Integrated Analysis of Bulk and Single Cell Transcriptomic Sequencing Reveals Immune Patterns of T Cell Regulation and Prognostic Biomarkers in Glioblastoma

Technical Summary: Integrated Analysis of Bulk and Single-Cell Transcriptomic Sequencing in Glioblastoma

Problem Statement
Glioblastoma (GBM) remains the most aggressive primary malignant tumor of the adult central nervous system, characterized by a poor prognosis and a 5-year survival rate of less than 10% for newly diagnosed patients. Despite standard interventions involving surgical resection, radiotherapy, and temozolomide chemotherapy, the immunosuppressive tumor microenvironment (TME) severely limits the efficacy of immunotherapies such as immune checkpoint inhibitors (ICIs) and CAR-T cells. A critical gap exists in understanding the heterogeneity of T cells within the GBM microenvironment, their interaction patterns with other cell populations, and the specific molecular mechanisms driving their dysfunction. Consequently, there is an urgent need to identify robust prognostic biomarkers and develop models that can guide personalized clinical interventions.

Methodology
This study employed a multi-omics approach integrating bulk RNA sequencing (bulk RNA-seq) and single-cell RNA sequencing (scRNA-seq) data from public repositories (TCGA, GEO). The analytical workflow proceeded as follows:

1. Core Gene Identification via WGCNA: Weighted Gene Co-expression Network Analysis (WGCNA) was performed on bulk transcriptomic datasets (GSE4290 and GSE50161) to construct co-expression networks. Modules highly correlated with GBM status were identified, and their intersection with differential expression analysis yielded a core gene set of 47 GBM-associated genes.
2. Single-Cell Resolution and Annotation: scRNA-seq data (GSE162631) comprising four normal and four GBM brain tissues were processed using the Seurat package. After rigorous quality control and dimensionality reduction (PCA, UMAP, t-SNE), cells were clustered and annotated into six major subtypes: endothelial cells, fibroblasts, macrophages, malignant cells, microglia, and T cells.
3. T-Cell Specific Analysis: The 47 core genes were mapped onto single-cell data using the UCell algorithm to determine enrichment. T cells were identified as the primary carriers of these core genes. Subsequent analyses included:
   • Differential Expression: Identification of 100 T-cell-specific differentially expressed genes (DEGs).
   • Trajectory and Communication: Pseudotime trajectory analysis (Monocle 2) to map T-cell state transitions and CellChat analysis to quantify ligand-receptor interactions between T cells and other stromal components.
   • Functional Enrichment: GO and KEGG pathway analyses to characterize biological functions.
4. Subtyping and Prognostic Modeling:
   • Consensus Clustering: TCGA-GBM samples were stratified into two molecular subtypes (C1 and C2) based on the expression of T-cell DEGs.
   • Immune and Drug Profiling: The subtypes were evaluated using ESTIMATE, multiple immune infiltration algorithms (CIBERSORT, XCELL, etc.), and drug sensitivity prediction (oncoPredict).
   • Risk Model Construction: A LASSO-Cox regression model was built using the T-cell DEGs to derive a 5-gene prognostic signature. The model's performance was validated via Kaplan-Meier survival analysis, ROC curves, and independent protein-level verification using the Human Protein Atlas (HPA) database.

Key Contributions and Results

• Identification of T-Cell Centric Core Genes: The study successfully identified a 47-gene core set associated with GBM. Single-cell analysis revealed that T cells exhibited the highest enrichment scores for this gene set among all cellular components, highlighting their central role in the GBM immune landscape.
• Mechanistic Insights into T-Cell Exhaustion: Pseudotime analysis demonstrated that T cells in GBM evolve along a trajectory toward terminal exhaustion, characterized by decreased cytotoxicity and the abnormal upregulation of immunoglobulin genes (e.g., IGHA2, IGHG2). Cell-cell communication analysis revealed that fibroblasts are the primary source of incoming signals to T cells, with the COL family (collagens) acting as a potent regulator of T-cell CD44 and integrin receptors, suggesting a mechanism of matrix-mediated immune suppression.
• Molecular Subtyping (C1 vs. C2): The study defined two distinct GBM subtypes based on T-cell gene expression:
  • C1 Subtype: Characterized by higher expression of immune checkpoint genes and sensitivity to PI3K inhibitors (e.g., Alpelisib, Dactolisib) and alkylating agents (Carmustine).
  • C2 Subtype: Exhibited higher immune infiltration (Macrophages, Tregs), higher immune scores, and greater sensitivity to epigenetic modulators (HDAC inhibitors like Vorinostat) and CDK4/6 inhibitors (Ribociclib).
• Prognostic 5-Gene Signature: A robust prognostic model was constructed using five key genes: GNG11, CD5, COL1A1, CAMK4, and BCAS4.
  • Risk Factors: High expression of GNG11 and CD5 correlated with poor prognosis.
  • Protective Factors: High expression of COL1A1, CAMK4, and BCAS4 correlated with better survival outcomes.
  • Validation: The model demonstrated predictive accuracy with AUC values of 0.59 (1-year), 0.64 (3-year), and 0.69 (5-year). Immunohistochemical analysis from the HPA database confirmed that the protein expression levels of all five genes were significantly upregulated in GBM tissues compared to normal brain tissue, validating the transcriptomic findings at the proteomic level.

Significance and Claims
The authors claim that this study systematically elucidates the heterogeneity of the T-cell-mediated immune microenvironment in GBM. By integrating bulk and single-cell data, the research provides a theoretical basis for understanding the complex interplay between tumor stroma (specifically fibroblasts and ECM) and T-cell exhaustion. The development and validation of the 5-gene prognostic model offer a practical tool for clinical prognosis evaluation. Furthermore, the identification of subtype-specific drug sensitivities suggests a pathway toward personalized clinical interventions, potentially improving the diagnosis, prognosis assessment, and targeted immunotherapy strategies for GBM patients. The study acknowledges limitations regarding its reliance on retrospective public data and the need for future in vivo/in vitro validation but asserts that its findings provide a foundational framework for exploring GBM biomarkers and therapeutic targets.