Introduction
GBM is the most common and lethal primary brain tumor in adults [
1], accounting for 12–15% of intracranial tumors statistically [
2]. GBM is highly aggressive with an abysmal prognosis. Despite employing a combination of surgery, radiotherapy, chemotherapy, and immunotherapy, the survival rate remains dishearteningly low, with fewer than 30% of patients surviving past two years [
3]. Clinical diagnosis and treatment of GBM are challenging due to its aggressive nature, genetic complexity, and our limited understanding of the underlying disease mechanisms.
Epithelial membrane protein 3 (EMP3) belongs to the Peripheral Myelin Protein 22-kDa (PMP22) gene family. This molecular family plays a crucial role in regulating cell growth and orchestrating tissue-specific processes like chondrocyte differentiation and angiogenesis [
4,
5].
EMP3 is located on chromosome 19q13.3 [
6], encodes 163 amino acids at 18 kDa, contains four transmembrane structural domains and two n-linked glycosylation sites on the first extracellular loop [
7,
8], and is involved in cell proliferation, differentiation, apoptosis and migration [
9].
EMP3 was recently reported to be a tumor suppressor gene for several solid tumors, and is drawing attention as a novel prognostic marker [
10], such as glioma [
6], esophageal squamous cell carcinoma [
11], colorectal cancer [
12], and non-small cell lung cancer [
13]. In some kinds of malignancies like breast cancer, and hepatocellular carcinoma,
EMP3 can act as an oncogene [
14,
15]. Increasing evidence suggests that
EMP3 might be a driving force in tumorigenesis and the progression of certain cancers. Low
EMP3 expression has been linked to inhibiting the progression of gastric cancer, notably curtailing the migration and invasion of cancer cells [
16]. Interestingly,
EMP3 expression is found to be altered in glioma tissue, but the gene expression status and biological function of
EMP3 in glioblastoma remains unknown and unclear. Epithelial-to-mesenchymal transition (EMT) is the process by which epithelial cells acquire a mesenchymal stem cell phenotype. Notably, a large body of literature has demonstrated the involvement of EMT in the metastasis of tumor cells [
17,
18].
In this study, we conducted an in-depth exploration of the clinicopathological characteristics, prognostic implications, and diagnostic potential of EMP3 in glioblastoma patients, employing both bulk and single-cell analysis methods. Our experimental observations revealed the pivotal role of EMP3 in amplifying the malignant progression of GBM. Our findings suggest that EMP3 expression may amplify the aggressive trajectory of GBM via regulation of the epithelial-mesenchymal transition (EMT) process. These insights indicate potential advantages of therapeutically targeting EMP3 and modulating the EMT process.
Material and method
The data in this study were obtained from the CGGA database (
http://www.cgga.org.cn) and TCGA database (
https://portal.gdc.cancer.gov/).The expression of
EMP3 in different clinicopathological types of gliomas was analyzed. Based on the median expression levels of
EMP3, we analyzed the survival data of glioma patients, ROC curves were used to measure the diagnostic value of
EMP3 in glioma patients, and univariate and multifactorial analyses of the prognostic role of
EMP3 in glioma were performed using COX regression. We then developed nomogram models to analyze and assess the predictive value of
EMP3 expression in glioma patients at 1, 3 and 5 years of overall survival, and calibration curves were plotted at 1, 3 and 5 years to evaluate the performance of the nomogram.
Cell lines and cell culture
Human glioblastoma cells U87MG and U251MG (purchased from the Cell Bank of the Chinese Academy of Sciences, Shanghai, China) were cultured in DMEM high sugar medium supplemented with 10% fetal bovine serum (FBS, Every Green, China) and 1% double antibody: penicillin–streptomycin (Biyuntian, Shanghai, China), and then cultured at the atmosphere of 5% CO2 in 37 ◦C constant temperature incubator.
Quantitative reverse- transcription polymerase chain reaction (qRT- PCR)
According to the manufacturer’s protocol, total RNA was extracted using the TRIzol reagent (Seven, China). The purity and concentration of total RNA extracted from the cells were then measured by UV spectrophotometer (Nikon, Japan), and the corresponding mRNAs we needed were amplified by reverse transcription. All qPCR reactions were performed by a real-time fluorescent quantitative PCR instrument (BIO-RAD, USA), and the expression levels of the target genes were analyzed. Primers for the target genes: VIM, FOS, SNAI2 and TWIST1 were designed by Beijing Rui Bo Xing Ke Biotechnology Co Ltd, and the primer sequences are shown in Table
1. The expression levels of
EMP3 were calculated using the 2^ − △△CT method.
