Surface cytotoxic T lymphocyte-associated antigen 4 partitions within lipid rafts and relocates to the immunological synapse under conditions of inhibition of T cell activation

Abstract

T cell activation through the T cell receptor (TCR) involves partitioning of receptors into discrete membrane compartments known as lipid rafts, and the formation of an immunological synapse (IS) between the T cell and antigen-presenting cell (APC). Compartmentalization of negative regulators of T cell activation such as cytotoxic T lymphocyte-associated antigen-4 (CTLA-4) is unknown. Recent crystal structures of B7-ligated CTLA-4 suggest that it may form lattices within the IS which could explain the mechanism of action of this molecule. Here, we show that after T cell stimulation, CTLA-4 coclusters with the TCR and the lipid raft ganglioside GM1 within the IS. Using subcellular fractionation, we show that most lipid raft-associated CTLA-4 is on the T cell surface. Such compartmentalization is dependent on the cytoplasmic tail of CTLA-4 and can be forced with a glycosylphosphatidylinositol-anchor in CTLA-4. The level of CTLA-4 within lipid rafts increases under conditions of APC-dependent TCR-CTLA-4 coligation and T cell inactivation. However, raft localization, although necessary for inhibition of T cell activation, is not sufficient for CTLA-4-mediated negative signaling. These data demonstrate that CTLA-4 within lipid rafts migrates to the IS where it can potentially form lattice structures and inhibit T cell activation.

Figure 1.

Inducible surface expression of wild-type or mutant CTLA-4 on Jurkat T cell clones. (A) Jurkat T cells stably transfected for CTLA-4 on a doxycycline inducible enhancer-promoter. T cells were cultured with 5 μg/ml of doxycycline overnight (solid line) or without it (dotted line), and were analyzed by FACS^® for surface expression of CTLA-4, CD3, CD28, and CD45. Shown are representative T cell clones stably transfected with the wild-type CTLA-4 gene, a version of CTLA-4 that lacks the cytoplasmic tail (tailless CTLA-4), and a chimerical molecule consisting of the extracellular domain of CTLA-4 fused to the GPI-anchor of CD58 (GPI-anchored CTLA-4). The staining control is shown in gray. Note that the very high expression of GPI-anchored and tailless CTLA-4 after doxycycline induction is above 10⁴. (B) Percentage of wild-type (WT), tailless (TL), or GPI-anchored CTLA-4 that is expressed on the cell surface. T cell transfectants were biotinylated and underwent immunoprecipitation for biotinylated, i.e., surface, molecules (S), for total CTLA-4 (T), or for an isotype-matched irrelevant antibody control (C), using antibodies covalently cross-linked to beads. These immunoprecipitates were subsequently blotted for CTLA-4. Bands were analyzed by densitometry and plotted as percentage of total CTLA-4 that was biotinylated.

Figure 1.

Inducible surface expression of wild-type or mutant CTLA-4 on Jurkat T cell clones. (A) Jurkat T cells stably transfected for CTLA-4 on a doxycycline inducible enhancer-promoter. T cells were cultured with 5 μg/ml of doxycycline overnight (solid line) or without it (dotted line), and were analyzed by FACS^® for surface expression of CTLA-4, CD3, CD28, and CD45. Shown are representative T cell clones stably transfected with the wild-type CTLA-4 gene, a version of CTLA-4 that lacks the cytoplasmic tail (tailless CTLA-4), and a chimerical molecule consisting of the extracellular domain of CTLA-4 fused to the GPI-anchor of CD58 (GPI-anchored CTLA-4). The staining control is shown in gray. Note that the very high expression of GPI-anchored and tailless CTLA-4 after doxycycline induction is above 10⁴. (B) Percentage of wild-type (WT), tailless (TL), or GPI-anchored CTLA-4 that is expressed on the cell surface. T cell transfectants were biotinylated and underwent immunoprecipitation for biotinylated, i.e., surface, molecules (S), for total CTLA-4 (T), or for an isotype-matched irrelevant antibody control (C), using antibodies covalently cross-linked to beads. These immunoprecipitates were subsequently blotted for CTLA-4. Bands were analyzed by densitometry and plotted as percentage of total CTLA-4 that was biotinylated.

Figure 2.

