Sage Journals: Discover world-class research

Abstract

AIMS:

To evaluate the prognostic and clinicopathological features of glioma with BRMS1L expression.

METHODS:

Total 120 glioma samples were obtained as discovery cohort. CGGA, GSE and TCGA datasets were obtained as validation sets. Furthermore, Kaplan-Meier survival and multivariate Cox analysis were used to evaluate the survival distributions. Moreover, the functional role of BRMS1L was also analyzed by transwell assay.

RESULTS:

In the discovery cohort, decreased BRMS1L expression was significantly associated with high-grade glioma as well as the higher mortality in survival analysis (log-rank test, $p<$ 0.01). And the three validation cohorts showed the similar results. Furthermore, BRMS1L act as an independent prognostic factor in glioblastoma patients. Additionally, functional assay showed that ectopic of BRMS1L suppressed glioma cells’ invasion.

CONCLUSION:

BRMS1L plays as an anti-oncogene in GBM and indicates a new potential therapeutic target.

Keywords

Glioma BRMS1L prognostic invasion

1. Introduction

Glioma is the most common type of primary brain tumor, patients with gliomas have an invariable fatal outcome and poor prognosis [1, 2, 3, 13]. According to immunophenotypical similarity to a cell of putative origin, gliomas are subgroupped into astrocytoma, oligodendroglioma and oligoastrocytoma [9]. World Health Organization (WHO) grades II, III and IV gliomas are malignant, diffuse, and incurable [2]. Patients with grade IV gliomas (glioblastomas; GBM) suffers average survival barely exceeding 16 months even with the most aggressive therapies [5]. The GBM cells is characterized by uncontrolled growth, migration and spread throughout the brain tissue [6]. Even though increasing studies reveal underlying mechanisms of the malignant properties of glioma cells and explore reasonable therapeutic targets, the existing clinical prediction and treatment efficacy are still not satisfactory [5].

Figure 1.

(A) BRMS1L expression is positively correlated with tumor grade. This result is validated by GSE 16011 microarray database (B), CGGA RNAsequencing database (C), and TCGA RNAsequencing database (D). A single spot is the BRMS1L expression value of an individual patient. Lines in the middle are the mean expression value. Error bars represent standard deviation (SD). * $P<$ 0.05, ** $P<$ 0.01.

BRMS1L (breast cancer metastasis suppressor 1 like, BRMS1-like) is involved in Sin3A–histone deacetylase (HDAC) co-repressor complex that blocks target gene transcription [4]. Through recruitment of HDAC1 and deacetylation of histone H3K9 at the promoter, BRMS1L modulates epigenetic silencing of Wnt signaling receptor and inhibits breast cancer metastasis [4]. Downregulated BRMS1L expression also contributes to epithelial-mesenchymal transition and invasion in breast cancer [15]. Clinically, reduced BRMS1L in breast cancer is found to be correlated with metastasis and poor survival, also validating BRMS1L is an important breast cancer metastasis suppressor [4]. Meanwhile, the treatment of human lung cancer resulting in downregulated BRMS1L expression also indicates a possible anti-oncogene in lung cancer [8]. Recently, a Glioma Associated Protein-Protein Interaction Network analysis predicts the potential involvement of BRMS1L in the pathogenesis of glioma [14]. However, the explicit role of BRMS1L in glioma malignant phenotype and prognosis prediction needs to be assured.

In the present study, we demonstrated that BRMS1L expression was inversely correlated with glioma grade. Forced BRMS1L expression inhibited aggressive invasion of GBM cells. Loss of BRMS1L in GBM predicted a poor survival by our discovery cohort, GSE16011 and TCGA cohort. Targeting BRMS1L is promising to a novel therapeutic target in GBM treatment.

2. Materials and methods

2.1 Glioma samples and patients

Total 120 glioma samples (30 grade II, 30 grade III, and 60 grade IV) were confirmed by pathological diagnosis according to the 2016 WHO classification. After resection, all samples were immediately frozen in liquid nitrogen until RNA extraction. Extent of resection was graded as total or subtotal resection by using MRIs obtained within 72 h after surgical resection by two independent radiologists. Overall survival time (OS) was calculated from the date of histological diagnosis until death or the last follow-up. This study was approved by the institutional review boards of Nanjing Medical University, and written informed consent was obtained from all patients.

Figure 2.

The prognostic value of BRMS1L in GBM patients. Low BRMS1L expression confers a poor survival in 60 GBM patients (A), GSE16011 database (B), CGGA database (C), and TCGA database. High group, patients with higher BRMS1L level than the median value or equal to the median one. Low group, patients with lower BRMS1L value than the median one.