EMP3 expression levels were normalized to those of GAPDH.
Table 1
Part of primer sequences
VIM | 5’-GACGCCATCAACACCGAGTT-3’ |
FOS | 5’-GAAGACCGAGCCCTTTGAT-3’ |
SNAI2 | 5’-TGTGACAAGGAATATGTGAGCC-3’ |
TWIST1 | 5’-GTCCGCAGTCTTACGAGGAG-3’ |
GAPDH | 5’-UGACCUCAACUACAUGGUUTT-3’ |
Western Blotting (WB) analysis
Protein was extracted from the cells were resolved by SDS-PAGE and then transferred to PVDF membranes (IPVH00010, Millipore), and then incubated with primary antibodies diluted in blocking buffer at 4 °C overnight. The following primary antibodies were used: Mouse Anti-β-Actin (HC201, TransGen Biotech, 1/2000), HRP conjugated Goat Anti-Mouse IgG (H + L) (GB23301, Servicebio, 1/2000), Rabbit Anti EMP3 (DF14661, Affinity, 1/1000), Rabbit Anti TWIST1(AF4009, Affinity, 1/1000), Rabbit Anti SNAI2 (PB9439, Boster, 1/1000), Rabbit Anti FOS (AF5354, Affinity, 1/1000), Mouse Anti-VIM (60330–1-Ig, Proteintech,1/10000), HRP conjugated Goat Anti-Rabbit IgG (H + L) (GB23303, Servicebio, 1/2000), HRP conjugated Goat Anti-Mouse IgG (H + L) (GB23301, Servicebio, 1/2000)was used. Finally, the antigen–antibody reaction was visualized by the enhanced Pierce ECL Western blotting substrate kit (Thermo Scientific/ Pierce, Rockford, IL, USA).
Transient transfection
The negative control (NC),
si-EMP3-1, si-EMP3-2, si-EMP3-3 and GAPDH were purchased from Suzhou Jima Biotechnology Co. U87 cells and U251 cells at logarithmic growth were collected 24 h before transfection. Cells were cultured in 6-well plates (1–2 × 105 cells/well) at 80–90% concentration. A mixture of liposome 2000, siRNA and serum-free DMEM was added to the wells for transfection according to the manufacturer's instructions and replaced with a serum-containing medium 6–10 h later. The sequences of these siRNAs are listed in Table
2.
Table 2
Part of primer sequences
NC | 5’-CCTGAATCTCTGGTACGACTGC-3’ |
si-EMP3-1 | 5’-GCAGUAAUGUCAGCGAGAATT-3’ |
si-EMP3-2 | 5’-GUCUCUCCUUCAUCCUGUUTT-3’ |
si-EMP3-3 | 5’-GUCAGCGGCAUCAUCUACATT-3’ |
GAPDH | 5’-UGACCUCAACUACAUGGUUTT-3’ |
CCK-8 assay
Cells from the U87MG and U251MG control and si-EMP3 groups were inoculated in 96-well plates. Cell counting kit-8 (Dojindo Laboratories, Japan) was diluted to 10% using DMEM and added to the wells to measure the OD value at 450 nm using a microplate reader.
Transwell assays
Five thousand cells were inoculated into the upper chamber of the 24-well transwell chamber and cultured in a serum-free medium for 24 h. At the same time, 600 µL of complete medium was added to the lower chamber. The cells in the upper and lower chambers were then fixed in 4% paraformaldehyde for 30 min and stained with 0.1% crystal violet solution for 30 min. The upper chambers were washed with PBS, and the cells were counted by 200 × light microscopy.
Wound healing assays
The complete medium was removed from the 6-well plates and replaced with a serum-free medium. A 200uL gun was used to draw a uniform, straight line vertically, and PBS was washed three times to remove floating cells. Images of 0h and 24h scratches were recorded by microscopic photography.