CTLA-4 colocalizes in the IS with CD3 and GM1. Wild-type CTLA-4 transfected Jurkat T cells were induced with doxycycline or noninduced and incubated on poly-l-lysine coated confocal dishes with APC plus 100 ng/ml SEE where indicated. Similarly, tailless and GPI-anchored CTLA-4 in the absence of doxycycline (as these cells express significant amounts of CTLA-4 without it) were used. The samples were put on ice to prevent receptor internalization and immunostained for (A) CD3 (green) and CTLA-4 (red) or (B) GM1 (green) and CTLA-4 (red). The samples were then analyzed by confocal microscopy. Co-capping of CD3 and CTLA-4 was scored when both colocalized at the interface between T cell and APC and the cluster stained <50% of the cell surface. At least 100 doublets per group were counted and scored and results are shown at the bottom right of Fig. 2 A. APCs were identified based on their morphology and lack of staining for CTLA-4 and CD3. Light field pictures were taken simultaneously and merged with the color fields. Yellow indicates overlay of red and green signals.

Figure 2.

CTLA-4 colocalizes in the IS with CD3 and GM1. Wild-type CTLA-4 transfected Jurkat T cells were induced with doxycycline or noninduced and incubated on poly-l-lysine coated confocal dishes with APC plus 100 ng/ml SEE where indicated. Similarly, tailless and GPI-anchored CTLA-4 in the absence of doxycycline (as these cells express significant amounts of CTLA-4 without it) were used. The samples were put on ice to prevent receptor internalization and immunostained for (A) CD3 (green) and CTLA-4 (red) or (B) GM1 (green) and CTLA-4 (red). The samples were then analyzed by confocal microscopy. Co-capping of CD3 and CTLA-4 was scored when both colocalized at the interface between T cell and APC and the cluster stained <50% of the cell surface. At least 100 doublets per group were counted and scored and results are shown at the bottom right of Fig. 2 A. APCs were identified based on their morphology and lack of staining for CTLA-4 and CD3. Light field pictures were taken simultaneously and merged with the color fields. Yellow indicates overlay of red and green signals.

Figure 3.

Wild-type CTLA-4 is within lipid rafts in a tail-dependent manner. Lipid raft (R) and detergent-soluble (S) fractions were isolated from resting Jurkat T cells transfected with wild-type CTLA-4, tailless CTLA-4, or GPI-anchored CTLA-4 molecules induced with doxycycline or left noninduced. (A) Selected fractions were run on SDS PAGE and Western blotted for CTLA-4. Where indicated by an asterisk (*) the samples were diluted 10-fold with 1× sample buffer before loading in order to assess nonsaturated signals for CTLA-4 in these samples. CTLA-4 is detected as a broad band attributed to differential glycosylation and intermediate GPI-anchored molecules. (B) Band intensities from the film in panel A were quantified by densitometry and used to calculate the percentage of total CTLA-4 within lipid rafts (see Materials and Methods). (C) The fractions used in the experiment (A) were immunoblotted for ERK-1/-2, CD45, and GM1 to control for quality of the separation. ERK-1/-2 and CD45 were found in the soluble fraction while GM1 partitioned within lipid rafts.

Figure 3.

Wild-type CTLA-4 is within lipid rafts in a tail-dependent manner. Lipid raft (R) and detergent-soluble (S) fractions were isolated from resting Jurkat T cells transfected with wild-type CTLA-4, tailless CTLA-4, or GPI-anchored CTLA-4 molecules induced with doxycycline or left noninduced. (A) Selected fractions were run on SDS PAGE and Western blotted for CTLA-4. Where indicated by an asterisk (*) the samples were diluted 10-fold with 1× sample buffer before loading in order to assess nonsaturated signals for CTLA-4 in these samples. CTLA-4 is detected as a broad band attributed to differential glycosylation and intermediate GPI-anchored molecules. (B) Band intensities from the film in panel A were quantified by densitometry and used to calculate the percentage of total CTLA-4 within lipid rafts (see Materials and Methods). (C) The fractions used in the experiment (A) were immunoblotted for ERK-1/-2, CD45, and GM1 to control for quality of the separation. ERK-1/-2 and CD45 were found in the soluble fraction while GM1 partitioned within lipid rafts.

Figure 3.

Wild-type CTLA-4 is within lipid rafts in a tail-dependent manner. Lipid raft (R) and detergent-soluble (S) fractions were isolated from resting Jurkat T cells transfected with wild-type CTLA-4, tailless CTLA-4, or GPI-anchored CTLA-4 molecules induced with doxycycline or left noninduced. (A) Selected fractions were run on SDS PAGE and Western blotted for CTLA-4. Where indicated by an asterisk (*) the samples were diluted 10-fold with 1× sample buffer before loading in order to assess nonsaturated signals for CTLA-4 in these samples. CTLA-4 is detected as a broad band attributed to differential glycosylation and intermediate GPI-anchored molecules. (B) Band intensities from the film in panel A were quantified by densitometry and used to calculate the percentage of total CTLA-4 within lipid rafts (see Materials and Methods). (C) The fractions used in the experiment (A) were immunoblotted for ERK-1/-2, CD45, and GM1 to control for quality of the separation. ERK-1/-2 and CD45 were found in the soluble fraction while GM1 partitioned within lipid rafts.

