ITGB3BP is a potential biomarker associated with poor prognosis of glioma

Abstract Despite the growing recognition of ITGB3BP as an essential feature of various cancers, the relationship between ITGB3BP and glioma remains unclear. The main aim of this study was to determine the prognostic and diagnostic value of ITGB3BP in glioma. RNA‐Seq and microarray data from 2222 glioma patients were included, and we found that the expression level of ITGB3BP in glioma tissues was significantly higher than that in normal brain tissues. Moreover, ITGB3BP can be considered an independent risk factor for poor prognosis and has great predictive value for the prognosis of glioma. Gene Set Enrichment Analysis results showed that ITGB3BP contributes to the poor prognosis of glioma by activating tumour‐related signalling pathways. Some small‐molecule drugs were identified, such as hexestrol, which may specifically inhibit ITGB3BP and be useful in the treatment of glioma. The TIMER database analysis results revealed a correlation between the expression of ITGB3BP and the infiltration of various immune cells in glioma. Our findings provide the first evidence that the up‐regulation of ITGB3BP correlates with poor prognosis in human glioma. Thus, ITGB3BP is a potential new biomarker that can be used for the clinical diagnosis and treatment of glioma.


| INTRODUC TI ON
Glioma is the most common malignant disease of the central nervous system, originating from neural stem cells with significant heterogeneity. 1 Due to the malignant characteristics of glioma, such as infiltration and rapid growth, glioma patients often have a higher disability and mortality rate. Therefore, glioma poses a severe threat to human health. 2 Presently, surgery combined with radiotherapy and chemotherapy is the primary treatment for reducing the disability rate and extending the survival time of glioma patients. 3 However, because the tumour tissue boundary is unclear, the surgical resection rate is low, and the recurrence rate is high, leading to poor overall treatment outcomes. In recent years, studies on immune checkpoints have received the most attention, but the long-term survival rate and cure rate of glioma still have not been effectively improved. [4][5][6] This is mainly due to the specific pathogenesis of glioma, which remains unknown. Therefore, it is of great clinical significance to gain insight into the pathogenesis of glioma, optimize the treatment plan and identify potential clinical treatment targets.
Recently, many biomarkers have been associated with glioma diagnosis, treatment and prognosis. O6-methylguanine-DNA methyltransferase (MGMT) is a DNA damage repair protein. Strict regulation of MGMT methylation can reduce MGMT expression levels and promote cell apoptosis, improving responsiveness to temozolomide (TMZ) and moderately improving the survival rate. [7][8][9] Additionally, the co-deletion of 1p/19q and IDH1/2 mutations in patients with low-grade glioma indicate a better clinical prognosis; however, the effect of these molecular events in glioblastoma (GBM) is unclear. 10,11 Moreover, IDH mutations usually occur in low-grade glioma (LGG) and secondary glioblastoma, which may be related to the effect of d-2hg on DNA demethylase, and often occur simultaneously with a p53 mutation or 1p/19q co-deletion. 12,13 In addition, ATP-binding cassette (ABC) transporters, nestin, PI3K and CD133 have also been reported as glioma biomarkers. 14-17 Although many biomarkers have been identified, because of the variety and large number of glioma clinical subtypes, there are no specific biomarkers with diagnostic and prognostic value for each subtype. Therefore, there is an urgent need to identify specific biomarkers.
Invasion and metastasis are the hallmarks of malignant tumours, and the tumour microenvironment (TME) is the main factor mediating tumour metastasis. 18 Integrin-binding proteins play a pivotal role in the TME and regulate the occurrence and development of tumours through various signalling pathways. Studies have shown that the integrin subunit β3-binding protein (ITGB3BP) is significantly up-regulated via the TGFβ pathway, promoting the metastasis and invasion of breast cancer cells under hypoxic conditions, and is also an important element leading to poor breast cancer prognosis. 19,20 Moreover, the co-expression of ITGB3BP and cyclin-dependent kinase inhibitor 2A (CDKN2A) is very common in human solid tumours, particularly in bladder cancer. 21 In addition, ITGB3BP, an anti-apoptotic gene, is involved in cell adhesion, and its expression is up-regulated in colorectal cancer. 22 From the evidence above, it is clear that ITGB3BP, as a pathogenic gene, plays an important regulatory role in the pathological process of a variety of tumours.
However, to date, there are no studies on the relationship between ITGB3BP and glioma or the biological function of ITGB3BP in glioma.
Glioma research is one of the leading research directions for neuroscience surgeons. Therefore, we collected thousands of tissue samples from multiple databases for comprehensive analysis to reveal the regulatory effect of ITGB3BP in the pathological process of glioma. In this study, we investigated the biological function of ITGB3BP in glioma and its relationship with survival prognosis for the first time. Our findings provide the first evidence that ITGB3BP is a glioma oncogene that can affect the survival time of patients through the cell cycle and other signalling pathways; thus, targeting ITGB3BP may be a promising treatment strategy.