Statistical analysis
Statistical analyses and visualization were performed in R (version 3.6.3),IBM SPSS Statistics (version 25.0), and GraphPad Prism (version 7.0.0). All data are expressed as mean ± standard error of the mean. Student's t-test or one-way ANOVA was used to compare the differences among multiple groups. Statistical significance was set at P < 0.05.
Discussion
GBM is a high-grade cancer of glial cell origin with a poor prognosis. It is highly resistant to conventional treatments such as surgery, radiotherapy and chemotherapy, and surviving cells exhibit a more aggressive nature [
23]. Patients with GBM often present with headaches, neurological deficits, seizures, nausea, vomiting, or other neurological sequelae. Many patients have a severe disease burden at the time of initial presentation and a poor prognosis, with a median survival of fewer than two years [
24]. The current backbone of systemic treatment for GBM is temozolomide (TMZ). This alkylating agent targets the N-7 or O-6 position of guanine residues in DNA. the role of TMZ was established in the EORTC 26981/22981 and NCIC CE.3 trials, also known as the Stupp trial, which compared postoperative radiotherapy alone with postoperative radiotherapy and TMZ treatment followed by up to 6 cycles of TMZ maintenance treatment. Overall survival rates were significantly longer in the TMZ group, 27.2% at two years and 10% at five years, compared to 10.9% and 1.9% in the radiotherapy alone group [
25]. The addition of TMZ has shown a survival benefit, particularly for patients with
MGMT promoter methylation, which is present in approximately 50% of tumor patients [
26]. As GBM is susceptible to tolerance to various treatment modalities, there is currently no clinically effective treatment for GBM. In recent years, new therapeutic strategies such as molecularly targeted therapies, immunotherapy and tumor electric field therapy have offered hope for highly drug-resistant GBM. Targeted therapies based on BRAFV600E mutation, NF1 mutation, EGFRvIII mutation and anti-angiogenesis are already available after preliminary clinical trials [
27]. The discovery of new targets for developing precision-targeted therapy for GBM patients is now promising.
EMP3 expression is upregulated in brain tumors, particularly in GBM. Ernst et al. [
28] showed high
EMP3 levels in GBM spherical cultures. Scrideli et al. [
29] also found higher expression of
EMP3 in GBM.
EMP3 mRNA expression has been shown to correlate with poor prognosis [
30], and deletion of
EMP3 protein reduces malignant behavior in GBM cells [
31]. Consistently with these previous findings, our single-cell analysis revealed
EMP3 correlates with EMT score in high-grade glioblastoma cells. We analyzed the expression levels of
EMP3 in different clinicopathological types of glioma based on mRNA-seq data from glioma patients in the CGGA and TCGA databases. Our results elucidate that
EMP3 is enriched in more malignant glioma subtypes and that univariate and multifactorial analyses show that
EMP3 expression is an independent risk factor for OS in glioma patients. Patients who has higher mRNA level of
EMP3 perform worse prognosis in clinical practice. Functional experiments showed that the knockdown of
EMP3 significantly suppressed the malignant phenotype of tumor cells in terms of proliferation and migration. Consistent with other literature reports, Anke Zhang et al. [
32] demonstrated
EMP3 as a novel predictor for clinical progression and clinical outcomes in glioma. However, what are the mechanisms and reasons behind it? So far, there is still no clear indication. In our study, silencing of
EMP3 inhibits the malignant behavior of glioblastoma cells by regulating the EMT process mechanistically. GUO et al. [
3] found that low expression of
EMP3 can regulate the EMT process, hinder its development, and thus inhibit the invasion of gastric cancer cells, which corresponds to our results. Overall, our results reveal that
EMP3 can predict poor prognosis in glioblastoma patients.
EMT process involves some fundamental processes, including embryonic evolution, tissue formation, wound healing and tissue fibrosis. In addition, EMT has been shown to promote tumor cell growth, drug resistance and proliferation. High-expression of mesenchymal gene signature predicts poor prognosis in glioma patients [
33], suggesting that EMT-like processes are closely associated with GBM invasion. Several key signaling pathways, including transforming growth factor beta (TGF-β), Wnt, Notch and Hedgehog, are known to be involved in EMT [
34]. These signaling pathways ultimately lead to the activation of EMT transcription factors (EMT-TFs). Several transcription factors have been identified as master regulators of EMT, including
SNAI2 factors, bHLH factors (E12 and E47,
TWIST1 and
TWIST2) [
35].