Figure 4.

Most lipid raft-associated CTLA-4 is on the cell surface. (A) Wild-type, tailless, and GPI-anchored CTLA-4 T cell transfectants were induced with doxycycline or noninduced, surface biotinylated, and separated into raft and soluble fractions. These fractions were immunoprecipitated for CTLA-4 and for biotin, and Western blotted for CTLA-4. The light chain of the immunoprecipitating antibodies is indicated. The GPI-anchored and tailless CTLA-4 samples were diluted 10-fold before loading on the gel, based on total expression of CTLA-4 to prevent signal saturation. In addition, three times fewer cells were used for tailless CTLA-4 and GPI-anchored CTLA-4 transfectants compared with wild-type CTLA-4 transfectants. Antibody-coated beads rotated in lysis buffer are shown as negative control. Representative isotype-matched control antibody immunoprecipitations (Ctrl.) are shown. (B) CTLA-4 immunoprecipitates of lipid raft and soluble fractions from noninduced and induced, biotinylated wild-type CTLA-4 transfected T cells were sequentially immunoblotted with avidin and with anti–CTLA-4 antibody. Signals were quantified and the ratio of biotinylated CTLA-4 (sCTLA-4) over total CTLA-4 (tCTLA-4) for each fraction plotted in percentage.

Figure 4.

Most lipid raft-associated CTLA-4 is on the cell surface. (A) Wild-type, tailless, and GPI-anchored CTLA-4 T cell transfectants were induced with doxycycline or noninduced, surface biotinylated, and separated into raft and soluble fractions. These fractions were immunoprecipitated for CTLA-4 and for biotin, and Western blotted for CTLA-4. The light chain of the immunoprecipitating antibodies is indicated. The GPI-anchored and tailless CTLA-4 samples were diluted 10-fold before loading on the gel, based on total expression of CTLA-4 to prevent signal saturation. In addition, three times fewer cells were used for tailless CTLA-4 and GPI-anchored CTLA-4 transfectants compared with wild-type CTLA-4 transfectants. Antibody-coated beads rotated in lysis buffer are shown as negative control. Representative isotype-matched control antibody immunoprecipitations (Ctrl.) are shown. (B) CTLA-4 immunoprecipitates of lipid raft and soluble fractions from noninduced and induced, biotinylated wild-type CTLA-4 transfected T cells were sequentially immunoblotted with avidin and with anti–CTLA-4 antibody. Signals were quantified and the ratio of biotinylated CTLA-4 (sCTLA-4) over total CTLA-4 (tCTLA-4) for each fraction plotted in percentage.

Figure 5.

CTLA-4 in lipid rafts is homodimeric. Lipid rafts and soluble fractions from doxycycline induced and noninduced wild-type CTLA-4 T cell transfectants were immunoprecipitated for CTLA-4 or an isotype-matched irrelevant control (Ctrl.). Immunoprecipitates were analyzed by SDS PAGE in nonreducing conditions (NR) and reducing (R) conditions, and Western blotted for CTLA-4.

Figure 6.

The levels of CTLA-4 within lipid rafts increase under conditions of T cell inhibition. (A) Doxycycline-induced, wild-type T cell transfectants were incubated with APC plus or minus 100 ng/ml SEE for 30 min, and separated into lipid raft and soluble fractions. These fractions were immunoprecipitated for CTLA-4 and Western blotted for CTLA-4. The CTLA-4 immunoprecipitations from the soluble fractions were diluted 10-fold to prevent signal saturation. (B) Wild-type T cell transfectants were stimulated as in panel A and then surface biotinylated before raft isolation. Biotinylated proteins were immunoprecipitated from the fractions and Western blotted for CTLA-4. The soluble fractions were not diluted before loading. (C) Wild-type T cell transfectants were stimulated with SEE and APC for 48 h in the absence (−) or presence (+) of doxycycline to induce CTLA-4 expression. The amount of IL-2 in the supernatant was determined by ELISA. (D) Doxycycline induced Jurkat T cells were preincubated with or without anti–human CTLA-4 ScFv F′ab blocking single chain fragment for 30 min. Then, APCs and SEE were added to the cells and IL-2 production was measured after 48 h. Results were statistically significant as analyzed by one-way ANOVA (P < 0.05). (E) Non-induced and doxycycline-induced wild-type–transfected Jurkat T cells were stimulated with APC and SEE for 30 min and CTLA-4 expression on the cell surface was analyzed by flow cytometry. (Non-induced, nonstimulated, control antibody: light dotted line; noninduced, nonstimulated, CTLA-4 stained: thin line; doxycycline-induced, nonstimulated cells: dashed line; doxycycline-induced, stimulated cells: thick line.