| Data collection
The Chinese Glioma Genome Atlas (CGGA; http://www.cgga.org. cn/) database is used for data storage and analysis and includes genomic and detailed clinical data of thousands of glioma patients.
We compiled data from 748 patients in the CGGA RNA-Seq database and 268 patients in the CGGA microarray database. Tables S1 and S2 present the detailed clinical data.
The Cancer Genome Atlas (TCGA, http://cance rgeno me.nih. gov/) database collects various cancer data, allowing researchers to forge a deeper understanding of cancer and establish a foundation for accurate diagnosis and treatment. 23 We compiled data from 653 glioma patients from TCGA RNA-Seq database. Table S3 presents the specific clinical information.
Gene Expression Omnibus (GEO, https://www.ncbi.nlm.nih.gov/ geo/) includes high-throughput gene expression data referred to by worldwide research establishments and is the largest and most comprehensive public gene expression data resource. 24 We retrieved seven glioma datasets from GEO (GSE4290, GSE50161, GSE4412, GSE43378, GSE50025, GSE74187 and GSE83300). GSE4290 contained data from 77 glioma tissues and 23 normal brain tissues, GSE50161 contained data from 34 glioma tissues and 13 normal brain tissues, and the remaining five datasets contained data from 279 glioma samples. These datasets were used to verify the difference in ITGB3BP expression between glioma and normal tissues and the relationship between the expression level of ITGB3BP and the survival prognosis of patients with glioma. 25 Data on the expression of ITGB3BP in various human tumour tissues were obtained from the Gene Expression Profiling Interactive Analysis (GEPIA, http://gepia.cance r-pku.cn/.) database, which is a public database recently developed by a Chinese research group to analyse gene expression profiles of cancer and normal tissues, bridging the gap in cancer genomics data and supporting all expression analysis data at the isomer level. 26 The GEPIA data in this study included that from 163 high-grade glioma (HGG) samples and 207 normal brain samples.
The Human Protein Atlas (HPA; http://www.prote inatl as.org/) database provides a complete map of protein expression in all major tissues and organs of the human body and has the highest level of specificity, versatility and reproducibility among related databases. 27 We examined the protein expression of ITGB3BP in normal brain tissue and in HGG and LGG to understand the variation in protein expression in the collected tumour samples.

| Gene set enrichment analysis
Gene Set Enrichment Analysis (GSEA) is a tool for analysing microarray data at the whole-genome level. GSEA identifies differentially expressed genes among whole-genome datasets, allowing researchers to interpret the microarray results more efficiently and accurately. 28,29 We classified the data from the CGGA RNA-Seq, CGGA microarray and TCGA RNA-Seq digital datasets into groups with high and low ITGB3BP expression levels. GSEA 4.0.jar was used to investigate the signalling pathways related to ITGB3BP. The number of permutations was set at 1000, and 'KEGG cell signalling pathways' was selected for the gene data library.

