Identification of survival-associated key genes and long non-coding RNAs in glioblastoma multiforme by weighted gene co-expression network analysis

Abstract

Glioblastoma multiforme (GBM) is the most common and aggressive primary malignant brain tumour. However, the causes of GBM are not clear, and the prognosis remains poor. The aim of the present study was to elucidate the key coding genes and long non-coding RNAs (lncRNAs) associated with the survival time of GBM patients by obtaining the RNA expression profiles from the Chinese Glioma Genome Atlas database and conducting weighted gene co-expression network analysis. Modules associated with overall survival (OS) were identified, and Gene Ontology and pathway enrichment analyses were performed. The hub genes of these modules were validated via survival analysis, while the biological functions of crucial lncRNAs were also analysed in the publicly available data. The results identified a survival-associated module with 195 key genes. Among them, 33 key genes were demonstrated to be associated with OS, and the majority of these were involved in extracellular matrix-associated and tyrosine kinase receptor signalling pathways. Furthermore, LOC541471 was identified as an OS-associated lncRNA, and was reported to be involved in the oxidative phosphorylation of GBM with pleckstrin-2. These findings may significantly enhance our understanding on the aetiology and underlying molecular events of GBM, while the identified candidate genes may serve as novel prognostic markers and potential therapeutic targets for GBM.

Introduction

Glioblastoma multiforme (GBM) is an aggressive carcinoma that was first described in 1800 (1). According to the World Health Organisation classification of tumours of the central nervous system (CNS), GBM is classified as a grade IV malignant glial neoplasm with astrocytic differentiation (2). As one of the most commonly diagnosed malignant CNS tumours, it accounts for 45.6% of primary malignant brain tumours, with an annual incidence of 3.1 cases per 100,000 individuals in the United States (3). GBM is considered to be one of the most malignant primary intracranial tumours and has a dismal prognosis, <5% of patients surviving 5 years after diagnosis (3). Even after microsurgery and adjuvant temozolomide-based radiochemotherapy or radiotherapy alone, the prognosis of patients remains poor, with a median survival of 14 months (4). In recent years, clinical trials of angiogenesis inhibitor therapies have been performed on patients with recurrent GBM; however, no survival benefit was achieved (5). Furthermore, a randomised phase III clinical trial of epidermal growth factor receptor (EGFR) variant III-targeted drugs failed to confirm any survival benefit compared with the control group (6). Therefore, searching for novel candidate genes and studying the underlying mechanisms driving tumourigenesis are important to generate new therapeutic targets.

The development of microarray and high-throughput sequencing technology has provided new methods for investigating the molecular mechanisms underlying tumour behaviour and for screening drug targets. In 2012, the Chinese Glioma Genome Atlas (CGGA) was built, which provides genomic and clinical data for hundreds of samples that have been examined by whole-genome sequencing, whole-exome sequencing, DNA methylation microarray detection, as well as microRNA (miRNA), circular RNA, long non-coding RNA (lncRNA) and mRNA sequencing (7). This database represents a landmark achievement in glioma research in China.

Weighted gene co-expression network analysis (WGCNA) is a powerful method for describing the correlations among genes using microarray or RNA sequencing data. WGCNA can be applied to search for candidate biomarkers or therapeutic targets, as well as phenotype-associated modules or gene clusters based on co-expression networks (8). Compared with other analytical methods, WGCNA has the advantage of summarising and standardising the methods and functions in the R software package. Currently, this systematic method is widely used to study a number of complex diseases, particularly cancer. For instance, Lv et al (9) observed that LINC01314 functions as a tumour suppressor in hepatoblastoma. Furthermore, Clarke et al (10) indicated that KCNK5 was associated with poor outcomes of the basal-like molecular subtype in breast cancer. Using this analysis, the pathways involved in the co-expression network of cancer models and target genes can be identified. However, relevant research is currently lacking in GBM.

The present study aimed to explore the underlying molecular mechanisms, and identify novel prognostic biomarkers and treatment targets of GBM. The expression profiles of lncRNAs and mRNAs in GBM compared with those of normal controls were investigated, and differentially expressed RNAs were identified from the CGGA database. Next, the enriched pathways participating in the tumourigenesis of GBM were determined, and protein-protein interaction (PPI) networks were constructed. Finally, using public databases, potential prognostic biomarkers were confirmed, and the clinical significance and biological functions of survival-associated lncRNAs were identified.

