Integrated analysis of inflammatory response subtype-related signature to predict clinical outcomes, immune status and drug targets in lower-grade glioma

Abstract

Background: The inflammatory response in the tumor immune microenvironment has implications for the progression and prognosis in glioma. However, few inflammatory response-related biomarkers for lower-grade glioma (LGG) prognosis and immune infiltration have been identified. We aimed to construct and identify the prognostic value of an inflammatory response-related signature, immune infiltration, and drug targets for LGG. Methods: The transcriptomic and clinical data of LGG samples and 200 inflammatory response genes were obtained from public databases. The LGG samples were separated into two inflammatory response-related subtypes based on differentially expressed inflammatory response genes between LGG and normal brain tissue. Next, inflammatory response-related genes (IRRGs) were determined through a difference analysis between the aforementioned two subtypes. An inflammatory response-related prognostic model was constructed using IRRGs by using univariate Cox regression and Lasso regression analyses and validated in an external database (CGGA database). ssGSEA and ESTIMATE algorithms were conducted to evaluate immune infiltration. Additionally, we performed integrated analyses to investigate the correlation between the prognostic signature and N 6-methyladenosine mRNA status, stemness index, and drug sensitivity. We finally selected MSR1 from the prognostic signature for further experimental validation. Results: A total of nine IRRGs were identified to construct the prognostic signature for LGG. LGG patients in the high-risk group presented significantly reduced overall survival than those in the low-risk group. An ROC analysis confirmed the predictive power of the prognostic model. Multivariate analyses identified the risk score as an independent predictor for the overall survival. ssGSEA revealed that the immune status was definitely disparate between two risk subgroups, and immune checkpoints such as PD-1, PD-L1, and CTLA4 were significantly expressed higher in the high-risk group. The risk score was strongly correlated with tumor stemness and m6A. The expression levels of the genes in the signature were significantly associated with the sensitivity of tumor cells to anti-tumor drugs. Finally, the knockdown of MSR1 suppressed LGG cell migration, invasion, epithelial-mesenchymal transition, and proliferation. Conclusion: The study constructed a novel signature composed of nine IRRGs to predict the prognosis, potential drug targets, and impact immune infiltration status in LGG, which hold promise for screening prognostic biomarkers and guiding immunotherapy for LGG.

Keywords: drug targets; immune characteristics; inflammatory response; lower-grade glioma; prognostic signature.

FIGURE 1

The overall procedure flow chart illustrating the data collection and analysis process.

FIGURE 2

Identification of the DE-IRGs and sub-clusters based on these genes. (A) A PPI network showing the interactions of the DE-IRGs (interaction score = 0.9). Circles in red denote upregulated genes, and green symbolizes downregulated genes in the LGG samples compared with normal brain tissues. (B) The correlation network of the DE-IRGs. The red line showed a positive correlation, and the blue line showed a negative correlation. The depth of the color reflects the strength of the relevance. (C) Patients with LGG were classified into two clusters according to the consensus clustering matrix (k = 2). (D) Kaplan–Meier curves for the OS of patients between the two clusters. DE-IRGs: differentially expressed inflammatory response genes; PPI: protein–protein interaction; LGG: lower-grade glioma; OS: overall survival.

FIGURE 3

Construction of the prognostic IRRG signature. (A) Univariate Cox regression analysis of OS based on IRRGs in the TCGA cohort. (B) LASSO regression of the 74 OS-related IRRGs. (C) Cross-validation for tuning the parameter in the LASSO regression. IRRG: inflammatory response subtypes-related gene; OS: overall survival; TCGA: The Cancer Genome Atlas; LASSO: the least absolute shrinkage and selection operator.

FIGURE 4

Validation of the prognostic IRRG signature. (A) KM curves for the OS of patients with LGG in the high- and low-risk groups in the TCGA cohort. (B) The ROC curve at 1-, 3-, 5-years for survival prediction of the signature in the TCGA cohort. Distribution of the risk score (C) and survival status (D) for each patient in the TCGA cohort. PCA plot (E) and t-SNE analysis (F) based on the nine prognostic signature genes in the TCGA cohort. (G) Similarly. KM curves for the OS of each patient in the CGGA cohort. (H) The ROC curve analysis of the risk score signature in the CGGA cohort. Distribution of the risk score (I) and survival status (J) for each patient with LGG in the CGGA cohort. PCA plot (K) and t-SNE analysis (L) based on the nine prognostic signature genes in the CGGA cohort. IRRG: inflammatory response-related gene; KM: Kaplan–Meier; OS: overall survival; LGG: lower-grade glioma; TCGA: The Cancer Genome Atlas; ROC: receiver operating characteristic; PCA: principal component analysis; t-SNE: T-distributed Stochastic Neighbor Embedding; CGGA: Chinese Glioma Genome Atlas.

FIGURE 5

Assessment of the risk scores and the predictive value of clinical variables. Forest charts of the risk scores combining common clinical variables based on the univariate Cox regression analysis in the TCGA cohort (A) and the CGGA cohort (B). Forest charts of the risk scores combining common clinical variables based on the multivariate Cox regression analysis in the TCGA cohort (C) and the CGGA cohort (D) showed the significance and HR values of risk scores and clinical characters. (E) Heatmap presented the association of risk and clinical information based on the nine-gene signature. **p < 0.01, ***p < 0.001. TCGA: The Cancer Genome Atlas; CGGA: Chinese Glioma Genome Atlas; HR: hazard ratio.

