Integration of Single-Cell and Bulk RNA-seq Data to Identify the Cancer-Associated Fibroblast Subtypes and Risk Model in Glioma.

Cancer-associated fibroblasts (CAFs) are an important component of the stroma. Studies showed that CAFs were pivotally in glioma progression which have long been considered a promising therapeutic target. Therefore, the identification of prognostic CAF markers might facilitate the development of novel diagnostic and therapeutic approaches. A total of 1333 glioma samples were obtained from the TCGA and CGGA datasets. The EPIC, MCP-counter, and xCell algorithms were used to evaluate the relative proportion of CAFs in glioma. CAF markers were identified by the single-cell RNA-seq datasets (GSE141383) from the Tumor Immune Single-Cell Hub database. Unsupervised consensus clustering was used to divide the glioma patients into different distinct subgroups. The least absolute shrinkage and selection operator regression model was utilized to establish a CAF-related signature (CRS). Finally, the prognostic CAF markers were further validated in clinical specimens by RT‒qPCR. Combined single-cell RNA-seq analysis and differential expression analysis of samples with high and low proportions of CAFs revealed 23 prognostic CAF markers. By using unsupervised consensus clustering, glioma patients were divided into two distinct subtypes. Subsequently, based on 18 differentially expressed prognostic CAF markers between the two CAF subtypes, we developed and validated a new CRS model (including PCOLCE, TIMP1, and CLIC1). The nomogram and calibration curves indicated that the CRS was an accurate prognostic marker for glioma. In addition, patients in the high-CRS score group had higher immune infiltration and tumor mutation burden levels. Moreover, the CRS score had the potential to predict the response to immune checkpoint blockade (ICB) therapy and chemotherapy. Finally, the expression profiles of three CAF markers were verified by RT‒qPCR. In general, our study classified glioma patients into distinct subgroups based on CAF markers, which will facilitate the development of individualized therapy. We also provided insights into the role of the CRS in predicting the response to ICB and chemotherapy in glioma patients.