Human T cell crosstalk is induced by tumor membrane transfer

Abstract

Trogocytosis is a contact-dependent unidirectional transfer of membrane fragments between immune effector cells and their targets, initially detected in T cells following interaction with professional antigen presenting cells (APC). Previously, we have demonstrated that trogocytosis also takes place between melanoma-specific cytotoxic T lymphocytes (CTLs) and their cognate tumors. In the present study, we took this finding a step further, focusing on the ability of melanoma membrane-imprinted CD8+ T cells to act as APCs (CD8+ T-APCs). We demonstrate that, following trogocytosis, CD8+ T-APCs directly present a variety of melanoma derived peptides to fraternal T cells with the same TCR specificity or to T cells with different TCRs. The resulting T cell-T cell immune synapse leads to (1) Activation of effector CTLs, as determined by proliferation, cytokine secretion and degranulation; (2) Fratricide (killing) of CD8+ T-APCs by the activated CTLs. Thus, trogocytosis enables cross-reactivity among CD8+ T cells with interchanging roles of effectors and APCs. This dual function of tumor-reactive CTLs may hint at their ability to amplify or restrict reactivity against the tumor and participate in modulation of the anti-cancer immune response.

Fig 1. CD8⁺T-APCs activate anti-tumor CD8 T cells of the same antigen specificity.

(A-C) Cytokine production by effector CTLs in response to activation by T-APCs. (A) DiIC₁₈-labeled CD8⁺T-APCs (see materials and methods) were incubated for 6 hours with surface-biotinylated effector CTLs. Cytokine-producing effector CTLs were defined based on intracellular IFN-γ or TNF-α staining of CD8⁺streptavidin⁺ lymphocytes. Numbers in upper right quadrants indicate the percentage of IFN-γ ⁺ (upper panel) or TNF-α ⁺ (lower panel) effector CTLs. Labels indicate cells used as targets for CTLs: CTLs co-cultured with 624mel melanoma cells are designated T-APC; CTLs co-cultured with irrelevant M171 melanoma cells are designated non T-APC. (B) Confocal images of cytokine-producing effector CTLs. Calcein AM labeled CD8⁺T-APCs (green, upper panel) or non T-APCs (green, lower panel) were co-cultured for 6 hours with streptavidin-allophycocyanin-stained effector CTLs (red). Intracellular TNF-α production (blue) by effector CTLs is shown. Scale bars are 10 μm (upper panel) and 20 μm (lower panel). (C) Time period that CD8⁺T-APCs activate effector CTLs. Effector CTLs were co-cultured with CD8⁺T-APCs either immediately or 6, 24 and 48 hours after CD8⁺T-APC purification. Data are mean ± SE (n = 3 replicates/group) percentage of IFN-γ ⁺ effector CTLs, gated on CD8⁺ T cells. (D) CD8⁺T-APCs trigger degranulation of effector CTLs. CD8⁺T-APCs were generated as described above (1A) and co-cultured with effector CTLs. Cytolytic activity of T cells was measured by detection of surface CD107A on CD8⁺streptavidin⁺ effector CTLs. Number in upper right quadrants indicates the percentage of CD107A⁺ streptavidin⁺ effector CTLs. Data are representative of at least three independent experiments.

Fig 2. CD8⁺ T-APCs induce degranulation of effector CTLs with different antigen specificity.

(A, B) CD8⁺ clones with different antigen specificity were used as CD8⁺T-APCs and effector CTLs. Biotinylated effector CTLs were co-cultured with CD8⁺T-APCs, stained with anti-CD107A mAb and streptavidin-allophycocyanin and analyzed by flow cytometry. (A) The gp100_154–162-specific clone 1G2 was used as CD8⁺T-APC for the MART-1_26–35-specific CTL clones (2E2 and 2D11). (B) The MART-1_26–35-specific clones 2E2 and 2D11 were used as CD8⁺T-APC for the gp100_154–162-specific clone (1G2). Numbers in upper right quadrants indicate the percentage of CD107A⁺streptavidin⁺ lymphocytes, gated on the CD8⁺ population (effector CTLs). Data are representative of three independent experiments. (C) Graphic presentation of intra- and inter-clonal T cell cross talk.

Fig 3. The effect of CD8⁺T-APCs on effector CTLs is mediated by tumor-derived pMHC.

(A) Ova-expressing EG7 and parental EL4 target cell lines (Left column) and target-entrained CD8⁺T-APC (generated following co-culture of OT-I CD8⁺ T cells with designated targets, right column) were labeled with Ova_257–264/H-2Kb-specific mAb (black histogram). Grey histogram, background staining with isotype control antibody. (B) Proliferation of OT-I CD8⁺ T cells stimulated with CD8⁺T-APCs. CFSE-labeled OT-I T cells were left untreated (no target) or co-cultured with the following T-APCs: OT-I CD8⁺ pre-incubated with EL4 (EL4) or OT-I CD8⁺ pre-incubated with EG7 cells (EG7). ConA stimulation was used as positive control (right). Scale bars, proliferating lymphocytes that divided at least twice. Numbers, percentage of dividing CD8⁺ T cells. Data are representative of two independent experiments.

Fig 4. CD8⁺T-APC induce secondary trogocytosis by tumor specific CTL.

(A, B) CD8⁺T-APCs and non-T-APCs were generated by co-incubation with cognate or irrelevant melanoma, respectively. They were then sorted, labeled with DiIC18 (see Methods) and co-cultured with surface biotinylated 2C7 or 2E2 clones (effector CTL). The culture was then stained with anti-CD8 antibodies and streptavidin- allophycocyanin, and subjected to flow cytometry. (A) Histograms indicate secondary trogocytosis by the presence of DiIC18 on 2C7 or 2E2 CTLs, gated on CD8⁺streptavidin⁺ populations, following co-culture with CD8⁺T-APC (blue) or non-T-APC (grey). (B) Secondary trogocytosis was measured by presence of DiIC18 on the CD8⁺streptavidin⁺ population (effector CTL) and streptavidin-allophycocyanin on the CD8⁺DiIC18⁺ population (left column, non-T-APC, right column, T-APC). Numbers in upper right quadrants indicate the percentage of DiIC18- stained effector CTL (performing secondary trogocytosis). (C) PKH67-labeled CD8⁺T-APCs (red) were co-cultured with PKH26-labeled effector CTLs (blue). The lymphocytes were co-incubated in a chambered cover-glass and subjected to confocal microscopy. A snapshot series of 8 min is presented. Arrows, transfer of membrane fragments (secondary trogocytosis) from CD8⁺T-APC to CTL. Scale bars, 15 μm. Data are representative of at least five independent experiments (A, B) or of three experiments (C).

Fig 5. Assessment of CTL fratricide activity based on detection of cleaved caspase-3.

MART-1- or MUC-1- reactive CTLs were co-cultured with DDAO-SE-tagged CD8⁺T-APCs and non T-APCs, generated by co-incubation of 2E2 cells with 624mel and M171 melanoma cells, respectively. T-APC damage was examined based on detection of intracellular cleaved caspase-3 in the DDAO-SE⁺CD8⁺ population. Numbers in upper right quadrants indicate the percentage of cleaved caspase 3-positive CD8⁺T-APC cells. Data are representative of three independent experiments.