Comprehensive analyses reveal the role of histone deacetylase genes in prognosis and immune response in low-grade glioma

Abstract

Many studies have shown that Histone deacetylases (HDAC) is involved in the occurrence of malignant tumors and regulates the occurrence, proliferation, invasion, and migration of malignant tumors through a variety of signaling pathways. In the present, we explored the role of Histone deacetylases genes in prognosis and immune response in low-grade glioma. Using consensus clustering, we built the new molecular clusters. Using HDAC genes, we constructed and validated the prognostic model in two independent cohort datasets. Patients were divided into high-risk and low-risk groups. Then, we explored the molecular characteristics, clinical characteristics, tumor microenvironment and immune infiltration levels of two clusters and risk groups. Receiver operating characteristic analyses were built for model assessment. We finally detected the expression levels of signature genes between tumor and normal tissues. Low-grade can be separated into two molecular clusters using 11 HDACs genes. Two clusters had different clinical characteristics and prognosis. Nex, we constructed a prognosis model using six HDAC genes (HDAC1, HDAC4, HDAC5, HDAC7, HDAC9, and HDAC10), which was also validated in an independent cohort dataset. Furthermore, multivariate cox regression indicated that the calculated risk score was independently associated with prognosis in low-grade glioma, and risk score can predict the five-year survival probability of low-grade glioma well. High-risk patients can be attributed to multiple complex function and molecular signaling pathways, and the genes alterations of high- and low-risk patients were significantly different. We also found that different survival outcomes of high- and low- risk patients could be involved in the differences of immune filtration level and tumor microenvironment. Subsequently, using signature genes, we identified several small molecular compounds that could be useful for low-grade glioma patients' treatment. Finally, we detected the expression levels of signature genes in tumor tissues. our study uncovers the biology function role of HDAC genes in low-grade glioma. We identified new molecular subtypes and established a prognostic model based on six HDAC genes, which was well applied in two independent cohort data. The regulation of HDAC genes in low-grade glioma involved in multiple molecular function and signaling pathways and immune infiltration levels. Further experiments in vivo and vitro were required to confirm the present findings.

Fig 1. Molecular subtypes for low-grade glioma based on HDAC genes.

A: The flow chart of overall data processing. B: Circle plot showed the correlations among HDAC genes. C: Consensus matrix identified two clusters. D: Principal component analysis showed two components. E: tSEN further confirmed two molecular clusters. F: Kaplan-Meier survival curves of two molecular clusters. G: Correlations of molecular clusters with clinical characteristics and HDAC expressions.

Fig 2. Development and validations of a prognostic signature based on HDAC genes.

A: Forest plot of univariate cox regression for HDAC genes. B and C: LASSO regression identified the HDAC genes included in the model. D: Kaplan-Meier survival cures of high- and low-risk groups in TCGA. E: Risk score and survival time distributions of high- and low-risk groups in TCGA. F: PCA indicated two components in TCGA. G: Kaplan-Meier survival cures of high- and low-risk groups in CGGA. H: Risk score and survival time distributions of high- and low-risk groups in CGGA. I: PCA indicated two components in CGGA. J and K: Time-independent ROC for predicting survival at 1-year, 2-year, and 3-year using TCGA and CGGA datasets.

Fig 3. Stratified analyses for prognosis of high-and low-risk groups.

A and B: age< = 41 vs age>41. C and D: Male vs Female. E and F: WHO II vs III. G and H: primary vs recurrent. I and J: 1p19q non-codeletion and codeletion. K and L: IDH mutant vs wildtype. M and N: chemotherapy Yes vs No. O and P: Radiotherapy Yes vs No.

Fig 4. Correlations of risk score with clinical characteristics.

A: Heatmap showed the expression levels of HDAC genes between high- and low-risk groups. B: Heatmap showed the correlations of risk groups with clinical parameters and signature genes.

Fig 5. Risk assessment system for induvial prognosis.

A and B: univariate and multivariate cox regression of risk score for prognosis prediction in TCGA. C: Multi-ROC comparisons for predicting prognosis in TCGA. D and E: univariate and multivariate cox regression of risk score for prognosis prediction in CGGA. F: Multi-ROC comparisons for predicting prognosis in CGGA. G: Nomogram risk assessment system for individual prognosis. H, I and J: Calibrations fitting plots between actual and predicted probability at 1-year, 3-year, and 5-year.

Fig 6. Function and pathways analysis based on DGEs between high- and low-risk groups.

A: GO enrichment analysis. B: KEGG pathways analysis.

Fig 7. Gene alterations levels between high- and low-risk groups.

A and B: Top 20 gene alterations frequencies in high- and low-risk groups. C and D: Variant classifications of high- and low-risk groups. E and F: Co-occurrence and mutually exclusive genes of high- and low-risk groups.

Fig 8. Correlations of risk score with immune status.

A, B and C: Comparisons of ESTIMATE, stromal, and immune scores between two risk groups. D: Immune cell infiltrations between high- and low-risk groups. E: Immune function comparisons between high- and low-risk group. F-K: Scatter plot of associations between risk score and immune cells: Macrophages Mo, M1, Monocytes, Mast cells activated, T cells CD4 memory activated, and B cells naïve.

Fig 9. Chemotherapy sensitivity analysis for signature genes: HDAC1, HDAC4, HDAC5, HDAC7, HDAC9, HDAC10.

Fig 10. Expression levels of signature genes between tumor and normal tissue.

A: HDAC1. B: HDAC4. C: HDAC5. D: HDAC7. E: HDAC9. F: HDAC10.