Materials and methods

Data collection

RNA sequencing and clinical data of GBM patients were obtained from the CGGA database (http://www.cgga.org.cn). The gene expression levels were measured in terms of fragments per kilobase of transcript per million mapped reads. Various clinical data were also downloaded from the database, which included the patient gender, age, The Cancer Genome Atlas (TCGA) subtype, overall survival (OS), radiotherapy and chemotherapy details, and the mutation status of the genes isocitrate dehydrogenase (IDH), tumour protein p53 (TP53), EGFR, ATRX and enhancer of zeste homolog 2 (EZH2) (11,12). All cases with pathological diagnosis of GBM were included in the analysis. The exclusion criteria applied in the present study were as follows: i) Histologic confirmation of the diagnosis of any brain tumour type other than primary GBM; ii) history of radiotherapy or chemotherapy prior to histologic diagnosis; iii) patients with missing follow-up records; and iv) missing mutation information for the five aforementioned genes. According to these criteria, a total of 88 GBM samples were selected for inclusion in the current study (Table I). Furthermore, another gene expression and clinical dataset of GBM patients was downloaded from the TCGA database (https://cancergenome.nih.gov/), from which 162 GBM cases were selected. The dataset obtained from TCGA was analysed, and served as the validation dataset.

Table I.

Summary of patient characteristics.

Characteristic	Value
No. of patients	88
Age
≥40 years	73
<40 years	15
Sex
Male	55
Female	33
TCGA subtype
Classical	32
Mesenchymal	36
Neural	7
Proneural	13
OS
≥12 months	43
<12 months	45
History of radiotherapy	59
History of chemotherapy	52
Gene mutation
IDH	12
TP53	44
EGFR	23
ATRX	9
EZH2	13

TCGA, The Cancer Genome Atlas; OS, overall survival; IDH, isoci-trate dehydrogenase; TP53, tumour protein p53; EGFR, epidermal growth factor receptor; EZH2, enhancer of zeste homolog 2.

Construction of the gene co-expression network and identification of preserved modules

First, the expression data profiles from the CGGA were tested to confirm that they were suitable for the analysis. The standard deviation value for all samples of each gene was calculated, and the top 5,000 genes with the lowest standard deviation values were selected for subsequent analysis. Next, the co-expression network of genes was constructed using the WGCNA package in R (8). To calculate the scale-free topology fitting index r² that corresponded to different soft-thresholding parameter β values, functional pickSoftThreshold was used, and the β value was selected if r² reached 0.9. The soft-thresholding power β value was then set to 6, and the minModuleSize was set to 30. Subsequently, the gene expression profile was transformed into an adjacency matrix and a topological overlap matrix (TOM), which was defined as the sum of adjacency between the gene and all other genes for network generation. Next, the corresponding dissimilarity of TOM (dissTOM) was calculated, and dissTOM-based hierarchical clustering was used to produce a hierarchical clustering dendrogram of genes. Modules of clustered genes were then generated using the Dynamic Tree Cut algorithm. The module eigengene was calculated using the function moduleEigengenes, and a number of modules were merged according to a cut-off line for the module dendrogram. The interactions (correlations) of each module were analysed and visualised by heatmaps. To identify modules associated with patient characteristics, the Pearson's correlation test was used to evaluate the correlation of module eigengenes with the clinical traits, OS and mutation status, and correlations with P-values of <0.05 were considered to be statistically significant.

Gene ontology (GO) and pathway enrichment analysis

The functional enrichment of the genes of the identified module was assessed based on GO terms (13) and Kyoto Encyclopaedia of Genes and Genomes (KEGG) pathway (14) annotations. GO term analyses were performed using the DAVID database (https://david.ncifcrf.gov/) and Panther database (http://www.pantherdb.org) (15), which are essential tools for the success of high-throughput gene function analysis. Pathway analysis was also conducted using multiple online databases, including the DAVID database, KEGG pathway database (http://www.genome.jp/kegg) and STRING online database (http://string-db.org) (16). P-values of <0.05 were considered to denote statistically significant differences in GO term enrichment and KEGG pathway analyses, and the false discovery rate was utilised to correct the P-values.

