Intratumoral antigen signaling traps CD8+ T cells to confine exhaustion to the tumor site

Abstract

Immunotherapy advances have been hindered by difficulties in tracking the behaviors of lymphocytes after antigen signaling. Here, we assessed the behavior of T cells active within tumors through the development of the antigen receptor signaling reporter (AgRSR) mouse, fate-mapping lymphocytes responding to antigens at specific times and locations. Contrary to reports describing the ready egress of T cells out of the tumor, we find that intratumoral antigen signaling traps CD8⁺ T cells in the tumor. These clonal populations expand and become increasingly exhausted over time. By contrast, antigen-signaled regulatory T cell (T_reg) clonal populations readily recirculate out of the tumor. Consequently, intratumoral antigen signaling acts as a gatekeeper to compartmentalize CD8⁺ T cell responses, even within the same clonotype, thus enabling exhausted T cells to remain confined to a specific tumor tissue site.

Fig. 1. The AgRSR mouse fate-maps antigen-signaled CD8+ T cells.

(A) Transgenes in the antigen receptor signaling reporter (AgRSR) and AgRSR-LSL-EYFP mouse. (B) Lymphocytes receiving antigen and tamoxifen signaling, and their progenies are marked EYFP+. (C-D) Range of Katushka expression of OT-I x AgRSR-LSL-EYFP and WT CD8+ T cells following stimulation with SIINFEKL variant peptides, by flow cytometry histograms (C) and normalized fluorescence (D). Data are representative of two independent experiments. (E) AgRSR-LSL-EYFP splenocytes were adoptively transferred to B2m^-/- and RAG2^-/- strains, challenged with Listeria and tamoxifen treated. (F) Representative flow cytometry plots of splenocytes at day 7 post infection (Left), and summary plot of all mice (Right). Data are pooled from two independent experiments (n=7 per condition). (G) AgRSR-LSL-EYFP mice were immunized with SIINFEKL with or without tamoxifen. (H) Summary plot of all mice. Data are pooled from three independent experiments (n=6 per condition). Dots represent mice (F, H). Mean ± SEM as shown (D, F, H). Statistical testing via unpaired two-tailed students t-test (****, p <0.0001; ***, p <0.001).

Fig. 2. The AgRSR mouse tracks expansion of antigen-signaled CD8+ T cells in the tumor immune response.

(A) AgRSR-LSL-EYFP mice were tamoxifen treated pre- (5 days before) or post- (7 days after) subcutaneous injection of YUMMER1.7 melanoma cells, and CD8+ T cells were assessed 15 days later. (B) Representative flow cytometry histograms (Left) and summary plot of all experiments (Right). Data are pooled from two independent experiments (n≥5 per condition). (C-D) Representative flow cytometry histograms (Left) and summary plot (Right) of PD-1 expression in CD8+ T cells from the spleen (C) and the tumor (D) of CD8+ T cells 8 days after tamoxifen labelling. Data are pooled from two independent experiments (n=6). (E) AgRSR-LSL-EYFP mice were implanted with YUMMER1.7 melanoma cells, treated with tamoxifen 7 days later, and CD8+ T cells from the spleen, tumor, draining lymph node and non-draining lymph node were sampled at indicated days. (F) EYFP+ percentages of CD8+ T cells in the indicated tissues. Data are pooled from four independent experiments (n≥5 per condition). Dots represent mice (B, C, D, F). Mean ± SEM as shown (B, C, D, F). Statistical testing via paired two-tailed students t-test and ordinary one-way ANOVA (****, p <0.0001; **, p <0.01).

Fig. 3. Antigen-signaled CD8+ T and Treg cells partition their differentiation state by tissue site.

(A) EYFP+ T cells were sorted from the tumor and spleen of AgRSR-LSL-EYFP mice 8 days after intraperitoneal tamoxifen injection and subject to single-cell RNA and VDJ analysis (n=4). (B) Largest antigen-signaled T cell clonal populations ranked by size, displaying clonal population size (frequency) within the tumor or spleen. (C-G) Antigen-signaled CD8+ T cell (C-E) and Treg (in F-G) clonal populations with at least two cell members in both the tumor and spleen assessed for mean exhaustion (C), cytotoxicity (D) and effector Treg gene set expression score (F) in each tissue. Scores from the same antigen-signaled clonal population are linked by a line. (E, G) Violin plot comparing the gene set scores of individual cells in G2M/ S phase across tissues. Dots represent cells (E, G) and antigen-signaled clonal populations (C, D, F). Statistical testing via paired two-tailed students t-test and Kruskal-Wallis test (****, p <0.0001). (AU) arbitrary units.

Fig. 4. Intratumoral antigen signaling traps CD8+ T cells.