| Target drug explanation
The connectivity map (CMap) is a unique pharmacogenomic database devoted to revealing the functional relationships among drugs, genes and diseases through changes in gene expression. Using the GPL570 platform, we transformed genes (10 positively correlated and 10 negatively correlated) related to ITGB3BP into gene probe information via co-expression analysis. Drugs corresponding to the gene probe files were searched using https://porta ls.broad insti tute. org/cmap/. p < 0.01 and enrichment <−0.75 were set as the thresholds to identify inhibitors of ITGB3BP.

| Meta-analysis
We conducted a systematic search from databases such as PubMed, Web of Science and Embase and found no prior research report on the relationship between ITGB3BP and the prognosis of glioma.
Since this study explored the prognostic role of ITGB3BP in glioma for the first time, we could not obtain relevant data from published articles. Therefore, we used meta-analysis to reveal the overall prognostic significance of ITGB3BP in glioma patients from the CGGA RNA-Seq, CGGA microarray, TCGA RNA-Seq and GEO datasets.
The hazard ratio (HR) and 95% confidence interval (CI) were used to evaluate the correlation between the expression of ITGB3BP and the survival prognosis of patients with glioma. In addition, the heterogeneity between different data sets was evaluated using the Q test (I 2 statistics). When I 2 was <50%, the fixed-effects model was used for meta-analysis; otherwise, the random-effects model was implemented.

| Analysis of immune infiltration
Tumour Immune Estimation Resource (TIMER, https://cistr ome. shiny apps.io/timer/), based on RNA-Seq expression profile data, is widely used to detect the infiltration of immune cells in tumour tissues. 30 The database uses high-throughput sequencing data to analyse the infiltration of immune cells (CD8 + T cells, B cells, neutrophils, CD4 + T cells, dendritic cells and macrophages) in tumour tissues. We used the TIMER 'gene' module to evaluate the relationship between the level of ITGB3BP expression and the level of immune cell infiltration of glioma. We also analysed the relationship between ITGB3BP and immune escape genes (CD274, PDCD1 and PDCD1LG2) to explore whether ITGB3BP affects the prognosis of glioma patients by mediating immune escape.

| Cell culture and tissue preparation
The human glioma cell lines (T98, U251 and A172) and the corre-

| RT-qPCR
RT-qPCR was used to detect the expression of ITGB3BP in glioma cells and tissues. Total RNA Kit I (Omega, Biotek) and Trizol Reagent (Thermo Fisher Scientific) were used to extract total RNA from cells and tissues respectively. RNA reverse transcription was performed using NovoScript Plus All-in-one 1st Strand cDNA Synthesis SuperMix (Novoprotein). NovoStart SYBR qPCR SuperMix Plus

| Statistical analysis
Perl and R (v4.0.5) were used for all data analyses in this study. The limma package was used to analyse the ITGB3BP expression levels in glioma and normal brain tissues, and the Kaplan-Meier method was used to identify the association between ITGB3BP levels and OS in glioma patients. Univariate and multivariate analyses were performed using Cox or multi-Cox regression, respectively, to determine whether ITGB3BP could be used as an independent predictor of glioma prognosis. Cox regression was used for the receiver operating characteristic (ROC) and area under the curve (AUC) analyses. The Wilcoxon or Kruskal-Wallis tests were used to investigate the link between the clinical characteristics or prognosis of tumour patients and the aberrant expression of ITGB3BP.
The co-expression of ITGB3BP with other genes was evaluated using the cor.test function in R. Statistical significance was set at p < 0.05.