2.2 RNA extraction and quantitative real-time reverse transcription PCR (qRT-PCR)

Total RNA was extracted from glioma samples by using TRIzol reagent (Invitrogen). RNA was quantified by absorbance reading at A260/A280 $=$ 1.6–1.8, qRT-PCR was performed using the SYBR Premix ExTaq TM (TaKaRa) according to the manufacturer’s instruction. Primers for BRMS1L and GAPDH were obtained from Invitrogen as follows, BRMS1L forward primer: 5’-AGTGAAAACGGAACCACCTG-3’, BRMS1L reverse primer: 5’-CCATCAGGCCTCTTAAACCA-3’. GAPDH forward primer: 5’-CCACCCATGGCAAA TTCCATGGCA-3’, reverse primer: 5’-TCTAGACGG CAGGTCAGGTCCACC-3’. The relative level of BRMS1L was normalized to GAPDH.

2.3 BRMS1L expression analysis in the validation datasets

Chinese Glioma Genoma Atlas (CGGA) RNA sequencing database (http://www.cgga.org.cn), GSE16011 data (http://www.ncbi.nlm.nih.gov) and Cancer Genome Atlas (TCGA) RNA sequencing database (http://cancergenome.nih.gov) were obtained as validation sets. In the three datasets, samples with definite WHO classification were used for grade expression analysis, and only GBM samples with survival data were employed for survival analysis.

2.4 Cell line, cell culture and transfection

U87 and U251 cell lines were purchased from the Chinese Academy of Sciences Cell Bank; cells were maintained in a 37 ${}^{\circ}$ C, 5% CO ${}_{2}$ incubator, cultured by DMEM supplemented with 10% fetal bovine serum (FBS). BRMS1L overexpressing vectors and negative control vectors were constructed by GeneChemCo., Inc (China). Cells of 70%–80% confluence were transfected with the vectors by using Lipofectamine 2000 reagent (Invitrogen) according to the manufacturer’s instructions and screened by the aminoglycoside G418.

2.5 Invasion assay

Transwell assay was applied to assess cells’ invasion ability, 1 $\times$ 10 ${}^{5}$ cells were plated into 24-well Boyden chambers with an 8-um pore polycarbonate membrane, which was coated with 20 ug of Matrigel (BD Biosciences). Cells in the upper chamber were cultured with 200 ul serum-free medium, while the medium containing 20% FBS was added to the lower chamber to play as a chemoattractant. Twenty-four hours later, non-invasive cells were removed from the upper well by cotton swabs, and the invasive cells were fixed with 4% paraformaldehyde, then stained with 0.1% crystal violet, and photographed ( $\times$ 200).

Table 1
BRMS1L and other clinicopathological factors affecting 60 GBM patients’ overall survival

Variable	Univariable regression			Multivariable Regression
	HR	95% CI	$p$ value	HR	95% CI	$p$ value
Gender
Female	1.00
Male	0.922	0.474–1.792	0.810
Age
$<$ 60	1.00			1.00
$\geqslant$ 60	2.340	1.214–4.510	0.011	2.090	1.020–4.282	0.044
KPS score
$<$ 80	1.00			1.00
$\geqslant$ 80	0.316	0.162–0.619	0.001	0.363	0.169–0.781	0.009
Extent of resection
Total	1.00			1.00
Subtotal	2.233	1.167–4.271	0.015	1.608	0.774–3.343	0.203
BRMS1L expression
Low	1.00			1.00
High	0.460	0.237–0.891	0.021	0.307	0.143–0.659	0.002
Radiotherapy $+$ TMZ
No	1.00			1.00
Accept	0.393	0.206–0.750	0.005	0.314	0.153–0.646	0.002

HR, hazard ratio.

2.6 Western blot

Seventy-two hours after transfection, total protein was collected using RIPA and quantified by BCA method (KeyGen). Thirty micrograms of protein were subjected to immunoblot analysis with rabbit anti-BRMS1L antibody (Abcam, 1:1000), followed by HRP-conjugated mouse anti-rabbit secondary antibodies. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as control.

2.7 Statistical analysis

Statistical analysis was performed on SPSS 16.0 and GraphPad Prism 5.0 statistical software. Student’s t-test and one-way ANOVA test were used to investigate the significance of difference. Kaplan–Meier survival analysis was applied to analyze the survival distributions, and the log-rank test was employed to assess the statistical significance between stratified survival groups. Univariate and multivariate Cox regression analysis were used to evaluate prognostic value of BRMS1L in GBM patients. Patient whose BRMS1L expression was lower than the median value was defined as low expression, while higher than the median value was classified into high expression group. A two-sided $p$ value $<$ 0.05 was considered statistically significant.