(A) OT-I CD8+ T cells from the whole tumor, draining and non-draining lymph nodes, and 1/5^th of the spleen were analyzed 8 days after intratumoral tamoxifen administration of B16-OVA bearing congenic marker mice that had been injected with activated OT-I x AgRSR-LSL-EYFP mice 4 days earlier (n=3). (B-C) Distribution of OT-I CD8+ T cells (B) and EYFP+ OT-I CD8+ T cells (C) across different tissues with enlarged view of lymphoid tissues (Top right of respective figures). Asterisk denotes 1/5^th of the spleen was sampled. (D) EYFP+ and EYFP- CD8+ T cells were sorted from the tumor and draining lymph node of AgRSR-LSL-EYFP mice 8 days after intratumoral tamoxifen injection and subject to bulk TCR-seq analysis (n=3). The top 200 largest EYFP+ CD8+ T cell clonal populations in the tumor of each mouse were analyzed. (E) Distribution of cells across the tumor and draining lymph node, with enlarged view of the draining lymph node (Top right). (F) Fraction of EYFP+ clonal populations detected exclusively in the tumor that had TCR overlap with EYFP -cells in the draining lymph node. Dots represent mice (B, C, E, F).

Fig. 5. Intratumoral antigen-signals trap and exhaust CD8+ T cells but not Tregs.

(A) EYFP+ T cells were sorted from the whole pre-enriched tumor and spleen of AgRSR-LSL-EYFP mice 8 days after intratumoral tamoxifen injection and subject to single-cell RNA and VDJ analysis (n=3). (B) Largest antigen-signaled T cell clonal populations ranked by size and colored by tissue origin. (C) Fraction of antigen-signaled CD8+ T cell and Treg clonal populations for which no cells with the same TCR were identified in EYFP+ cells of the spleen. (D) Effect of intratumoral antigen signaling in CD8+ T cell and Treg clonal populations. (E-F) The average exhaustion and effector Treg score of antigen-signaled CD8+ T cell (E) and Treg (F) clonal populations obtained from intraperitoneal fate-mapping (Left section – as in Fig. 3C, 3F) and intratumoral fate-mapping (Right section and labelled with arrows for Treg clonal populations). Dots represent antigen-signaled clonal populations (E, F) and mice (C). Statistical testing via paired two-tailed students t-test and Kruskal-Wallis test (****, p <0.0001; *, p<0.05). (AU) arbitrary units.

Fig. 6. Time elapsed since antigen signaling impacts the differentiation state of CD8+ T cells.

(A) EYFP+ T cells were sorted from the tumor and spleen of AgRSR-LSL-EYFP mice 18 days after intraperitoneal tamoxifen injection and subject to single-cell RNA and VDJ analysis (n=3). (B) Largest antigen-signaled T cell clonal populations ranked by size, displaying clonal population size (frequency) within the tumor or spleen. (C) TrajClust, a computational algorithm to cluster clonal differentiation patterns, was applied to simulated datasets containing clonal populations with five different differentiation patterns. Results from an unsupervised clustering of these clonal populations by established methods (Left) and TrajClust (Right) (D) Unsupervised clustering of the largest antigen-signaled CD8+ T cell clonal populations found 8 days (Fig. 3A) or 18 days (Fig. 6A) after intraperitoneal tamoxifen injection by TrajClust. Each cluster is denoted clonal differentiation pattern 1-4. (E) Pie chart showing the properties of clonal populations from each clonal differentiation pattern. Clonal populations are labelled by their TCR reactivity groups identified by GLIPH2 analysis (Top), their tissue distribution (Middle) and their time elapsed since antigen signaling (Bottom).

Fig. 7. Spatial distribution changes of antigen-signaled T cell clonal populations.

(A-G) Tracking changes in antigen-signaled T cell clonal populations over days 8 and 18. (A, E) Changes in the fraction of cells in G2M/ S phase across tissues for antigen-signaled CD8+ T cell (A) and Treg (E) populations. (B, F) Changes in the spatial distribution bias of antigen-signaled CD8+ T cell (B) and Treg (F) clonal populations. The spatial distribution bias was calculated by dividing the clonal population size (frequency) in the spleen by that of the tumor. (C, G) Changes in the mean exhaustion and Treg effector gene set score of tumor cells from antigen-signaled CD8+ T cell (C) and Treg (G) clonal populations. (D) Mean exhaustion and cytotoxic scores of antigen-signaled CD8+ T cell clonal populations, restricted to analysis of their cells from the tumor (Small dots) or the secondary lymphoid tissue (Large cross – mean of all clonal populations). Individual clonal populations are colored by time elapsed since antigen signaling. Dots represent antigen-signaled clonal populations (A-G). Statistical testing via paired two-tailed students t-test and Kruskal-Wallis test (****, p <0.0001; ***, p <0.001; *, p<0.05; ns, p>0.05). (AU) arbitrary units.