Antigen-presenting B cells promote TCF-1+ PD1- stem-like CD8+ T-cell proliferation in glioblastoma

Abstract

Understanding the spatial relationship and functional interaction of immune cells in glioblastoma (GBM) is critical for developing new therapeutics that overcome the highly immunosuppressive tumor microenvironment. Our study showed that B and T cells form clusters within the GBM microenvironment within a 15-μm radius, suggesting that B and T cells could form immune synapses within the GBM. However, GBM-infiltrating B cells suppress the activation of CD8⁺ T cells. To overcome this immunosuppression, we leveraged B-cell functions by activating them with CD40 agonism, IFNγ, and BAFF to generate a potent antigen-presenting B cells named B_Vax. B_Vax had improved antigen cross-presentation potential compared to naïve B cells and were primed to use the IL15-IL15Ra mechanism to enhance T cell activation. Compared to naïve B cells, B_Vax could improve CD8 T cell activation and proliferation. Compared to dendritic cells (DCs), which are the current gold standard professional antigen-presenting cell, B_Vax promoted highly proliferative T cells in-vitro that had a stem-like memory T cell phenotype characterized by CD62L⁺CD44^- expression, high TCF-1 expression, and low PD-1 and granzyme B expression. Adoptive transfer of B_Vax-activated CD8⁺ T cells into tumor-bearing brains led to T cell reactivation with higher TCF-1 expression and elevated granzyme B production compared to DC-activated CD8⁺ T cells. Adoptive transfer of B_Vax into an irradiated immunocompetent tumor-bearing host promoted more CD8⁺ T cell proliferation than adoptive transfer of DCs. Moreover, highly proliferative CD8⁺ T cells in the B_Vax group had less PD-1 expression than those highly proliferative CD8⁺ T cells in the DC group. The findings of this study suggest that B_Vax and DC could generate distinctive CD8⁺ T cells, which potentially serve multiple purposes in cellular vaccine development.

Keywords: B cells; GBM; anti-tumor response; immunological synapse; stem-like memory CD8+ T cell.

Figure 1

Spatial analysis of B T and myeloid cells in GBM TME. (A) Representative Spatial Multiplex seqIF™ images generated using the COMET™ platform (Lunaphore Technologies) showing cellular infiltration in the GBM TME from patient NU02569 and NU02594. The TME was analyzed for prevalence of B cells (CD20⁺), cytotoxic T cells (CD8⁺), myeloid cells (CD163⁺), epithelial cells (CD31⁺), tumor cells and astrocytes (GFAP⁺) combined with DAPI staining. For each patient, upper images from left to right: low magnification of full panel, high magnification of the full panel, high magnification background stain (DAPI, GFAP and CD31). Lower images from left to right: High magnification of CD8⁺, CD20⁺, CD163⁺ cells. White arrows highlight CD8⁺ or CD20⁺ cells. Yellow arrow and box mark the magnified region. Dashed arrows and boxes mark the regions of the high magnification images. (B) Paired correlation analysis (n=21) of density of CD8⁺ T cells, myeloid cells (CD163⁺ TMEM119^-), and tumor cells (SOX2⁺) based on radial distance from the nearest B-cell from images obtained with Vectra 3 Automated Quantitative Pathology Imaging System.

Figure 2

B_vax demonstrates superior antigen uptake and presentation function. (A) Illustration of B_Vax and B_Naive generation. (B) KEGG transcriptomic pathway analysis from bulk RNA sequencing comparing upregulated gene pathways in Bvax compared to B_Naive. Of note, antigen processing and presentation pathway is upregulated in Bvax compared to B_Naive. (C) Differential gene expression analysis between B_Vax and B_Naive with a curated gene list based on antigen uptake, processing and presentation gene sets. Heatmap was generated based on z scores calculated from FPKM. (D) Quantitative PCR analysis showing fold-change expression of IL2, IL7, IL15, and IL15Ra in B_Vax compared to B_Naive. (E) CD8⁺ T cell viability after 3 days of culture with B_Vax or B_Naive with or without exogenous IL15. (F) Proliferation of OT1 T cells after co-cultured with B_Vax, B_Naive or without APC (all supplemented with exogenous IL15) for 3 days. Histogram represented mean ± SEM. ns, p>0.05, ***p<0.001, ****p<0.0001.