PPI network construction and analysis

To identify the gene-encoded proteins and construct the PPI network of the identified module, the genes were mapped to the STRING database. The results obtained from this database were then imported into Cytoscape software (version 3.6.0; https://cyto-scape.org/) to analyse the interactional associations among the gene-encoding proteins and their degrees in GBM (17). In addition, significant genes from the PPI network complex were selected according to their degree of importance. The corresponding proteins may be the core proteins or key candidate genes that have significant physiological regulatory functions.

Survival analysis and validation of the genes in the TCGA dataset

To confirm the reliability of the identified genes from the CGGA data, GBM data from TCGA were then used to perform validation with the GEPIA database (http://gepia.cancer-pku.cn) (18). Through this database, the expression levels of all genes of interest in GBM and other tumours can be obtained. Furthermore, Kaplan-Meier curves were generated based on the GEPIA database. The OS was estimated using the log-rank test, and P<0.05 was considered to denote statistically significant data.

KEGG analysis of lncRNA-correlated mRNAs in GBM

In TCGA data, mRNAs having a Spearman's correlation with lncRNA of >0.4 were considered to be lncRNA-correlated mRNAs. These were then analysed by KEGG pathway enrichment analysis. A P-value of <0.05 was applied to identify the significant pathways.

Further analysis of candidate lncRNAs in GBM

The genes nearby lncRNAs were analysed by genomic region enrichment of annotations tool (GREAT version 3.0.0) (19). The potential targets of lncRNA were predicted by searching the miRDB database (http://www.mirdb.org).

Results

Gene co-expression network of GBM

To detect and explore the possible biological function of the key survival-associated genes, WGCNA was performed based on the mRNA and lncRNA profiles derived from the CGGA database. According to the exclusion criteria mentioned earlier, RNA sequencing results and the clinical data of 88 GBM samples were downloaded from the CGGA database. For module detection, 5,000 coding and non-coding RNAs were selected for further analysis from the original 21,000 genes according to the standard deviation values. One outlier sample was removed from the sample network. The TCGA subtype, gender, age, OS, radiotherapy and chemotherapy information, and the mutation status of IDH, TP53, EGFR, ATRX and EZH2 were defined as clinical traits (Fig. 1A). Analysis of the network topology was first performed for various soft-thresholding power β values to determine the relative balanced scale independence and mean connectivity of the weighted gene co-expression network. As shown in Fig. 1B, power 6 was the lowest power at which the scale-free topology fitting indices r² reached 0.90; thus, this power was selected in order to produce a hierarchical clustering tree (dendrogram) of the 5,000 genes. In total, 21 modules were identified by hierarchical clustering and dynamic branch cutting, and each module was assigned a unique colour as an identifier (Fig. 1C and D). The largest module contained 826 genes, while the smallest contained 58 genes. The grey module represented a gene set that was not assigned to any of the modules.

Figure 1.

Network visualization plots of weighted gene co-expression network analysis. (A) Dendrogram of samples and heatmap of clinical and molecular traits. (B) Analysis of network topology for different soft-thresholding powers. The higher panel displays the influence of soft-thresholding power (x-axis) on the scale-free fit index (y-axis). The lower panel shows the influence of soft-thresholding power (x-axis) on the mean connectivity (degree; y-axis). (C) Clustering of MEs. Hierarchical clustering of module eigengenes that summarise the modules yielded in the clustering analysis. Joint branches of the dendrogram represent genes that are positively correlated. (D) Dendrogram of selected genes, indicating clusters with dissimilarity based on topological overlap, along with the assigned module colours. ME, module eigengene; CGGA, Chinese Glioma Genome Atlas; TCGA, The Cancer Genome Atlas; OS, overall survival; IDH, isocitrate dehydrogenase; TP53, tumour protein p53; EGFR, epidermal growth factor receptor; EZH2, enhancer of zeste homolog 2.