Figure 6.

The levels of CTLA-4 within lipid rafts increase under conditions of T cell inhibition. (A) Doxycycline-induced, wild-type T cell transfectants were incubated with APC plus or minus 100 ng/ml SEE for 30 min, and separated into lipid raft and soluble fractions. These fractions were immunoprecipitated for CTLA-4 and Western blotted for CTLA-4. The CTLA-4 immunoprecipitations from the soluble fractions were diluted 10-fold to prevent signal saturation. (B) Wild-type T cell transfectants were stimulated as in panel A and then surface biotinylated before raft isolation. Biotinylated proteins were immunoprecipitated from the fractions and Western blotted for CTLA-4. The soluble fractions were not diluted before loading. (C) Wild-type T cell transfectants were stimulated with SEE and APC for 48 h in the absence (−) or presence (+) of doxycycline to induce CTLA-4 expression. The amount of IL-2 in the supernatant was determined by ELISA. (D) Doxycycline induced Jurkat T cells were preincubated with or without anti–human CTLA-4 ScFv F′ab blocking single chain fragment for 30 min. Then, APCs and SEE were added to the cells and IL-2 production was measured after 48 h. Results were statistically significant as analyzed by one-way ANOVA (P < 0.05). (E) Non-induced and doxycycline-induced wild-type–transfected Jurkat T cells were stimulated with APC and SEE for 30 min and CTLA-4 expression on the cell surface was analyzed by flow cytometry. (Non-induced, nonstimulated, control antibody: light dotted line; noninduced, nonstimulated, CTLA-4 stained: thin line; doxycycline-induced, nonstimulated cells: dashed line; doxycycline-induced, stimulated cells: thick line.

Figure 6.

The levels of CTLA-4 within lipid rafts increase under conditions of T cell inhibition. (A) Doxycycline-induced, wild-type T cell transfectants were incubated with APC plus or minus 100 ng/ml SEE for 30 min, and separated into lipid raft and soluble fractions. These fractions were immunoprecipitated for CTLA-4 and Western blotted for CTLA-4. The CTLA-4 immunoprecipitations from the soluble fractions were diluted 10-fold to prevent signal saturation. (B) Wild-type T cell transfectants were stimulated as in panel A and then surface biotinylated before raft isolation. Biotinylated proteins were immunoprecipitated from the fractions and Western blotted for CTLA-4. The soluble fractions were not diluted before loading. (C) Wild-type T cell transfectants were stimulated with SEE and APC for 48 h in the absence (−) or presence (+) of doxycycline to induce CTLA-4 expression. The amount of IL-2 in the supernatant was determined by ELISA. (D) Doxycycline induced Jurkat T cells were preincubated with or without anti–human CTLA-4 ScFv F′ab blocking single chain fragment for 30 min. Then, APCs and SEE were added to the cells and IL-2 production was measured after 48 h. Results were statistically significant as analyzed by one-way ANOVA (P < 0.05). (E) Non-induced and doxycycline-induced wild-type–transfected Jurkat T cells were stimulated with APC and SEE for 30 min and CTLA-4 expression on the cell surface was analyzed by flow cytometry. (Non-induced, nonstimulated, control antibody: light dotted line; noninduced, nonstimulated, CTLA-4 stained: thin line; doxycycline-induced, nonstimulated cells: dashed line; doxycycline-induced, stimulated cells: thick line.

Figure 6.