| Aberrant overexpression levels of ITGB3BP were observed in glioma
To determine the expression level of ITGB3BP in glioma, we first analysed the mRNA expression of ITGB3BP in the GEPIA database ( Figure 1A). The results showed that the expression level of ITGB3BP in GBM was significantly increased (GBM: 163; normal brain tissue: 207). To compile credible results, we further used the data in the GEO dataset to compare the expression levels of ITGB3BP in glioma tissues and normal tissues. This finding was unexpected and suggested that the average expression level of ITGB3BP in glioma tissue was higher than that in normal brain tissue ( Figure 1B,C). We also analysed the expression level of ITGB3BP protein in glioma using immunohistochemistry (IHC) data. The protein level was not significantly increased in glioma samples compared to that in normal brain tissue samples ( Figure S1). Finally, to further validate the changes in the mRNA expression level of ITGB3BP in glioma, we used RT-qPCR to verify the expression of ITGB3BP in glioma cell lines (T98, U251 and A172) and tissues. As predicted, the mRNA expression level of ITGB3BP in glioma cell lines and tissues was significantly up-regulated compared to that in the corresponding normal cell lines (HA) and normal tissues ( Figure 1D,E). In general, these observations indicate that the mRNA expression level of ITGB3BP in glioma tissues is significantly up-regulated, suggesting that ITGB3BP may act as a tumour promoter and play a significant role in tumourigenesis. The expression level of ITGB3BP in secondary and recurrent glioma was significantly higher than that in primary glioma ( Figure 2B).

|
ITGB3BP was also positively correlated with age and chemotherapy ( Figure 2C,E), and the expression level of ITGB3BP in grade IV glioma (glioblastoma) was significantly higher than that in grade III glioma (anaplastic astrocytoma, anaplastic oligodendroglioma and anaplastic oligoastrocytoma), which in turn showed higher ITGB3BP expression levels than grade II glioma (astrocytoma, oligodendroglioma and oligoastrocytoma) ( Figure 2G). However, the expression level of ITGB3BP in the 1p19q co-deletion and IDH mutation group was significantly lower than that in the wild-type group ( Figure 2D,F). These results suggest that ITGB3BP is associated with poor prognosis in patients with glioma.

| Poor prognosis of glioma was associated with abnormal overexpression of ITGB3BP
To confirm the effect of high expression of ITGB3BP on the prognosis of patients, we obtained three datasets from the CGGA and TCGA databases and divided them into two groups according to the expression level of ITGB3BP. Kaplan-Meier survival curves were drawn to identify the OS of each group, as shown in Figure 3A-C.
We found that the OS of the up-regulated group was lower than that of the down-regulated group (p < 0.001). This indicates that the up-regulation of ITGB3BP is related to poor outcome. Subsequently, ROC curves were plotted to determine whether the expression level of ITGB3BP has clinical diagnostic value for glioma. As shown in

| GSEA revealed the effect of high expression of ITGB3BP on the cell signalling pathway
Based on the above results, we clarified that the overexpression of ITGB3BP can lead to poor prognosis in glioma patients; however, the mechanism by which ITGB3BP participates in regulating the pathological progress of glioma needs to be further explained. GSEA is one of the most common bioinformatics methods that can provide the biological regulatory function of target genes. Therefore, we used GSEA to evaluate three different datasets from the CGGA and and Table 1).

| Functional correlation of ITGB3BP and other genes with potential therapeutic compounds
To further explore the functions of ITGB3BP, we investigated genes co-expressed with ITGB3BP using correlation analysis. As shown in Figure S2 and Table 2 (Table 3), and their molecular and structural formulas were identified at https://pubch em.ncbi.nlm.nih.gov/ (Figure 6).
These compounds are expected to become new treatments for glioma in the future.

| Meta-analysis verified the prognostic risk of ITGB3BP in patients with glioma
To clarify the impact of ITGB3BP on the survival prognosis of patients with glioma, we further used a meta-analysis based on publicly available data for verification. However, since no previous studies have reported the relationship between high ITGB3BP expression and survival time in glioma patients, we could not incorporate any published data into the meta-analysis. Therefore, we queried the original data for survival time and survival status from the public databases and incorporated them into the metaanalysis to obtain more evidence to support our research results that ITGB3BP is a risk factor for glioma patients.