Identification of the modules significantly associated with survival

To explore the survival significance of the selected modules, correlations between the OS and module eigengenes were analysed. It was observed that three modules (black, brown and light cyan) were positively correlated with OS (P<0.05). Among them, the black module had the lowest P-value. Another five modules (turquoise, blue, purple, tan and grey) were negatively correlated with OS (Fig. 2; P<0.05), with the purple module exhibiting the lowest P-value. According to the previous analysis of the survival significance of modules, the black and purple modules were selected for further analysis with OS, since they had the lowest P-values. Limited progress in targeted therapy for GBM has been made in recent years, and the majority of the therapeutic targets for this disease are oncogenes (20); thus, the purple module was further analysed in the present study, since the genes in purple module are considered to be oncogenes. In total, 195 genes, including 193 protein-coding and 2 non-coding genes, were identified in the purple module. The locations of certain of these identified genes on human chromosomes are displayed in the Circos plot in Fig. 3.

Figure 2.

Module-trait associations were evaluated according to the correlations between MEs and clinical traits. Each row corresponds to an ME, while each column corresponds to a trait. Each cell contains the corresponding correlation (top line) and P-value (bottom line). The table is colour-coded by correlation, according to the coloured legend. ME, module eigengene; TCGA, The Cancer Genome Atlas; OS, overall survival; IDH, isocitrate dehydrogenase; TP53, tumour protein p53; EGFR, epidermal growth factor receptor; EZH2, enhancer of zeste homolog 2.

Figure 3.

Identification of significant genes and their chromosomal locations in the purple module. The Circos plot shows the significant genes on the human chromosomes. From the outside and inwards, the first layer of the Circos plot is a chromosome map of the human genome; the black and white bars represent chromosome cytobands, while red bars represent centromeres. Due to limited space, a number of the genes are labelled in the second circle. In the third layer, coding and non-coding RNAs are marked in blue and red, respectively. The fourth layer represents the GS values of genes, while the innermost circle indicates the P-value of the GS. The network in the centre of the plot represents the core network. GS, gene significance.

Enrichment analysis of coding genes in the purple module

The functions and pathway enrichment of the candidate genes in GBM identified from the purple module were analysed using multiple online databases, including the DAVID, Panther, KEGG pathway and STRING databases. The GO functional enrichments of genes were determined with the DAVID and Panther databases, with a P-value of <0.05 indicating statistical significance of the data. The protein-coding genes in the purple module were mapped to the GO database to determine their potential functions. GO terms were divided into three functional groups, including biological processes (BP), cell composition (CC) and molecular function (MF). The enriched GO terms for the genes are presented in Fig. 4. For the candidate genes in the purple module, the top three enriched GO terms in each category were as follows: Extracellular structure organisation, extracellular matrix (ECM) organisation and multicellular organismal macromolecule metabolic process in the BP category; ECM, extracellular region and extracellular space in the CC category; and collagen binding, ECM structural constituent and growth factor binding in the MF category (Fig. 4A and B). These results revealed that the majority of the genes were significantly enriched in extracellular structure, binding, cell parts and cell growth.

Figure 4.

GO analysis and significantly enriched GO terms of significant genes in glioblastoma multiforme. GO analysis classified the significant genes into three categories: Biological process, molecular function and cellular component. (A) Top 50 significantly enriched GO terms ordered by (A) their function and (B) gene counts. GO, Gene Ontology.

Subsequently, functional and signalling pathway enrichment of genes in the purple module was performed using the online databases DAVID, STRING and KEGG. The top enriched KEGG pathways for the candidate genes included ECM organisation, collagen formation, integrin cell surface interactions, degradation of the ECM, collagen biosynthesis and modifying enzymes, ECM proteoglycans, assembly of collagen fibrils and other multimeric structures, ECM-receptor interaction, collagen degradation and signalling by receptor tyrosine kinases (Fig. 5). The majority of the identified pathways were cancer-associated signalling pathways.

Figure 5.

Significantly enriched pathways of significant genes in glioblastoma multiforme, included in the purple module. The top 20 enriched pathways are presented.

PPI network analysis of survival-associated coding genes in GBM

Data from the STRING database revealed that several of the genes interacted with each other. In total, 113 of the 193 candidate protein-coding genes were filtered to form the PPI network complex (Fig. 6). The network contained 113 nodes and 648 edges. Among the 113 nodes, the most significant 19 hub node genes were identified using a degree of ≥10 as the filtering criterion. These genes were COL1A1, COL1A2, DCN, COL3A1, FN1, MMP9, COL6A1, COL5A2, COL6A2, ITGA5, FBN1, CTGF, ITGA4, PDGFRB, LUM, PPIB, MMP14, ITGA11 and COPB1 (Table II). Among these genes, COL1A1 exhibited the highest node degree (degree =30), and FN1 shares the highest degree with 33 validated coding genes.