FIGURE 6

The nomogram for predicting 1-, 3-, and 5-year survival outcomes of LGG patients integrating prognostic markers including grade, gender, age, IDH1 status, 1p/19q codeletion, ATRX status, and MGMT promoter status in the TCGA cohort. LGG: lower-grade glioma; TCGA: The Cancer Genome Atlas.

FIGURE 7

Immune-related analysis in the TCGA and CGGA cohorts. Comparison of the ssGSEA scores of immune cells between low- and high-risk groups in the TCGA cohort (A) and the CGGA cohort (B). Comparison of the ssGSEA scores of immune-related pathways between low- and high-risk groups in the TCGA dataset (C) and the CGGA dataset (D). The relationship between the risk score and ESTIMATE score (E), immune score (F), stromal score (G), and tumor purity (H) in the TCGA cohort (I) Heatmap of the ssGSEA scores integrating the tumor purity, ESTIMATE score, immune score, and stromal score of each sample calculated by ESTIMATE's algorithm between the high- and low-risk groups in the TCGA dataset. (J) Comparison of the expression level of immune checkpoints among high- and low-risk groups in the TCGA dataset. *p < 0.05, **p < 0.01, ***p < 0.001. TCGA: The Cancer Genome Atlas; CGGA: Chinese Glioma Genome Atlas; ssGSEA: single-sample gene set enrichment analysis; ESTIMATE: Estimation of Stromal and Immune cells in Malignant Tumor tissues using Expression data.

FIGURE 8

Functional enrichment analyses based on the DEGs between the high- and low-risk groups in the TCGA cohort. (A) The bubble graph for GO enrichment presented the top 10 terms in BP, CC, and MF. Circle size corresponded to enriched counts of genes, and the circle indicated the q-values and the significance of the enriched GO terms. (B) The bar plot for the KEGG pathway analysis. The bar length corresponded to the number of enriched genes in the corresponding pathway. The gradual color indicated the different degrees of KEGG enrichment, with red representing the highest magnitude of KEGG enrichment. DEG: differentially expressed gene; TCGA: The Cancer Genome Atlas; GO: Gene Ontology; BP: Biological Process; CC: Cellular Component; MF: Molecular Function; KEGG: Kyoto Encyclopedia of Genes and Genomes.

FIGURE 9

The correlation between the prognostic gene expression and drug sensitivity. The top 16 drugs with the highest correlation with gene expression in the predictive model were screened. The vertical axis shows the Z-scores of the drugs, and the horizontal axis represents the gene expression. The larger the Z-score, the more sensitive the cancer cell is to the drug.

FIGURE 10

Effect of ABCC3 knockdown on the sensitivity of glioma cells to arsenic trioxide. The cell viability of U251 (A) and T98G cells (B) in the Si-ABCC3 group was significantly inhibited compared with the control group following treatment with arsenic trioxide (4 μM) for 24, 48, and 72 h. EdU staining showed that the proportion of EdU-positive cells of U251 (C) and T98G cells (D) in the Si-ABCC3 group markedly reduced compared to the control group after being treated with arsenic trioxide (4 μM) for 48 h. Photographs (C) and (D) magnification: ×200; scale bar: 50 μm. The data are presented as the mean ± SD for at least three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001. EdU: 5-Ethynyl-2′-deoxyuridine.

FIGURE 11

Correlation between the risk score and stemness index and m6A. The relationship between the risk score and stemness index based on mDNAsi (A) and the stemness index based on mRNAsi (B) in the TCGA dataset. (C) The expression level of m6A-related genes between the high- and low-risk groups in the TCGA dataset. *p < 0.05, **p < 0.01, ***p < 0.001. ns: non-sense; m6A: N6-methyladenosine; mDNAsi: DNA methylation-based stemness index; mRNAsi: mRNA expression-based stemness index; TCGA: The Cancer Genome Atlas.

FIGURE 12

MSR1 knockdown impairs migration, invasion, EMT, and proliferation. (A) Relative RNA expression of MSR1 in LGG samples and normal brain tissues. (B) The transfection efficiency of the MSR1 siRNA in SHG44 and HS683 cells was assessed via qRT-PCR. (C) Transwell migration assay and (D) transwell invasion assay presented that MSR1 downregulation remarkably reduced the migration, invasion, EMT, and proliferation of SHG44 and HS683 LGG cell lines. (E) Western blot analysis demonstrated the expression changes in EMT markers (ZO-1 and vimentin) in the indicated LGG cell lines after MSR1 knockdown with siRNA. (F) CCK8 assay evaluated the proliferation ability between control and Si-MSR LGG cells. *p < 0.05, **p < 0.01, ***p < 0.001. MSR1: macrophage scavenger receptor 1; EMT: epithelial-mesenchymal transition; LGG: lower-grade glioma; siRNA: small interfering RNA.