Molecular typing of gliomas on the basis of integrin family genes and a functional study of ITGA7.

Gliomas are highly malignant tumors of the central nervous system, and their complex molecular heterogeneity poses major therapeutic challenges. Integrins are important members of the class of cell adhesion molecules (CAMs), consisting of α-subunits and β-subunits that form 24 different heterodimers. To elucidate the complex role of integrins in glioma pathogenesis, we analyzed integrin family genes. We used a scoring system based on gene set enrichment analysis (GSEA) to identify prognostic biomarkers and nonnegative matrix factorization (NMF) to establish a new integrin-based molecular classification of gliomas. Subsequent analyses of the clinical relevance of the molecular subtypes and the underlying mechanisms demonstrated a strong correlation between integrin-based molecular subtypes and glioma malignancy. We further characterized the different clinical features and tumor microenvironments (TMEs) associated with these subtypes. We identified subtype-specific driver genes using the limma R package and weighted gene coexpression network analysis (WGCNA). We subsequently identified key integrin-mediated genes that significantly contribute to poor prognosis through a combined approach of machine learning (ML) and protein‒protein interaction (PPI) network analysis. Finally, we performed in vitro cellular experiments on the integrin family gene ITGA7 and demonstrated that ITGA7 can serve as a biomarker for gliomas. Our findings provide important insights into the multifaceted roles of integrins in glioma biology, provide an opportunity for the discovery of novel targeted therapies on the basis of the subtype-specific vulnerability of integrins, and provide a basis for the study of the role of ITGA7 in gliomas.