Figure 6.

PPI network of the significant genes in glioblastoma multiforme. Using the STRING online database, 113 genes were selected and used to construct the PPI network. PPI, PPI, protein-protein interaction.

Table II.

Degrees of the top 19 key genes in the protein-protein interaction network.

Gene symbol	Degree
COL1A1	30
COL1A2	25
DCN	24
COL3A1	22
FN1	20
MMP9	20
COL6A1	18
COL5A2	16
COL6A2	16
ITGA5	16
FBN1	16
CTGF	14
ITGA4	13
PDGFRB	12
LUM	11
PPIB	11
MMP14	11
ITGA11	11
COPB1	10

Validation of the survival-associated genes in TCGA dataset

To confirm the reliability of the 193 survival-associated coding genes identified from the CGGA, GBM datasets were also downloaded from TCGA database (including 162 GBM samples), and the RNA sequencing data and survival information of these datasets were subjected to Kaplan-Meier survival analysis. The results revealed significantly different OS between the high and low expression groups for 33 genes in the TCGA GBM, including ADAM12, B4GALT7, CD248, CHPF2, COL6A1, COL6A2, CYGB, DCBLD2, DERL2, DUSP6, EFNB2, EMILIN1, EPHA2, FAP, FBLN1, FN1, FZD1, HOXB2, HSP90B1, IGFBP4, LAMB1, LOXL1, MMP11, NID2, P4HB, PCOLCE, PDIA3, PLOD1, PMAIP1, PPIB, RARRES1, TBL2 and THY1 (P<0.05; Fig. 7). Therefore, the expression levels of these 33 genes may be used as predictors of OS in GBM patients, indicating that they may be candidate genes involved in GBM that deserve further investigation.

Figure 7.

Survival rate calculated by the Kaplan-Meier survival curve in patients separated according to the median expression level of each gene, using data collected from The Cancer Genome Atlas. The top 8 of the 33 key genes are presented in the figure, while the analysis of the remaining genes is not shown. Survival curves of (A) ADAM12, (B) B4GALT7, (C) CD248, (D) CHPF2, (E) COL6A1, (F) COL6A2, (G) CYGB and (H) DCBLD2 are shown.

Validation and survival analysis of candidate lncRNAs in GBM

According to the co-expression network, two lncRNAs, namely LOC541471 and LOC284494, were identified as survival-associated key non-coding genes in GBM. Next, the genes nearby these two lncRNAs were analysed. The results demonstrated that 8 genes were located near the transcription start site (Fig. 8A and B). To confirm the reliability of the identified lncRNAs, TCGA GBM data were subsequently used to perform validation by GEPIA. As shown in Fig. 8C, LOC541471 was significantly overexpressed in the GBM data-sets obtained from TCGA. Next, the expression levels of these two lncRNAs in other types of cancer were explored using TCGA. It was observed that LOC541471 was overexpressed in 17 other cancer types (Fig. 8D), while LOC284494 was overexpressed in 3 other cancer types, as compared with their corresponding normal tissues (data not shown). Subsequently, survival analysis was performed in the TCGA GBM data, and a significant association with survival was observed only for LOC541471 (Fig. 8E). Therefore, LOC541471 was regarded as a core lncRNA in the network and warrants further investigation.

Figure 8.

Location, expression and survival analysis of candidate lncRNAs in GBM data obtained from TCGA. (A) Distribution of the distances to the nearest transcription start site of the two lncRNAs. (B) Absolute distance to transcription start site. (C) Expression of LOC541471 in GBM and normal brain tissue. (D) Expression of LOC541471 in other tumours and corresponding normal tissues. (E) Kaplan-Meier analysis of the overall survival of patients with high or low LOC541471 expression levels in the TCGA dataset. lncRNA, long non-coding RNA; GBM, glioblastoma multiforme; TCGA, The Cancer Genome Atlas.

