer, CNN3 can promote colon cancer lymph node metastasis and enhance colon cancer chemotherapy resistance by affecting β-catenin and p53. Therefore, the current study explored which biological processes of glioma cells were specifically affected by CNN3.
To compare the effects of different expression levels of CNN3 on the biological function of glioma cell lines, glioma cells with knocked down CNN3 expression were constructed. Cell transfection using shRNA is a commonly used and effective gene knockdown method. Three shRNAs with different knockdown sites (sh-CNN3-1, sh-CNN3-2 and sh-CNN3-3) were employed, and their CNN3 knockdown efficiency was detected at the mRNA and protein level by using RT-qPCR and western blotting, respectively. The results showed that the knockout efficiency and stability of sh-CNN3-1 and sh-CNN3-3 were lower than those of sh-CNN3-2; thus, sh-CNN3-2 was selected for subsequent functional experiments.
As aforementioned, CNN3 plays different biological functions in different types of tumor cells. Therefore, to explore the specific biological activities that CNN3 affects in glioma cells, CCK-8 assays were performed in the current study to explore the effect of CNN3 on the proliferation of glioma cells. The results showed that, compared with that of the NC group, the cell proliferation rate of the CNN3-knockdown group was significantly decreased in both the U251 and T98G cell lines. Subsequently, Transwell invasion and migration assays were performed on U251 and T98G cells in both the NC and CNN3 knockdown groups. The results revealed that, after knocking down CNN3 expression, the invasion and migration abilities of cancer cells were significantly decreased in both glioma cell lines. Thus, it was concluded that CNN3 could affect the proliferation, invasion and migration abilities of glioma cells.
To explore the effect of CNN3 on glioma growth in vivo, U251 transfected with sh-CNN3 and sh-NC were subcutaneously inoculated into mice (4 mice in each group) to construct a glioma xenograft model. One mouse in the NC group was inoculated unsuccessfully, which meant the glioma cells did not grow in the mouse subcutaneous tissue. The volume of subcutaneous tumors in mice was estimated every 7 days after inoculation (on days 7, 14, 21, 28 and 35) according to the formula ‘tumor volume = long diameter x short diameter2 x 0.5’(22). The mice were sacrificed on day 35 after inoculation, and the subcutaneous tumors of each mouse were removed and their weights recorded. The results showed that the growth rate and quality of tumors of subcutaneous tumors in the CNN3 knockdown group were significantly lower than those in the control group. However, due to budget constraints and limited laboratory equipment, the number of mice included in the present experiments was relatively small, and glioma cells were not directly inoculated into the mouse brains, which should be conducted in future research.
Finally, to investigate the specific mechanism by which CNN3 affects glioma growth, the online tool STRING was used to analyze other genes that may interact with CNN3. The results revealed that CNN3 had a strong potential to interact with ERK. Previous studies showed that ERK could affect the migration and invasion of cells by activating the MAPK/ERK signaling pathway (23, 24). Both ERK1 and ERK2 can be phosphorylated simultaneously in glioma cells(25, 26), and phosphorylated ERK1/2 can further activate glioma growth-related pathways. Therefore, the present study calculated the correlation between CNN3 and the expression of related genes (MEK1/2 and ERK1/2) in the MAPK signaling pathway using Pearson’s correlation analysis and the mRNA expression data of GBM and LGG in TCGA. The results showed that CNN3 was significantly correlated with the mRNA transcription levels of MEK1/2 and ERK1/2 (P < 0.05). Therefore, it was concluded that CNN3 may affect the occurrence and development of glioma through the MEK/ERK signaling pathway, but this hypothesis needs further experimental verification.
In conclusion, the present study demonstrated that CNN3 functioned as an oncogene in glioma in vitro and in vivo. Bioinformatics analysis suggested that CNN3 had a negative correlation with the prognosis of patients with glioma and had a significant correlation with the MEK/ERK signaling pathway in glioma. Overall, the present results suggest that CNN3 may potentially be a novel target and biomarker for the treatment of glioma.