| Relationship of ITGB3BP with immune cells
We explored the correlation between ITGB3BP expression levels and immune cell infiltration using the TIMER database. The results showed that ITGB3BP expression was positively correlated with

| DISCUSS ION
The main characteristics of tumours are increased proliferation, angiogenesis, migration and invasion. 31 Presently, studies have confirmed that ITGB3BP is related to the pathogenesis of various tumours [19][20][21][22] ; however, the relationship between ITGB3BP and glioma has not been reported. Therefore, this study aimed to analyse the clinical and prognostic effects of ITGB3BP on glioma. For the first time, we revealed a high expression level of ITGB3BP in glioma from the GEPIA, GEO and HPA databases. Further, we confirmed by RT-qPCR that ITGB3BP expression was significantly increased in glioma cells and tissues. The above evidence demonstrated that ITGB3BP was overexpressed in glioma. Second, our analysis results showed that the expression level of ITGB3BP was closely related to a series of important clinical features of glioma, and survival analysis showed that increased expression of ITGB3BP was not conducive to the survival of patients. Meta-analysis results also further confirmed that abnormally high expression of ITGB3BP can lead to a decline in the survival rate of patients with glioma. Hence, we believe that the overexpression of ITGB3BP plays a critical role in the poor prognosis of glioma patients. Consistently, through univariate, multivariate and ROC analyses, we concluded that ITGB3BP could be used as an independent predictive factor for the prognosis of glioma, with a specific diagnostic value. Finally, we used GSEA and TIMER database analysis to explore the possible mechanism of ITGB3BP in glioma and identified potential therapeutic compounds that may target ITGB3BP by CMap analysis. This is the first study to reveal the relationship between ITGB3BP and glioma, to the best of our knowledge. ITGB3BP may be an oncogene that affects the pathological process of glioma, but the specific mechanism still needs to be further explored.
Prior studies have noted that ITGB3BP affects the occurrence and development of tumours through different signal transduction pathways. For example, Yamanoi et al. found that the abnormally high expression of ITGB3BP could activate the hedgehog pathway in ovarian cancer, which led to the resistance of ovarian cancer cells to chemoradiotherapy. 32 In this study, we confirmed that ITGB3BP could lead to a poor prognosis of glioma, but the regulatory effect of ITGB3BP overexpression on the cell signalling pathway needs to be further elucidated. The results of GSEA suggested that the high ITGB3BP expression group exhibited increased activity of a variety  48 In addition, aspirin, a traditional anti-inflammatory and anti-platelet drug, assists in the prognosis of colorectal cancer. 49 Therefore, we believe that the above four small-molecule drugs have potential therapeutic effects for glioma patients via inhibition of ITGB3BP expression.
Immunosuppressive cells infiltrating the TME are key factors in the formation of the tumour-suppressive immune microenvironment. These immunosuppressive cells include CD4 + T cells, CD8 + T cells, B cells, neutrophils, dendritic cells and macrophages, which can assist immune evasion, thereby promoting the occurrence and development of tumours. 50,51 Previous studies have pointed out that tumour-infiltrating immune cells are key regulators of tumour growth and progression, which are related to the biological behaviour of glioma and patient survival. 52 In this study, we observed that ITGB3BP

| CON CLUS ION
Our study revealed that the overexpression of ITGB3BP was associated with the molecular and clinical characteristics of malignancy and predicted poor prognosis in glioma. Collectively, these findings provide new insight into the role in glioma of ITGB3BP, which might serve as a potential biomarker and novel therapeutic target for diagnosis and treatment.

ACK N OWLED G EM ENTS
We would like to thank all members of CGGA, TCGA, GEO, and GEPIA database for providing a good platform for researchers.