Characterization of Hypoxia Signature to Evaluate the Tumor Immune Microenvironment and Predict Prognosis in Glioma Groups

Abstract

Glioma groups, including lower-grade glioma (LGG) and glioblastoma multiforme (GBM), are the most common primary brain tumor. Malignant gliomas, especially glioblastomas, are associated with a dismal prognosis. Hypoxia is a driver of the malignant phenotype in glioma groups; it triggers a cascade of immunosuppressive processes and malignant cellular responses (tumor progression, anti-apoptosis, and resistance to chemoradiotherapy), which result in disease progression and poor prognosis. However, approaches to determine the extent of hypoxia in the tumor microenvironment are still unclear. Here, we downloaded 575 LGG patients and 354 GBM patients from Chinese Glioma Genome Atlas (GGGA), and 530 LGG patients and 167 GBM patients from The Cancer Genome Atlas (TCGA) with RNA sequence and clinicopathological data. We developed a hypoxia risk model to reflect the immune microenvironment in glioma and predict prognosis. High hypoxia risk score was associated with poor prognosis and indicated an immunosuppressive microenvironment. Hypoxia signature significantly correlated with clinical and molecular features and could serve as an independent prognostic factor for glioma patients. Moreover, Gene Set Enrichment Analysis showed that gene sets associated with the high-risk group were involved in carcinogenesis and immunosuppression signaling. In conclusion, we developed and validated a hypoxia risk model, which served as an independent prognostic indicator and reflected overall immune response intensity in the glioma microenvironment.

Keywords: gene set enrichment analysis; glioma; hypoxia; hypoxia risk model; immune response; tumor microenvironment.

Figure 1

Characterization of hypoxia risk signature to predict prognosis of glioma. (A) Protein–Protein Interaction interactions among 200 hypoxia-associated genes. The 20 genes with the highest interaction degrees were labeled; (B) Construction of a hypoxia risk signature to predict glioma prognosis by univariate and multivariate Cox regression; (C,D) Spearman correlation analysis of five hypoxia genes in the CGGA and TCGA databases.

Figure 2

Prognostic value of the hypoxia risk signature in glioma. (A) A heatmap showing five hypoxia gene expression profiles in high and low hypoxia risk groups from the CGGA and TCGA databases; (B) Patient status distribution in the high and low hypoxia risk groups. The dot presents patient status ranked by the increasing risk score. The X axis is patient number and Y axis is survival time; (C) Mortality rates of the high and low hypoxia risk groups; (D) Kaplan-Meier overall survival curves for patients assigned to high and low hypoxia risk groups based on the risk score; (E,F) The prognosis values of hypoxia signature in LGG in the CGGA and TCGA data; (G,H) The prognosis values of hypoxia signature in GBM in the CGGA and TCGA data.

Figure 3

Hypoxia gene expression is correlated with clinicopathological features of gliomas. (A,B) Heatmaps showing five hypoxia gene expression profiles in different WHO grades from the CGGA and TCGA databases; (C,D) The expression levels of five hypoxia genes in gliomas with different WHO grades; (E) The expression levels of five hypoxia genes in gliomas with different IDH status; (F) The expression levels of five hypoxia genes in gliomas with different 1p/19q codeletion status; *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

Figure 4

Prognostic value of the hypoxia risk signature in glioma. (A,B) ROC curves showing the predictive efficiency of the hypoxia risk signature on the 1-, 3-, and 5-years survival rate; (C–F) Univariate and multivariate Cox analyses evaluating the independent prognostic value of the hypoxia signature in terms of OS in glioma patients.

Figure 5

GSEA enrichment between low and high hypoxia risk groups. (A) GSEA revealing that genes in the high hypoxia risk group were enriched for hallmarks of malignant tumors in the CGGA data; (B) The results were further validated by the TCGA data. Normalized enrichment score (NES) > 1 and nominal p-value (NOM p-val) < 0.05 were considered significant gene sets.

Figure 6

Immune landscape between low and high hypoxia risk glioma patients. (A) Relative proportion of immune infiltration in high and low hypoxia risk patients; (B–G) Box plots visualizing significantly different immune cells between high-risk and low hypoxia risk patients; (H) GSEA analysis revealing immune-related biological processes correlated with hypoxia signature.

Figure 7

High hypoxia risk score indicates an immunosuppressive microenvironment. (A,B) Heatmap of the gene profiles involved in the negative regulation of the Cancer-Immunity Cycle in high and low hypoxia risk groups in the CGGA and TCGA databases; (C) Correlation between PD-L1 expression and hypoxia risk score; (D) PD-L1 expression in high and low hypoxia risk groups; (E) Correlation between PD1 expression and hypoxia risk score; (F) PD1 expression in high and low hypoxia risk groups; (G) CTLA-4 expression in high and low hypoxia risk groups; (H) LAG3 expression in high and low hypoxia risk groups; (I,J) Tumor immunosuppressive cytokine expression in high and low hypoxia risk groups. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.