Figure 3.

BRMS1L is a tumor anti-oncogene in glioma. (A and B) BRMS1L protein is elevated by BRMS1L over-expression vectors both in U87 and U251 glioma cells. (C and D) Transwell assay shows that both U87 and U251 cells’ invasion ability are significantly inhibited by ectopic of BRMS1L expression. NC means negative control vectors treatment group, Vector means BRMS1L over-expression vectors treatment group.

3. Results

3.1 BRMS1L expression is associated with glioma grade

To investigate BRMS1L mRNA expression pattern in glioma, 120 glioma samples (30 grade II, 30 grade III, and 60 grade IV) were analyzed by qRT-PCR, we found that BRMS1L level was negatively correlated with tumor grade ( $p<$ 0.01), as shown in Fig. 1A, BRMS1L expression was lowest in WHO IV glioma, while WHO II glioma showed the highest expression. And this result was further validated in CGGA, TCGA and GSE16011 databases.

3.2 Reduced BRMS1L expression predicts poor survival in GBM patients

To explore the association of BRMS1L expression level with prognosis of GBM patients, Kaplan-Meier survival curve analysis with a log-rank test of the above 60 grade IV GBM patients was employed, as shown in Fig. 2, GBM patients with lower BRMS1L expression (median survival is 263 days) had a worse overall survival time by contrast to those with higher BRMS1L expression (median survival is 432 days) ( $p<$ 0.05). Furthermore, the same result was also observed in the CGGA (median OS: high group 493 days, low group 286.5 days, $p<$ 0.01) and GSE (median OS: high group 0.84 years, low group 0.65 years, $p<$ 0.05) validation sets. Although there’s no statistically significant in the TCGA dataset, overall survival days in high BRMS1L group still showed a decreased trend compared to low group (median OS: high group 419 days, low group 360 days, $p=$ 0.13).

3.3 BRMS1L is an independent prognostic biomarker for GBM patients

To further reveal the prognostic value of BRMS1L in GBM patients, a univariate Cox regression analysis was used. As shown in Table 1, lower BRMS1L expression was a risk factor for GBM patients. Additionally, other factors including ender age ( $\geqslant$ 60), low KPS score ( $<$ 80), subtotal resection, and did not accept radiotherapy $+$ TMZ treatment were all risk factors for GBM patients’ survival days. Next, we performed multivariate Cox proportional hazards analysis incorporating the above factors. The analysis indicated that BRMS1L value was an independent prognostic marker for the overall survival of GBM patients.

3.4 BRMS1L suppresses glioma cell invasion

To assess the functional role of BRMS1L in GBM, gain-of-function assay was done by transfecting BRMS1L overexpression vectors into U87 and U251 glioma cell lines to elevate BRMS1L expression (Fig. 3A and B), and we found that ectopic of BRMS1L suppressed glioma cells’ invasive ability both in U87 and U251 glioma cells (Fig. 3C and D).

4. Discussion

In spite of some minor improvements in progression-free survival (PFS) of GBM by the combination of temozolomide (TMZ) and other targeted treatments, no satisfactory overall survival has been accomplished [12]. Thus, it is an urgent need that we advance glioma profiles understanding and identify novel genetic alterations of glioma, especially in primary and recurrent GBM. This could contribute to a deeper understanding of the characteristics of GBM, and imply potential targets for personalized therapeutic strategies [7, 11]. Aberrant BRMS1L expression plays a crucial role in carcinogenesis of breast cancer [4]. A previous study have also identified the potential role of loss BRMS1L in the pathogenesis of glioma [14]. Thus, we hypothesized that BRMS1L loss might be associated with glioma. To test this hypothesis, we analyzed the expression of BRMS1L in our cohort and other several databases. We found that BRMS1L levels were significantly lower in GBM compared with low grade glioma (WHO grade II and III) in all cohorts. In survival analyses, our cohort also indicated that the down-regulated expression of BRMS1L correlated with poor outcome in GBM patients and could be used as an independent diagnostic marker. Cross-validation from the CGGA and cross-validation from the GSE16011 and TCGA cohorts also confirmed BRMS1L as a significant biomarker for predicting GBM patients’ outcome.