Functional analysis of lncRNA LOC541471 in GBM

As mentioned earlier, LOC541471 was found to be the core lncRNA in the co-expression network. To understand how this lncRNA is involved in GBM, a deeper insight into the GBM expression data from TCGA was required. Spearman's correlation of LOC541471 was calculated with ~24,300 coding genes in 156 GBM patients. It was observed that pleckstrin-2 (PLEK2) is the mRNA exhibiting the highest correlation index (r=0.68) with lncRNA LOC541471, and this gene is located on human chromosome 3p24.1 (Fig. 9A and B). According to the TCGA data, PLEK2 is significantly overexpressed in GBM tissues as compared with the normal tissue (Fig. 9C). Next, a total of 436 mRNAs that had a Spearman's correlation with LOC541471 of >0.4 were further examined by KEGG pathway analysis. Pathways were considered as enriched at a P-value of <0.05 (Fig. 9D), and the most enriched pathway was oxidative phosphorylation (OXPHOS). Finally, the expression data of GBM samples obtained from TCGA and CGGA were analysed, and the expression of LOC541471 in IDH1 wild-type was observed to be much higher compared with that in the IDH1 mutant group (Fig. 10).

Figure 9.

Functional analysis of lncRNA LOC541471 in GBM. (A) Schematic representation of the location of the PLEK2 gene on human chromosome 14q24.1. (B) Spearman's correlation analysis examining the association of LOC541471 with PLEK2 in the GBM dataset obtained from The Cancer Genome Atlas. (C) Expression of PLEK2 in GBM and normal brain tissue. (D) Top 20 significantly enriched pathways of all protein-coding genes with Pearson's correlation with LOC541471 of >0.4. lncRNA, long non-coding RNA; GBM, glioblastoma multiforme; PLEK2, pleckstrin-2.

Figure 10.

Association between LOC541471 and the IDH1 subtype of GBM. (A) Expression data of LOC541471 in the GBM dataset obtained from (A) TCGA and (B) CGGA. ^***P<0.001. GBM, glioblastoma multiforme; CGGA, Chinese Glioma Genome Atlas; TCGA, The Cancer Genome Atlas; IDH1, isocitrate dehydrogenase 1.

Cytoplasmic lncRNAs can act as competing endogenous RNAs to modulate the functions of miRNAs (21). By searching the miRDB database (22), it was observed that LOC541471 was located in the cytoplasm, and 37 possible targets miRNAs were identified (Table III). The top 5 target miRNAs were hsa-miR-548t-3p, hsa-miR-548ap-3p, hsa-miR-548aa, hsa-miR-4288 and hsa-miR-3138. Taken together, these results predicted the potential mechanism of LOC541471; however, further studies are required to demonstrate this additional mechanism of LOC541471 in GBM.

Table III.

Predicted miRNAs targeted by LOC541471.

Target rank	Target score	miRNA name
1	84	hsa-miR-548t-3p
2	84	hsa-miR-548ap-3p
3	84	hsa-miR-548aa
4	82	hsa-miR-4288
5	77	hsa-miR-3138
6	76	hsa-miR-93-3p
7	74	hsa-miR-4508
8	71	hsa-miR-939-3p
9	69	hsa-miR-5585-5p
10	68	hsa-miR-4668-5p
11	66	hsa-miR-4278
12	66	hsa-miR-4492
13	65	hsa-miR-2110
14	65	hsa-miR-519d-5p
15	64	hsa-miR-3148
16	63	hsa-miR-1246
17	62	hsa-miR-4456
18	62	hsa-miR-1178-3p
19	61	hsa-miR-3945
20	61	hsa-miR-4306
21	60	hsa-miR-877-3p
22	60	hsa-miR-1302
23	59	hsa-miR-7975
24	59	hsa-miR-1252-5p
25	59	hsa-miR-6853-3p
26	58	hsa-miR-6794-5p
27	58	hsa-miR-4716-3p
28	58	hsa-miR-146b-3p
29	57	hsa-miR-6754-5p
30	57	hsa-miR-3622a-5p
31	57	hsa-miR-542-3p
32	55	hsa-miR-651-3p
33	55	hsa-miR-1296-5p
34	55	hsa-miR-1231
35	54	hsa-miR-136-5p
36	54	hsa-miR-874-3p
37	54	hsa-miR-4267

