Summary Responses to anti-PD-1 immunotherapy occur but are infrequent in bladder cancer. The specific T cells that mediate tumor rejection are unknown. T cells from human bladder tumors and non-malignant tissue were assessed with single-cell RNA and paired T cell receptor (TCR) sequencing of 30,604 T cells from 7 patients. We find that the states and repertoires of CD8 + T cells are not distinct in tumors compared with non-malignant tissues. In contrast, single-cell analysis of CD4 + T cells demonstrates several tumor-specific states, including multiple distinct states of regulatory T cells. Surprisingly, we also find multiple cytotoxic CD4 + T cell states that are clonally expanded. These CD4 + T cells can kill autologous tumors in an MHC class II-dependent fashion and are suppressed by regulatory T cells. Further, a gene signature of cytotoxic CD4 + T cells in tumors predicts a clinical response in 244 metastatic bladder cancer patients treated with anti-PD-L1.

Graphical Abstract Highlights d High-resolution map of human NK cells shows tissue-driven distribution across ages d Differentiated NK cells predominate in blood, bone marrow, spleen, and lungs d Tissue-resident NK cells exhibit specific adaptations in mucosal and lymphoid sites d Lymph nodes and intestines are reservoirs for precursor and immature NK cells

The common pathogen Epstein-Barr virus (EBV) transforms normal human B cells and can cause cancer. Latent membrane protein 2A (LMP2A) of EBV supports activation and proliferation of infected B cells and is expressed in many types of EBV-associated cancer. It is not clear how latent EBV infection and cancer escape elimination by host immunity, and it is unknown whether LMP2A can influence the interaction of EBV-infected cells with the immune system. We infected primary B cells with EBV deleted for LMP2A, and established lymphoblastoid cell lines (LCLs). We found that CD8+ T cell clones showed higher reactivity against LMP2A-deficient LCLs compared to LCLs infected with complete EBV. We identified several potential mediators of this immunomodulatory effect. In the absence of LMP2A, expression of some EBV latent antigens was elevated, and cell surface expression of MHC class I was marginally increased. LMP2A-deficient LCLs produced lower amounts of IL-10, although this did not directly affect CD8+ T cell recognition. Deletion of LMP2A led to several changes in the cell surface immunophenotype of LCLs. Specifically, the agonistic NKG2D ligands MICA and ULBP4 were increased. Blocking experiments showed that NKG2D activation contributed to LCL recognition by CD8+ T cell clones. Our results demonstrate that LMP2A reduces the reactivity of CD8+ T cells against EBV-infected cells, and we identify several relevant mechanisms.

BackgroundGlioblastoma (GBM) is refractory to immune checkpoint inhibitor (ICI) therapy. We sought to determine to what extent this immune evasion is due to intrinsic properties of the tumor cells versus the specialized immune context of the brain, and if it can be reversed.MethodsWe used CyTOF mass cytometry to compare the tumor immune microenvironments (TIME) of human tumors that are generally ICI-refractory (GBM and sarcoma) or ICI-responsive (renal cell carcinoma), as well as mouse models of GBM that are ICI-responsive (GL261) or ICI-refractory (SB28). We further compared SB28 tumors grown intracerebrally versus subcutaneously to determine how tumor site affects TIME and responsiveness to dual CTLA-4/PD-1 blockade. Informed by these data, we explored rational immunotherapeutic combinations.ResultsICI-sensitivity in human and mouse tumors was associated with increased T cells and dendritic cells (DCs), and fewer myeloid cells, in particular PD-L1+ tumor-associated macrophages. The SB28 mouse model of GBM responded to ICI when grown subcutaneously but not intracerebrally, providing a system to explore mechanisms underlying ICI resistance in GBM. The response to ICI in the subcutaneous SB28 model required CD4 T cells and NK cells, but not CD8 T cells. Recombinant FLT3L expanded DCs, improved antigen-specific T cell priming, and prolonged survival of mice with intracerebral SB28 tumors, but at the cost of increased Tregs. Targeting PD-L1 also prolonged survival, especially when combined with stereotactic radiation.ConclusionsOur data suggest that a major obstacle for effective immunotherapy of GBM is poor antigen presentation in the brain, rather than intrinsic immunosuppressive properties of GBM tumor cells. Deep immune profiling identified DCs and PD-L1+ tumor-associated macrophages as promising targetable cell populations, which was confirmed using therapeutic interventions in vivo.