The invasive property of brain-tumor cells makes an important contribution to the ineffectiveness of current treatment modalities [10]. To further understand biological role of BRMS1L in glioma invasive phenotype, we overexpressed BRMS1L in U87 and U251 GBM cells and detected their invasion change. And we found that ectopic expression of BRMS1L inhibited GBM cells’ invasion ability. However, further in vivo model invasion assay of GBM is needed to prove BRMS1L targeting to be a useful therapeutic strategy, which can impede the malignant characteristics of GBM.

In summary, low levels of BRMS1L is associated with glioma progression and plays as an independent poor prognosis biomarker. The up-regulation of BRMS1L suppresses glioma cells’ invasion. All of those indicate that BRMS1L is a novel prognostic biomarker with potential anti-invasion therapeutic implications in GBM.

References

Zeng

A.L.

Yan

Liu

Y.W.

et al., Tumour exosomes from cells harbouring PTPRZ1-MET fusion contribute to a malignant phenotype and temozolomide chemoresistance in glioblastoma, Oncogene 36(38) (2017), 5369–5381.

Zeng

Liu

et al., IDH1/2 mutation status combined with Ki-67 labeling index defines distinct prognostic groups in glioma, Oncotarget 6(30) (2015), 30232–30238.

Wang

Y.A.

et al., Epigenetic activation of WNT5A drives glioblastoma stem cell differentiation and invasive growth, Cell 167(5) (2016), 1281–1295e18..

Gong

X.B.

et al., BRMS1L suppresses breast cancer metastasis by inducing epigenetic silence of FZD10, Nature Communications 5 (2014), 5406.

Zhang

Cheng

Ren

et al., Tumor purity as an underlying key factor in glioma, Clinical Cancer Research 23(20) (2017), 6279–6291.

Binda

Visioli

Giani

et al., Wnt5a drives an invasive phenotype in human glioblastoma stem-like cells, Cancer Research 77(4) (2017), 996–1007.

Furnari

F.B.

Cloughesy

T.F.

Cavenee

W.K.

et al., Heterogeneity of epidermal growth factor receptor signalling networks in glioblastoma, Nature Reviews Cancer 15(5) (2015), 302–310.

Chiang

I.T.

Wang

W.S.

Liu

H.C.

et al., Curcumin alters gene expression-associated DNA damage, cell cycle, cell survival and cell migration and invasion in NCI-H460 human lung cancer cells in vitro, Oncology Reports 34(4) (2015), 1853–1874.

Kros

J.M.

Mustafa

D.M.

Dekker

L.J.

et al., Circulating glioma biomarkers, Neuro-Oncology 17(3) (2015), 343–360.

10.

Rao

J.S.

, Molecular mechanisms of glioma invasiveness: The role of proteases, Nature Reviews Cancer 3(7) (2003), 489–501.

11.

Zeng

et al., Polycomb group expression signatures in the malignant progression of gliomas, Oncology Letters 13(4) (2017), 2583–2590.

12.

Chen

You

et al., Comprehensive portrait of recurrent glioblastoma multiforme in molecular and clinical characteristics, Oncotarget 6(31) (2015), 30968–30974.

13.

Zeng

et al., miR-423-5p contributes to a malignant phenotype and temozolomide chemoresistance in glioblastomas, Neuro-Oncology 19(1) (2017), 55–65.

14.

Pan

Yang

et al., Revealing the potential pathogenesis of glioma by utilizing a glioma associated protein-protein interaction network, Pathology Oncology Research: POR 21(2) (2015), 455–462.

15.

Zhang

Qian

Zhang

et al., Autocrine/paracrine human growth hormone-stimulated MicroRNA 96-182-183 cluster promotes epithelial-mesenchymal transition and invasion in breast cancer, The Journal of Biological Chemistry 290(22) (2015), 13812–13829.

Decreased BRMS1L expression is correlated with glioma grade and predicts poor survival in glioblastoma via an invasive phenotype

Abstract

AIMS:

METHODS:

RESULTS:

CONCLUSION:

Keywords

1. Introduction

2.1 Glioma samples and patients

2.3 BRMS1L expression analysis in the validation datasets

2.4 Cell line, cell culture and transfection

2.5 Invasion assay

Table 1 BRMS1L and other clinicopathological factors affecting 60 GBM patients’ overall survival

2.7 Statistical analysis

3.1 BRMS1L expression is associated with glioma grade

3.2 Reduced BRMS1L expression predicts poor survival in GBM patients

3.3 BRMS1L is an independent prognostic biomarker for GBM patients

3.4 BRMS1L suppresses glioma cell invasion

4. Discussion

References

Table 1
BRMS1L and other clinicopathological factors affecting 60 GBM patients’ overall survival