Immunological Analysis of Brain Cancer

Division Of Basic Sciences - Nci

Investigators

Linked publications & trials

Abstract

Aim 1: We have been developing viral vector-based T cell-inducing immunotherapy using ChAdOx-1 and MVA vectors and assessing their efficacy against mouse syngeneic glioblastoma in vivo. Last year, using an SB28 syngeneic orthotopic glioblastoma model expressing the model tumor antigen P1A, we demonstrated that prophylactic treatment with a combination of two viral vectors expressing the model antigen resulted in significant protection. This year, we further examined the efficacy of the treatment in a therapeutic context, initiating therapy one week after tumor implantation when the tumors were established in the brain. We found that the treatment significantly prolonged the survival of mice with glioblastoma. High-dimensional flow cytometry analysis of the immune cell composition in tumors, both with and without treatment, revealed a significant increase in a subset of P1A-specific CD8+ T cells from the treated group. We discovered that the adoptive transfer of the CD8 T cell subset from tumors in mice that received immunotherapy protected the mice from tumor development, demonstrating that the P1A-specific CD8 T cell subset was sufficient to reduce tumor growth. A potential caveat of the study is that we have used tumor cells expressing the model tumor antigen at a supraphysiological level, which could more effectively induce T cell responses. To address this concern, we identified tumor-associated antigens (TAAs) in a parental mouse glioblastoma cell line, SB28, using the bioinformatics pipeline we developed (named METRO), and created viral vectors expressing those TAAs. Among five TAAs, we found that immunotherapy with viral vectors expressing TAA#1 induced protection in both prophylactic and therapeutic settings. Furthermore, we observed that the same subset of CD8 T cells, which we identified in the SB28-P1A model among TAA#1-specific CD8 T cells, significantly increased in mice that received the immunotherapy. Currently, we are examining a detailed mechanism for T cell induction in the brain and approaches, including NKT cell-targeted strategies, to enhance the efficacy of the therapy. Aim 2: We have employed multiple approaches to test our hypothesis that CD1d-NKT cells contribute to immune regulation in glioblastoma. In the first subproject, we have been testing how glioblastoma's altered metabolism produces a unique set of lipids recognized by NKT cells and how these glioblastoma-enriched lipids modulate the functions of NKT cells. Previously, we identified that human IDH-wild type glioblastoma cells have distinct lipid profiles compared to human IDH-mutated low-grade glioma cell lines and normal human astrocytes. A class of glioblastoma-enriched lipids identified was sulfatide. The glioblastoma-enriched lipids were recognized by and activated human NKT cells. Interestingly, glioblastoma-enriched sulfatide induced the expression of CD69 on human iNKT cells, but none of them induced the production of cytokines. Furthermore, one glioblastoma-enriched lipid, lyso-sulfatide, suppressed activation and cytokine production induced by the agonistic antigen alpha-galactosylceramide (aGC). The suppressive activity of lyso-sulfatide was not due to the induction of cell death in iNKT cells. It was not the result of competition against other lipids for binding to the antigen-presenting molecule CD1d, because preincubation of antigen-presenting cells (B cells) with either aGC or lyso-sulfatide separately before mixing with iNKT cells still demonstrated the suppression of iNKT cell activation. We also observed similar reactivity of glioblastoma patient-derived ex vivo iNKT cells against the lipids. These results suggest that glioblastoma produces lipids that modulate NKT cell function to suppress tumor immunity. In the second subproject, we aim to develop a new NKT cell agonist that can enhance the effects of the viral vector-based immunotherapy mentioned in Aim 1. NKT cells have been shown to modulate the immunosuppressive functions of myeloid cells into immunoenhancing functions. We have been studying the antigenicity of newly synthesized analogs of sulfaide, designed based on the structure of 7DW8-5, a previously reported human NKT cell agonistic lipid antigen, using human iNKT cells. One of the analogs, SAP-3, induced significant activation, measured by the expression of CD69 and CD25, in healthy donor-derived ex vivo iNKT cells as well as glioblastoma patient-derived ex vivo iNKT cells. The activation was completely abrogated by an anti-CD1d blocking antibody, indicating that SAP-3/CD1d complexes were recognized by human iNKT cell T cell receptors. SAP-3 also induced a significant increase in IL-2 production. Since IL-2 is a T cell growth factor, we hypothesized that SAP-3 enhances the CD8 T cell response induced by a protein antigen. We tested this hypothesis by stimulating healthy donor peripheral blood mononuclear cells (PBMCs) with an influenza matrix protein-derived peptide (FMP) in the presence or absence of SAP-3. Most PBMC donors possess memory CD8+ T cells specific for FMP, as they have been exposed to influenza in their lives. We found that adding SAP-3 to the stimulation culture did not change the number of FMP-specific CD8+ T cells. However, the presence of SAP-3 improved the quality of FMP-specific CD8+ T cells, as the proportion of polyfunctional CD8+ T cells producing IFN-g, TNF-a, and IL-2 simultaneously was significantly increased. Aim 3: The roles and presence of MAIT cells in glioblastoma remain poorly understood. Last year, we discovered that 15 out of 22 human glioblastoma tissue samples contained MAIT cells, using scRNAseq combined with TCR VDJ sequencing. Most MAIT cells in glioblastoma tissues expressed RORC, indicating they are IL-17 producing cells. Furthermore, we found that the expression levels of MAIT signature genes negatively correlated with disease prognosis, suggesting that MAIT cells contribute to the immune suppression observed in glioblastoma patients. This year, we conducted a spatial analysis of MAIT cells and granulocytes in glioblastoma tissues using high-dimensional tissue imaging (CODEX) across nine samples from glioblastoma patients. Notably, MAIT cell signature-positive glioblastoma samples exhibited higher expression levels of signature genes associated with immunosuppressive myeloid cells. We observed considerable heterogeneity among the samples, which aligns with the traits of glioblastoma. We identified MAIT cells infiltrating the tumor tissues and also noted clusters of granulocytes in some tissues. In one sample, we found a correlation between the counts of MAIT cells and neutrophils.