The levels of CTLA-4 within lipid rafts increase under conditions of T cell inhibition. (A) Doxycycline-induced, wild-type T cell transfectants were incubated with APC plus or minus 100 ng/ml SEE for 30 min, and separated into lipid raft and soluble fractions. These fractions were immunoprecipitated for CTLA-4 and Western blotted for CTLA-4. The CTLA-4 immunoprecipitations from the soluble fractions were diluted 10-fold to prevent signal saturation. (B) Wild-type T cell transfectants were stimulated as in panel A and then surface biotinylated before raft isolation. Biotinylated proteins were immunoprecipitated from the fractions and Western blotted for CTLA-4. The soluble fractions were not diluted before loading. (C) Wild-type T cell transfectants were stimulated with SEE and APC for 48 h in the absence (−) or presence (+) of doxycycline to induce CTLA-4 expression. The amount of IL-2 in the supernatant was determined by ELISA. (D) Doxycycline induced Jurkat T cells were preincubated with or without anti–human CTLA-4 ScFv F′ab blocking single chain fragment for 30 min. Then, APCs and SEE were added to the cells and IL-2 production was measured after 48 h. Results were statistically significant as analyzed by one-way ANOVA (P < 0.05). (E) Non-induced and doxycycline-induced wild-type–transfected Jurkat T cells were stimulated with APC and SEE for 30 min and CTLA-4 expression on the cell surface was analyzed by flow cytometry. (Non-induced, nonstimulated, control antibody: light dotted line; noninduced, nonstimulated, CTLA-4 stained: thin line; doxycycline-induced, nonstimulated cells: dashed line; doxycycline-induced, stimulated cells: thick line.

Figure 6.

The levels of CTLA-4 within lipid rafts increase under conditions of T cell inhibition. (A) Doxycycline-induced, wild-type T cell transfectants were incubated with APC plus or minus 100 ng/ml SEE for 30 min, and separated into lipid raft and soluble fractions. These fractions were immunoprecipitated for CTLA-4 and Western blotted for CTLA-4. The CTLA-4 immunoprecipitations from the soluble fractions were diluted 10-fold to prevent signal saturation. (B) Wild-type T cell transfectants were stimulated as in panel A and then surface biotinylated before raft isolation. Biotinylated proteins were immunoprecipitated from the fractions and Western blotted for CTLA-4. The soluble fractions were not diluted before loading. (C) Wild-type T cell transfectants were stimulated with SEE and APC for 48 h in the absence (−) or presence (+) of doxycycline to induce CTLA-4 expression. The amount of IL-2 in the supernatant was determined by ELISA. (D) Doxycycline induced Jurkat T cells were preincubated with or without anti–human CTLA-4 ScFv F′ab blocking single chain fragment for 30 min. Then, APCs and SEE were added to the cells and IL-2 production was measured after 48 h. Results were statistically significant as analyzed by one-way ANOVA (P < 0.05). (E) Non-induced and doxycycline-induced wild-type–transfected Jurkat T cells were stimulated with APC and SEE for 30 min and CTLA-4 expression on the cell surface was analyzed by flow cytometry. (Non-induced, nonstimulated, control antibody: light dotted line; noninduced, nonstimulated, CTLA-4 stained: thin line; doxycycline-induced, nonstimulated cells: dashed line; doxycycline-induced, stimulated cells: thick line.

Figure 7.

Wild-type CTLA-4, but not GPI-anchored CTLA-4, inhibits IL-2 production upon coligation with the TCR. (A) Wild-type CTLA-4 and GPI-anchored CTLA-4 were incubated with or without 100 ng/ml doxycycline, and incubated with anti-CD3 or anti-CD3 plus anti–CTLA-4 coated beads and soluble anti-CD28 for 48 h. IL-2 in the supernatant was determined by ELISA. The percentage change in IL-2 production in response to anti-CD3 plus anti-CTLA-4 is shown as a bar graph. The response to anti-CD3 beads was taken as 100% for each transfectant. (B) Doxycycline-induced Jurkat T cells were preincubated with or without blocking anti-CTLA-4 ScFv F′ab. After 30 min, beads coated with the indicated antibodies were added to the cells in the presence of soluble anti-CD28 antibodies. IL-2 production was measured after 48 h. Result were statistically significant as analyzed by one-way ANOVA (P < 0.05).

Figure 7.

Wild-type CTLA-4, but not GPI-anchored CTLA-4, inhibits IL-2 production upon coligation with the TCR. (A) Wild-type CTLA-4 and GPI-anchored CTLA-4 were incubated with or without 100 ng/ml doxycycline, and incubated with anti-CD3 or anti-CD3 plus anti–CTLA-4 coated beads and soluble anti-CD28 for 48 h. IL-2 in the supernatant was determined by ELISA. The percentage change in IL-2 production in response to anti-CD3 plus anti-CTLA-4 is shown as a bar graph. The response to anti-CD3 beads was taken as 100% for each transfectant. (B) Doxycycline-induced Jurkat T cells were preincubated with or without blocking anti-CTLA-4 ScFv F′ab. After 30 min, beads coated with the indicated antibodies were added to the cells in the presence of soluble anti-CD28 antibodies. IL-2 production was measured after 48 h. Result were statistically significant as analyzed by one-way ANOVA (P < 0.05).