Discussion

GBM is the most common and aggressive primary brain tumour type. The therapeutic options consist of microsurgery, and treatment with radiotherapy plus adjuvant chemotherapy with temozolomide and targeted drugs. Despite a large number of basic and clinical studies that have revealed the mechanisms underlying the formation and development of GBM in the past decades, the prognosis for this disease remains considerably poor (23-25). Therefore, searching for new candidate genes is an important part of studying the disease tumourigenesis and would help identify novel therapeutic targets. A number of previous studies have used regulatory network methods to analyse gene expression data, and these approaches focus not only on the differences, but also on the correlations between gene expression datasets (26-28). Such studies have provided numerous relevant molecular mechanisms and valuable biomarkers for GBM. However, these methods use a hard threshold to determine the correlation between genes and do not consider the changes in the correlation intensity between genes under different conditions. Therefore, more appropriate methods are required.

WGCNA is a novel statistical method analysing gene correlations that is based on scale-free topology and construction of a weighted network via soft thresholds. It is not only used to construct gene networks and detect sub-networks, but also to identify hub genes and select candidate biomarker genes (8). In general, module checking in WGCNA requires a knowledge-independent process, and has been widely and successfully applied in biological function analysis in various diseases (29-31). Furthermore, it is well known that the mutation of genes contributes to GBM tumourigenesis; however, the role of lncRNAs has not yet been fully researched.

In the present study, a comprehensive analysis of lncRNA and mRNA profiling data of GBM patients obtained from the public database CGGA was performed. Subsequently, the modules associated with the OS of GBM cases were identified, the coding genes of the module were analysed by GO and KEGG enrichment analyses, and the PPI network was constructed. Next, genes associated with survival were identified by Kaplan-Meier survival analysis. Finally, the core lncRNAs and their biological functions were identified, and their expression was validated using the GBM dataset obtained from TCGA. In conclusion, our work has identified a gene set involving the survival and tumourigenesis of GBM in which lncRNAs play a critical role.

The gene expression profile of GBM obtained from CGGA was initially examined in the present study. A total of 5,000 genes with the lowest standard deviation values were selected, since these were considered as more likely to provide good data quality. Next, the purple module was identified by WGCNA, in which genes were significantly associated with OS of GBM patients.

Subsequent GO and KEGG pathway analyses of mRNAs in the purple module revealed that the majority of the genes were significantly enriched in extracellular structure, binding, cell part and cell growth. They were also involved in pathways relevant to tumour progression and migration, such as ECM organisation, collagen formation, integrin cell surface interactions, degradation of the ECM, collagen biosynthesis and modifying enzymes, and ECM-receptor interactions. These results can explain the correlations between the purple module and the OS of GBM patients.

According to the present study analysis, the top enriched GO terms were associated with ECM. It is well known that ECM is considered to be the key component in the spread of gliomas throughout the brain (32). In addition, the ECM micro-environmental composition (33) and its mechanical force (34) influence the migration of glioma cells. ECM-binding proteins were considered to be indispensable elements of GBM migration (35). Furthermore, the degradation of ECM-components involves various proteases and hyaluronidases, which are important not only in promoting glioma cell migration by ECM decomposition, but also in releasing various growth and chemotactic factors that stimulate growth, survival and migration of GBM (36).

The current study identified 195 genes in the purple module, which were markedly associated with OS. Additionally, there were 33 candidate genes associated with the OS of patients with GBM according to TCGA data, and these may serve as prognostic biomarkers for GBM. Data from the STRING database further revealed that 113 genes of the 193 candidate protein-coding genes were filtered into the PPI network complex. Among these candidate genes, COL1A1 exhibited the highest degree in the network. COL1A1 encodes the major component of type I collagen, which is the fibrillar collagen found in the majority of connective tissues. Previous studies have reported that COL1A1 is upregulated in the microvasculature of proliferating GBMs (37). In addition, COL1A1 may be considered for use in stratifying patients with GBM into subgroups according to the risk of recurrence at diagnosis, as well as for prognostic and therapeutic evolution (38). These previous observations are consistent with the current research findings, suggesting that COL1A1 may also be a potential target for GBM therapy.