Figure 3

DC and B_Vax drive different T cell proliferation patterns in vitro. Proliferation dye-labeled B_Vax- or DC-activated CD8+ T cells for 3 days with the addition of 6.25ng/ml IL15 on Day 0 and 48h post-co-culture. Abbreviations were used as High: highly proliferative cells; Interm: Intermediately proliferative cells. (A) In-vitro proliferation of CD8 ⁺ T cells was measured by proliferation dye (eFluor450) dilution vs CD69 expression for B_Vax, DC, B_Naive, and no antigen-presenting cells (APC). Highly proliferative CD8 ⁺ T cells were identified with the lowest proliferation dye expression. Intermediately differentiated CD8 ⁺ T cells were the middle transitional population. No APC was used as baseline proliferation control. (B) Total proliferation (High + intermediately proliferative) in all CD8 ⁺ T cells and the ratio of highly to intermediately proliferative CD8 ⁺ T cells were plotted with mean ± SEM. APC were co-cultured with CD8 ⁺ T cells at 1:2, 1:6, 1:8, and 1:10 ratios. (C) CD44 vs CD62L expression in intermediately differentiated (red) and highly differentiated (blue) CD8+ T cells. (D) Expression of PD-1, TCF-1, and Granzyme B in total CD8 ⁺ T cell activated with B_Vax (red) or DC (blue) at 1:2 APC : CD8 ratio. (E–G) characterization of highly proliferative CD8 ⁺ T cells by the expression of PD-1 (E), TCF-1 (F), and Granzyme B (G). (H) After adoptively transferring 3M ex-vivo activated CD8 ⁺ T cells to tumor-bearing Rag1 KO mice through an intracranial cannula, hosts’ brains were collected 3 days post-treatment for immunophenotyping. The expression of PD-1, TCF-1, and Granzyme B of the donor CD8 ⁺ cells were shown in the histogram with mean ± SEM. ns, p>0.05, *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. (I) TCR clone overlaps from naïve (No APC), B_Vax, and DC-activated CD8⁺ T cells TCR sequencing.

Figure 4

B_Vax expands stem-like CD8 ⁺ T cell in vivo. (A) Schema of co-delivery APC and CD8 ⁺ T cell and immunophenotyping. Briefly, CD45.1 host mice were implanted with 100K CT2A tumor cells through a cannula system and received a lymphodepleting dose of brain irradiation for 3 consecutive days. Co-delivery of 1.5M APC and 5M CD8 ⁺ T cells from NUR-GFP⁺ (CD45.2) tumor-bearing donors. 3 days post-treatment, tumors were collected and processed for immunophenotyping. (B) Representative flow cytometry plot indicated tumor microenvironment (TME) mostly comprised of donor CD8 ⁺ T cells (CD45.2) and host CD11b⁺ cells (CD45.1), whereas all the donor APC didn’t survive 3 days post treatment. (C) B_Vax-activated CD8 ⁺ T cells accumulated more than DC-activated CD8 ⁺ T cells or No APC-CD8 ⁺ T cells in the TME as quantified by the percentage of CD8 in CD45⁺ cells. (D) In a parallel experiment, CD8 ⁺ T cells were labeled with CellTrace Violet before adoptive transfer into tumor-bearing mice. B_Vax-activated CD8 ⁺ T cells (pink) showed more total proliferation than DC-activated CD8+ T cells and no APC-CD8+ T cells as measured by loss of proliferation dye and proliferation. (E) Representative flow cytometry plot of T cell proliferation dye expression after adoptive transfer of CD8 ⁺ T cells with B_Vax, DC, or no APCs. The histogram shows the ratio of highly proliferative T cells to intermediately proliferative T cells in each of the 3 groups. (F) Flow cytometry measured the expression of PD-1 vs proliferation status of donor CD8+ T cells co-implanted with B_Vax (pink), DC (blue), and CD8+ T cell monotherapy. Pre-adoptive transfer CD8 ⁺ T cells were freshly isolated from spleens and lymph nodes of tumor-bearing mice. Reduced PD-1 expression was observed in highly proliferative CD8 T cells co-implanted with B_Vax. All histograms plotted mean ± SEM. ns, p>0.05, *p<0.05, ***p<0.001.