Association of IL۱۸ and CSF۱R overexpression leads to immunosuppressive phenotypes in microglial cells in GBM microenvironment: A transcriptome-wide meta-analysis

GBM is the most aggressive brain tumor in which glial cells undergo several changes leading to tumorprogression. The role of astrocytes in GBM progression is well established, but its role in connection with otherglial cells and induction of invasion still remains unclear. Hereby, we used single RNA sequencing data analysisin order to perform an integrative analysis of glioblastoma tissue samples. Cell type identification was performedusing SingleR package which executes cell type identification based on machine learning algorithms. Amongabout ۱۰۰۰۰ cells data analyzed, we obtained three main clusters of cells including astrocyte, microglia, andoligodendrocytes. Interestingly, each cell type included several clusters. We hypothesized the presence of variouscell phenotypes for each cell type among the identified clusters. Accordingly, clustering was performed to identifythe more similar clusters for identified cell types. This process led to find two phenotypes for astrocyte includingactive astrocyte and astrocyte/stem cell, three phenotypes for microglia including CD۱۶۳/CD۱۴ microglia, CD۳۳+or steady state microglia, and active microglia, immature oligodendrocyte, and neurons. Studying the genenetworks and related signaling pathways using Cytoscape and GSEA analysis led to find that astrocytes showproliferative and growth factor producing activities, while active astrocyte phenotype shows mitochondrialdysfunction and apoptotic changes. Microglial subpopulations show a wide range of diversity in expression levelsof marker genes including interleukins. Accordingly, it seems that each phenotype of microglia show increase indifferential expression levels of interleukins and growth factors, so that active microglia show higher expressionlevels of VEGF, while CD۱۶۳/CD۱۴ microglia shows higher expression levels of TGF-β and HBEGF, and CD۳۳+microglia show higher expression levels of PDGFB in comparison to other types. Similar results found forastrocytes, so that stem cell/astrocytes show higher expression levels of EGFR and PDGFR than active astrocytes,while active astrocytes show higher expression levels of VEGF. In addition, our findings revealed that astrocytesoverexpression of CSF۱ as a tumorigenic factor. However, overexpression of CSF۱R in microenvironment werefound in microglia subpopulations. In addition, studying the common genes between three phenotypes ofmicroglia showed that there are ۱۲ common genes between CD۱۶۳/CD۱۴, and active microglia microglia, ۷۵common genes between CD۱۶۳/CD۱۴ and CD۳۳+, and ۱۱۲ common genes between CD۳۳+ and active microglia.GSEA analysis showed that PI۳K, ECM-receptor interaction, and Focal adhesion signaling pathways are amongthe most frequent singling pathways in microglia and astrocytes withing the microenvironment glioblastoma.According to the results of pathway enrichment analysis, immunosuppressive microglia show overexpression ofcytokines and inflammatory responses. Our results for the first time showed association of overexpressionbetween CSF۱R and IL۱۸ as tumorigenic and inflammatory cytokine leading to transition of microglia toimmunosuppressive phenotype. Considering that the simultaneous overexpression of IL۱۸ and CSF was found inother cancers as prognostic factor. On the other hand, it was found that using IL-۱۸and CSF as adjuvants for tumor vaccines can induce immune response in colon and spleen cancer. In line withthese findings, we found overexpression of CSF۱R in various clusters of microglial cells which was associatedwith overexpression of IL۱۸ in all those clusters. Considering that no studies has reported the association betweenCSF۱R and IL۱۸ in GBM or other types of brain tumors, it is the first time that this association is studied.

کلیدواژه ها:

glioblastoma multiforme ، reactive astrocyte ، GBM microenvironment ، IL۱۸ ، CSF۱R ، microglia