GADD45G as a novel prognostic biomarker and therapeutic target in glioma: Integrative analysis of bulk and single-cell RNA sequencing

Cong Shen¹Haiping Cai²Chengliang Mao²Jiahao Yang²Kai Tang^2*

• ¹Shantou University Medical College, Shantou, China, Shantou, Guangdong Province, China
• ²Department of Neurosurgery, Guangdong Provincial People's Hospital (Guangdong Academy of Medical Sciences), Southern Medical University, Guangzhou, China

Background: Gliomas make up almost half of primary central nervous system tumors. Despite advancements in surgery and neuro-oncology, developing an effective treatment remains challenging. The protein Growth Arrest and DNA-Damage-Inducible, Gamma (GADD45G) is crucial for key cellular functions like DNA repair, genomic stability, and apoptosis. While GADD45G dysregulation has been found in various cancers, its role in glioma is still unclear. Methods: We analyzed unified pan-cancer datasets (TCGA, TARGET, GTEx) from UCSC Xena and integrated glioma data from CGGA and GEO (GSE108476). Prognostic value was assessed via multivariate Cox regression and Kaplan-Meier survival analysis using Gliovis. Single-cell RNA-seq data (GSE103224, GSE138794, GSE173278) were processed with Seurat (R 4.2.2), with Harmony for batch correction and UMAP for visualization. Malignant subclusters were annotated using marker genes. Functional enrichment and cell-type proportion estimation were conducted. Single-cell analysis revealed GADD45G expression patterns and identified its top correlated genes in malignant glioblastoma cells. Overexpression of GADD45G was performed to investigate its impact on cell function. Western blot analysis was used to examine the role of GADD45G in glioma cell invasion and migration. Results: Through comprehensive analysis across multiple datasets, it was found that GADD45G expression is higher in glioma patients compared to normal individuals, and its expression is generally higher in lower-grade gliomas than in glioblastoma. Cox regression analysis indicated that GADD45G has a protective effect. Survival curves further demonstrated that elevated GADD45G levels are associated with improved overall survival in patients. In this study, we identified four highly heterogeneous GBM cell subpopulations using single-cell data. The MES-like cells was significantly associated with poor prognosis. Spearman correlation analysis revealed the correlation between GADD45G and VIM. Further experiments revealed that GADD45G modulates glioma cell invasion and migration, potentially through its effects on EMT-like phenotypic features. Conclusion: GADD45G expression is significantly associated with glioma outcomes and may serve as a promising biomarker for prognosis evaluation. Its involvement in regulating EMT-like phenotypic traits further highlights its therapeutic potential.