Abstract Glioblastoma (GBM) and brain metastases remain largely resistant to immunotherapeutic intervention. T cell exhaustion, a differentiation state characterized by loss of function and persistence, contributes to this resistance. Two exhaustion subsets, progenitor (Tex_prog) and terminal (Tex_term), have been identified, where only Tex_prog remain responsive to immunotherapy. To date, the dynamics and characteristics of these exhausted populations in GBM and brain metastases remain unclear. Herein, we identify a striking loss of Tex_prog in a murine model of GBM throughout tumor progression. We elucidate the requirements for Tex_prog to Tex_term transition through characterization of T cell exhaustion progression at the RNA (paired scRNA and TCR sequencing) and protein level (flow-cytometry). Tex_prog are enriched for pathways of migration, cell-cell adhesion, and differentiation, whereas Tex_term primarily upregulate cytotoxic pathways. TCRseq revealed that clonal expansion is concentrated to the Tex_term cluster, suggesting a link between Tex_prog transition and proliferation in response to antigen. Likewise, we defined the necessity of hematopoietic antigen presentation, but not tumor-derived antigen presentation for the Tex_prog to Tex_term transition. Of the antigen presenting cells in the TME, tumor-associated macrophages (TAM) had the greatest expression of tumor-antigen loaded MHC I, illustrating their capacity to cross-present. By depleting TAM, we observed disruption of the Tex_prog transition. Similar results were seen in subcutaneous and intracranial melanoma, highlighting the role of TAM in this process across tumor histology and location. Lastly, we identified additional inflammatory cell-cell interactions between T cells and TAM in mice and humans that may contribute to this phenomenon. Taken together, we define T cell exhaustion dynamics in models of primary and metastatic brain tumors, where tumor-antigen presentation by TAM is a critical mediator of the loss of Tex_prog. We extend these findings to human data to better define these complex interactions, which we aim to leverage to overcome immunotherapy resistance.
Read full abstract