Antigen-independent acquisition of MHC class II molecules by human T lymphocytes

Abstract

We report here that human T lymphocytes have the capacity of acquiring large amounts of MHC class II molecules from various types of antigen-presenting cells (APC) in an antigen-independent manner. The transfer of MHC class II molecules from APC to T cell required direct cell-to-cell contact and appeared to involve the interaction of numerous adhesion molecules between these cells. Depletion of cholesterol from the plasma membrane reduced the amount of MHC class II transferred onto the T cells. Most significantly, the newly acquired MHC class II molecules were capable of efficiently presenting antigen to T helper cells. These results suggest that T cells are able to interact with other T cells to regulate immune responses by presenting MHC peptide complexes that have been snatched away from nearby APC.

Fig. 1

Transfer of MHC class II from APC to T cells. M35-EBV cells, which express high levels of HLA-DR, were used as APC. (A) Solid line corresponds to the staining of M35-EBV cells obtained with anti-HLA-DR antibody and the thin line represents the antibody isotype control. BLS-CD8 T cells (B–D) or CD28+ Jurkat T cells (E–G) were incubated with (or without) M35-EBV cells at 1:1 ratio for 3 h and HLA-DR expression on the gated Tcell populations was examined by flow cytometry. Dot plots in (B) and (E) correspond to cultures containing Tcells and APC, while dot plots in (C) and (F) are the cultures containing T cells alone. The data presented in the histograms (D) and (G) correspond to the cell populations gated on CD8+ T cells (B and C) or CD28+ Jurkat cells (E and F). Thin lines in the histograms (D) and (G) represent HLA-DR staining of BLS-CD8 T cells and Jurkat cells not incubated with M35-EBV (data gated from C and F). Solid lines in the histograms (D) and (G) correspond to the data gated on panels (B) and (E). The ‘DR transfer index’ (DRTI) was calculated by dividing the MFI of HLA-DR staining of the cells incubated with APC over the MFI of HLA-DR staining the control T cells (not incubated with APC). The percentage of HLA-DR+ cells for each population is shown for each histogram curve, calculated using the gates established with isotype controls. These results are representative of three independent experiments.

Fig. 2

Transfer of MHC class II from APC to Tcells. BLS-CD8 Tcells were incubated with M35-EBV cells (at a 1:1 ratio) for 3 h and stained for CD8 (red) and HLA-DR (green) as described in Methods. (A) Control BLS-CD8 Tcells not incubated with APC. (B) Control M35-EBV cells not incubated with T cells. (C and D) Two representative examples of cells from mixtures of BLS-CD8 T cells with M35-EBV cells.

Fig. 3

Antigen-independent transfer of MHC class II from APC to CD8+ T cell clone. The X283 CTL clone was co-cultured with mouse B cell lymphoma LB27.4 at a 1:1 ratio at 37°C overnight. (A) High level of mouse MHC II molecules (IA^b) on the LB27.4 cells (solid line, anti-IA^b antibody; dashed line, isotype control). (B) The amount of transferred mouse MHC class II molecules (IA^b) was measured by flow cytometry using anti-IA^b monoclonal antibody gating on human CD8+ T cells (solid line, X283 T cells co-incubated with LB27,4; dashed line, X283 T cells alone). This experiment was repeated twice with similar results.

Fig. 4

MHC class II transfer to Tcells by various APC. BLS-CD8 Tcells were co-cultured with human dendritic cells (A), with the TL-Hir HLA-DR+ T cell leukemia (B) or with normal (non-transformed) activated human B cells (C) for 3 h, and the HLA-DR expression on CD8⁺ T cells was analyzed by flow cytometry. Analysis was done as described in the legend to Fig. 1, and the DRTI is indicated. Dashed lines represent HLA-DR staining of BLS-CD8 Tcells not incubated with APC and solid lines the DR staining of the T cells co-incubated with APC. Histograms shown as inserts in each panel correspond to the levels of MHC class II found on each of the APC type used, where solid lines are the staining obtained with anti-HLA-DR antibody and the dashed lines are the isotype controls. These results are representative of three independent experiments.