The current study also identified that FN1 shares the highest degree with 33 validated coding genes. Fibronectin is a high-molecular-weight (~440 kDa) glycoprotein of the ECM that binds to membrane-spanning receptor proteins, known as integrins (39). Researchers have proven that FN1 is upregulated by TWIST1, which is known to promote epithelial-mesenchymal transition and/or GBM invasion (40). Furthermore, FN1 is associated with glioblastoma recurrence and can be regarded as a target for antiangiogenic therapy (41). The current study findings suggest that COL1A1 and FN1 are associated with migration, invasion, angiogenesis, recurrence and OS in GBM patients. Thus, these genes may serve important roles in the tumourigenesis of GBM.

The important roles of certain lncRNAs in GBM were clearly demonstrates in the current study. However, understanding the functions of these lncRNAs is challenging, since numerous lncRNAs are not included in the public databases. In the present study, it was demonstrated that LOC541471 was highly correlated with PLEK2 in GBM. Therefore, the biological functions of the lncRNA LOC541471 were further examined, and it was observed that it participates in the regulation of biological networks. KEGG pathway enrichment analysis for all protein-coding genes exhibiting Spearman's correlations with lncRNA LOC541471 of >0.4 revealed that the most enriched pathway was OXPHOS. OXPHOS is the metabolic pathway in which cells use enzymes to oxidise nutrients, thereby releasing energy that is used to produce adenosine triphosphate (42). This pathway serves an important role in the energy supply of GBM. Researchers have reported that IMP2 controls oxidative phosphorylation and is crucial for preserving glioblastoma cancer stem cells (43). Additionally, OXPHOS complexes may be clearly altered in GBM compared with normal brain tissue (44). Research progress in targeted therapies also indicates that the simultaneous targeting of glycolysis and OXPHOS is highly effective in blocking GBM tumourigenic phenotypes (45). Therefore, the present study results revealed that LOC541471 may serve a core role in the tumourigenesis of GBM, and may be a novel oncogene worth further study.

It should be noted that the current study examined a limited number of cases. The CGGA database only discloses a small part of the data, and these include gliomas from Grade I to Grade IV, as well as certain recurrent GBM cases. Therefore, only 88 cases met the requirements of the present study. It was attempted to add more data from other databases, however, standardisation of data from different sequencing platforms is not possible. Therefore, these 88 cases were selected for analysis. As more data become available in the CGGA database, more cases will be included in future studies.

In conclusion, using GBM data from the CGGA database and integrated WGCNA, a survival-associated gene module was identified, and 195 candidate key genes were obtained from this gene module. Among them, 33 key genes were proven to be correlated with OS, and the majority of the genes were involved in pathways associated with the ECM, ECM proteoglycans and tyrosine kinases receptor signalling. Furthermore, the lncRNA LOC541471 was identified as an OS-associated lncRNA, and appeared to serve a role in the OXPHOS of GBM through the PLEK2 gene. These findings may significantly enhance our understanding on the aetiology and underlying molecular events of GBM, and these candidate genes and pathways may serve as novel prognostic markers and potential therapeutic targets for GBM.

Funding

This study was supported by the National Natural Science Foundation of China (grant nos. 81472661 and 81872048), the National Key Research and Development Program of China (grant no. 2016YFA0500303), the Beijing Municipal Natural Science Foundation (grant no. 7161004) and the Beijing Municipal Administration of Hospitals Clinical Medicine Development of Special Funding Support (grant no. ZYLX201608).

Availability of data and materials

The analysed datasets generated during the study are available from the corresponding author on reasonable request.

Authors' contributions

XC, YSo and LZ conceived and designed the experiments. CP, LK and XX performed the GO and PPI analyses. CX, YSu, YG collected and performed the WGCNA. ZZ, WZ and LH constructed the Circos plot and other figures. XC, YSo and LZ wrote the manuscript. All authors read and approved the final manuscript.

Ethics approval and consent to participate

Not applicable.

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Abbreviations: WGCNA

weighted gene co-expression network analysis; GBM, glioblastoma multiforme; lncRNA, long non-coding RNA; GO, Gene Ontology; CNS, central nervous system; CGGA, Chinese Glioma Genome Atlas; PPI, protein-protein interaction; OS, overall survival; TCGA, The Cancer Genome Atlas; TOM, topological overlap matrix; KEGG, Kyoto Encyclopaedia of Genes and Genomes; BP, biological processes; CC, cell composition; MF, molecular function; ECM, extracellular matrix; OXPHOS, oxidative phosphorylation