Identification of a unique tumor cell subset employing myeloid transcriptional circuits to create an immunomodulatory microenvironment in glioblastoma

Abstract

Glioblastoma (GBM) is an aggressive primary brain tumor with unique immunity predominated by myeloid cells. GBM cells have been implicated to evade immune attack through hijacking myeloid-affiliated transcriptional programs to establish an immunosuppressive microenvironment. However, molecular features of immune-evading GBM cells in heterogeneous GBMs and their interactions with immune cells remain unclear. Herein, we employed single-cell RNA sequencing (scRNA-seq) and bulk RNA-seq data to develop an in silico method for delineating GBM immune signature and identifying new molecular subsets for immunotherapy. We identified a new GBM cell subset, termed TC-6, that harbored immune-invading signature and actively interacted with tumor-associated macrophages (TAMs) to orchestrate an immune-suppressive niche. Proinflammatory transcriptional factors STAT1, STAT2, IRF1, IRF2, IRF3, and IRF7 were identified as the core regulons defining TC-6 subsets. Further immune transcriptome analyses revealed three immune subtypes (C1, C2, and C3). C3 subtype GBMs were enriched with TC-6 cells and immunosuppressive TAMs, and exhibited an immunomodulatory signature that associated with reduced efficacy of anti-PD-1 treatment. Interferon-related DNA damage resistance signaling was upregulated in C3 GBMs, predicting shortened survival of GBM patients who received chemo-radiation treatment. Treatment of OSI-930 as a molecular agent targeting c-kit and VEGFR2 tyrosine kinases may compromise the immunomodulatory signature of C3 GBMs and synergize with chemo-radiation therapy. We further developed a simplified 11-gene set for defining C3 GBMs. Our work identified TC-6 subset as an immune-evading hub that creates an immunomodulatory signature of C3 GBMs, gaining insights into the heterogeneity of GBM immune microenvironment and holding promise for optimized anti-GBM immunotherapy.

Keywords

• glioblastoma
• molecular subtyping
• single-cell rna sequencing
• prognosis