Fig. 5

Direct cell–cell contact is required for MHC class II transfer. BLS-CD8 T cells and M-35 EBV cells (1:1) were co-cultured mixed (A) or separated by a semi-permeable membrane (0.4 μm) in a transwell dish (B) for 3 h. In a separate experiment, BLS-CD8 Tcells were co-cultured for 24 h with either M35-EBV cells (C), supernatant from a 2 day old M35-EBV cell culture (D) or exosomes purified from M35-EBV cells. The HLA-DR expression on CD8⁺ gated BLS-CD8 T cells was measured by flow cytometry and analyzed as described in the legend to Fig. 1, and the DRTI is indicated. Thin lines represent HLA-DR staining of the control BLS-CD8 T cells (not incubated with APC, supernatant or exosomes). These results are representative of three independent experiments.

Fig. 6

Kinetics of MHC class II transfer. BLS-CD8 (left panels) or Jurkat T cells (right panels) were incubated overnight with CFSE-labeled M35-EBV cells. Next day, the CFSE negative T cells were separated by flow sorting, placed on fresh medium and analyzed immediately (Day 0) and at various time points (Days 1 and 3) for HLA-DR expression by flow cytometry.

Fig. 7

Antibodies to adhesion molecules block MHC class II transfer to Tcells. BLS CD8 T cells and M35-EBV APC were co-incubated (1:1) in the presence of blocking antibodies at 10 μg/ml for 3 h and HLA-DR expression on BLS-CD8 Tcells was determined by flow cytometry, calculating the DRTI as explained in Methods. The numbers inside the bars indicate % inhibition of HLA-DR transfer. *P-values were derived by comparing the no antibody-treated sample (black bar) with each antibody treated sample (gray bars). mIg = normal mouse immunoglobulin control. These experiments were repeated three times with nearly identical results.

Fig. 8

Cholesterol depletion inhibits MHC class-II transfer to T cells. BLS-CD8 T cells (A and B) or Jurkat T cells (C and D) were incubated with 10 mM MCD (B and D) or in medium alone (A and C) for 10 min at room temperature. After this incubation, the T cells were co-cultured with M35-EBV cells (at 1:1 ratio) for 3 h and HLA-DR expression was analyzed by flow cytometry as described in previous figure legends.

Fig. 9

T cells with acquired MHC class II can function as APC for helper T cells. BLS-CD8 T cells were incubated with CFSE labeled, peptide-pulsed (10 μg/ml EBNA2_280–290 peptide for 4 h) M83-EBV cells (HLA-DQ2+) at a 1:1 ratio overnight. Next day, the CFSE negative T cells were purified by flow sorting (>98% purity, panel A) and were co-incubated with antigen specific HLA-DQ2 restricted M14-HTL at 1:1 ratio for 5 h. Antigen response of the M14-HTL was measured by intracellular staining for IFNγ. The percentage of IFNγ positive M14-HTL was detected by flow cytometric analysis by gating on the CD4+ Tcell population. As a positive control, M14-HTL were activated by 50 ng/ml PMA plus 1 μg/ml calcium ionophore (B). IFNγ expression by M14-HTL (gated on CD4+ cells) incubated with BLS-CD8 cells that acquired MHC class II (C). IFNγ expression of M14-HTL incubated with control BLS-CD8 T cells, without M83-EBV pre-incubation (D). Since the contamination of the flow sorted BLS-CD8 Tcells was ~1% (A), as an additional control BLS-CD8 Tcells were mixed with 1% peptide pulsed M83-EBV cells and incubated with M14-HTL (E), demonstrating that this level of contamination was not sufficient to activate the M14-HTL. This experiment was repeated twice with similar results.