SNAIL1/2 and
TWIST1/2 are considered the main regulatory factors driving the transcription pathway of EMT, and they converge to activate the expression of transcription factors [
36].
SNAIL1 and
SNAIL2 are involved in embryonic development, fibrosis, tumor development, and activation of EMT [
37], and together with other transcriptional regulatory factors, they control gene expression. Besides, there are also newly identified transcription factors, which is a crucial regulator of tumor progression [
38], and a critical transcription factor for EMT [
39], such as
FOSL2. In the SMAD dependent pathway, activation of the SMAD complex by the TGF-β receptor induces increased expression of the mesenchymal marker
VIM [
40]. Key transcription factors mediate the transition of cells from epithelial to mesenchymal transition, and these transcription factors mainly regulate the process intercellular adhesion, cell polarity, and vitality [
41], and metabolism, transcription, differentiation. Transcription factors inhibit genes associated with epithelial phenotype induction of mesenchymal gene expression leads to EMT characteristics in cell [
42]. Under specific physiological environments, the expression of SNAIL can be activated through various signaling pathways, including TGF-β, Wnt, Notch and growth factors acting on RTK [
43] indicate that SNAIL related EMT processes are influenced by multiple mechanisms Control driven [
44]. Related studies have found that down regulation of SNAI2 blocks the ability of TWIST to activate EMT in breast cells, suggesting that TWIST can indirectly induce E-cadherin transcriptional inhibition [
45].
In our study, we found a positive correlation between
EMP3 and the expression of
VIM, FOS, and
SNAI2 genes. Zhou W et al. found
SNAI2 increased the migration and invasion ability of tumors by promoting the loss of cell adhesion and polarity [
46]. We found that SNAI2, TWIST1 and FOS protein expressions were lower in
EMP3 knockdown U87 cells than U87 wide type. This result implied that
EMP3 might promote EMT process. Interestingly, VIM did not perform the same change as others. VIM plays an integral role in lamellipodia formation and maintenance of cell polarity in migrating cells [
47]. Dynamic reorganization of VIM might be continued in the lamellipodia for the formation of cellular polarity, without the influence of
EMP3. However, the varying mRNA expression of
TWIST1 in U87 and U251 cell lines following
EMP3 knockdown could be attributed to the fact that EMT can be active in different regions of the tumor and at different stages of cancer progression. This could result in the involvement of different EMT-associated transcription factors (EMT-TFs) at distinct stages of tumor progression, leading to EMT heterogeneity. Therefore, based on our observations, it can be inferred that
EMP3 might be promoting the malignant phenotype of GBM via the EMT process. Nevertheless, the exact mechanism underlying this process warrants further investigation.
Whether it is related to the tumor microenvironment of immunosuppression? Recently, Qun Chen et al. reported that EMP3 was associated with immunosuppression in GBM. Elevated EMP3 in GBM areas was accompanied by high expression of PD-L1 and abundant M2 Tumor-Associated Macrophage (TAM) recruitment but a lake of T cell infiltration. They found that EMP3 was a potent protein in M2 TAM polarization and recruitment that impaired the ability of GBM cells to secrete CCL2 and TGF-β1. Furthermore, EMP3 suppressed T cell infiltration into GBM tumors by inhibiting the secretion of CXCL9 and CXCL10 by macrophages and led to an effective response to anti-PD1 therapy [
48]. This opinion coincided with what some scholars considered, EMT is associated with cancer cell stemness, metastasis, chemotherapy resistance, and immune suppression [
41,
49]. Tumor epithelial mesenchymal plasticity (EMP) refers to the transformation of tumor cells between epithelial like cells and completely or partially mesenchymal like cells [
50,
51]. EMP is also a potential mediator causing resistance to immune checkpoint inhibitors [
52,
53]. Currently, the precise role of
EMP3 in GBM remains not fully elucidated, and the exact in vivo function of
EMP3 calls for further verification through integrated spatial omics and ex vivo experimental investigations. The swift discernment of
EMP3's functional mechanism could expedite the development of targeted therapies for GBM, and pave the way for discovering new therapeutic strategies to tackle chemoresistance, thereby improving clinical outcomes for GBM patients.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.