FBP1 is a potential prognostic biomarker and correlated with tumor immunosuppressive microenvironment in glioblastoma

Hypoxia has been shown to contribute to tumor immunosuppressive microenvironment and is an effective prognostic indicator. This study aimed to screen prognostic hypoxia-related genes (HRGs) in glioblastoma and investigate the association between HRGs and tumor immunosuppressive microenvironment. The glioblastoma-related mRNA data were collected from TCGA, GEO, and CGGA databases. Totally 200 HRGs were obtained from the GSEA website. The prognostic HRGs were screened by univariate Cox regression analysis. Somatic mutation data of glioblastoma from TCGA was visualized using the "maftools" of R package. Immune cell infiltration proportions were calculated by CIBERSORT. The TISIDB online tool was applied to analyze the relationship between HRGs and immunoinhibitors as well as the HRG expression in different glioblastoma immune and molecular subtypes. Hub gene's mRNA and protein levels in cell lines were determined by qRT-PCR and western blot, respectively. The effects of hub gene knockdown on cell viability and migration ability were evaluated employing CCK8 and wound healing assays. The univariate Cox regression showed that high level of FBP1 (fructose-1,6-bisphosphatase 1) was a poor prognostic biomarker, and FBP1 was mainly expressed in lymphocyte depleted immune subtype of glioblastoma. High FBP1 mRNA and protein levels have been successfully validated in vitro. The somatic mutation analysis suggested that TP53 mutation rate was the highest in the high FBP1 glioblastoma group, while EGFR mutation rate was the highest in the low FBP1 glioblastoma group. In the high FBP1 group, the infiltration proportions and types of immune cells were less, dominated by macrophages M2, and the expression of CTLA4, LAG3, TIGIT, PDL1, and PDL2 was significantly upregulated. The expression of FBP1 was positively correlated with several immunoinhibitors, such as IL-10 and TGFβ-1. In conclusion, we demonstrated that FBP1 could serve as a prognostic biomarker for glioblastoma. The immune microenvironment in the high FBP1 group might be suppressed by up-regulating immune checkpoints and immunoinhibitors.