The immune synapse clears and excludes molecules above a size threshold

Abstract

Natural killer cells assess target cell health via interactions at the immune synapse (IS) that facilitates signal integration and directed secretion. Here we test whether the IS also functions as a gasket. Quantitative fluorescence microscopy of nanometer-scale dextrans within synapses formed by various effector-target cell conjugates reveal that molecules are excluded in a size-dependent manner at activating synapses. Dextran sized ≤4 nm move in and out of the IS, but access is significantly reduced (by >50%) for dextran sized 10-13 nm, and dextran ≥32 nm is almost entirely excluded. Depolymerization of F-actin abrogated exclusion. Unexpectedly, larger-sized dextrans are cleared as the IS assembles in a zipper-like manner. Monoclonal antibodies are also excluded from the IS but smaller single-domain antibodies are able to penetrate. Therefore, the IS can clear and exclude molecules above a size threshold, and drugs designed to target synaptic cytokines or cytotoxic proteins must fit these dimensions.

Figure 1. Dextran is excluded from the activating pNK cell synapse in a size-dependent manner.

(a) Panels show bright-field (BF) images of pNK cells and 221 target cells overlaid with target cell nuclear staining (upper row) and the corresponding fluorescence image of dextran (lower row), size indicated. Scale bars, 10 μm. (b) Graph shows the mean raw fluorescence intensity of various sized dextrans (as indicated) for a line drawn perpendicular to the synapse. (c) Graph shows the relative intensity of dextran for each size tested. Error bars represent mean±s.d. n=38, 32, 33, 31, 38 and 35 conjugates from three independent experiments. Data were analysed using a one-way analysis of variance (ANOVA) with Tukey corrections. ****P<0.0001. (d) Graph shows the relative intensities of 4 and 54 nm dextran within the same synapse (connecting lines). n=29 conjugates from three independent experiments. Data were analysed using a Student’s t-test with Welch’s corrections. ****P<0.0001. (e) Graph shows the relative intensities of undiluted 4nm dextran or 32 nm dextran mixed 1:10 fluorescein-labelled:unlabelled within synapses formed by YTS cells and 221 cells. Error bars represent mean±s.d. n=32 conjugates for 4 nm dextran and 39 conjugates for 32 nm dextran from three independent experiments. Data were analysed using a two-tailed Student’s t-test with Welch’s corrections. ****P<0.0001. (f,g) Graphs show the relative intensities of 4 and 32 nm dextran within synapses formed by YTS cells and 221 cells for each labelled:unlabelled mixture ratio, as indicated. Error bars represent mean±s.d. n=23, 23 and 20 conjugates for 4 nm dextran and 22, 21 and 22 conjugates for 32 nm dextran from three independent experiments. Data were analysed using a one-way ANOVA with Tukey corrections. (h) Graph shows the relative intensities of 4 nm dextran at different concentrations, as indicated, within synapses formed by YTS cells and 221 cells. Error bars represent mean±s.d. n=26, 25, 25, 23, 23 and 29 conjugates from three independent experiments. Data were analysed using a one-way ANOVA with Tukey corrections.

Figure 2. Dextran is excluded in a size-dependent manner by activating but not inhibitory synapses.

(a,b) Graphs show the mean relative fluorescence intensity of dextran of the sizes indicated for synapses formed by pNK cells co-incubated with (a) Daudi cells or (b) K562 cells. Error bars show mean±s.d. of all data points. n=40, 39, 35 and 35 conjugates for Daudi cells and 40, 38, 38 and 42 conjugates for K562 cells from three independent experiments. Data were analysed using a one-way analysis of variance (ANOVA) with Tukey corrections. ****P<0.0001. (c) Panels show bright-field (BF) images of YTS/KIR2DL1 cells in conjugate with 221/Cw6 target cells overlaid with nuclear stained target cell (blue;upper row) and the corresponding fluorescence image of dextrans (lower row), sizes as indicated. Scale bars, 10 μm. (d) Graphs show mean raw fluorescence intensities of dextran (at the size indicated) along a line drawn perpendicular to the synapse. (e) Graph shows the mean relative fluorescence intensities of dextran within synapses formed by YTS/KIR2DL1 and 221/Cw6 cells. Bars show the mean for all data points. n=42, 45, 41, 45, 35 and 35 from three independent experiments. Data were analysed using a one-way ANOVA with Tukey corrections. (f) Panels show BF images overlaid with nuclear stained target cell (upper row) and the corresponding fluorescence image of dextran (lower row), size as indicated, of Jurkat cells in conjugate with superantigen-loaded Raji target cells. Scale bars, 10 μm. (g) Graph shows the mean relative fluorescence intensities of each dextran within synapses formed by T cells and Raji target cells. Bars show mean from all data points. n=34, 36, 33 and 37 from three independent experiments. Data were analysed using a one-way ANOVA with Tukey corrections. *P<0.1, ****P<0.0001.

Figure 3. Dextran below the size exclusion threshold can move in and out of the IS.

(a,b) Representative time-lapse confocal images of YTS and 221 cells before, and following, the addition of (a) 4 nm dextran or (b) 54 nm dextran (added at time=0). Panels show bright-field (BF) images overlaid with a nuclear stain marking target cells (blue) and corresponding fluorescence images of fluorescent dextran at the times were indicated. Scale bars, 10 μm. (c) Relative fluorescence intensity of 4 nm (·) and 54 nm (▪) dextran within the synapse at each time point, where time=0 indicates the time at which dextran was added. n=5 for each dextran size. (d) Representative images of a YTS cells co-incubated with 221 cells as the extracellular 4 nm dextran was diluted, final concentration as indicated. Panels show BF images overlaid with a nuclear stain marking target cells and corresponding fluorescence images of fluorescent dextran at each dilution. Scale bars, 10 μm. (e) Graph shows the mean relative fluorescence intensity of dextran within the synapse of the same conjugate as the extracellular dextran was diluted (connecting lines). n=8. Data were analysed using a one-way analysis of variance with Tukey corrections. (f,g) Graphs show relative fluorescence intensities of dextran in synapses formed by (f) YTS cells or (g) pNK cells co-incubated with 221 target cells, before (pre-wash) and following washing (post-wash). Fluorescence intensity of dextran in the synapse is shown relative to the mean intensity of the extracellular solution before washing. Error bars represent mean±s.d. n=38 and 40 for YTS cells and 35 and 39 for pNK cells from three independent experiments. Data were analysed using a two-tailed Student’s t-test with Welch’s corrections. ****P<0.0001.

Figure 4. Larger extracellular molecules are cleared from the IS during assembly.

(a,b) Panels show representative time-lapse bright-field (BF) images overlaid with target cell nuclear stain (upper row) and corresponding fluorescence image (lower row) at time points indicated of synapse formation between a YTS cell and a 221 cell in the presence of (a) 4 or (b) 54 nm dextran on fibronectin-coated slides. Time=0 indicates initial cell–cell contact. Scale bars, 10 μm. (c) Graph shows relative fluorescence intensity over time of 4 nm (−) and 32 nm (--) dextran within the synapses of YTS cells and 221 target cells before and during synapse formation. Time=0 indicates initial cell–cell contact. Red lines indicate 95% confidence interval (CI). n=5 for 4 nm dextran and n=5 for 32 nm dextran. (d,e) pNK cells were plated on slides coated with MICA-Fc and ICAM1-Fc in the presence of (d) 4 nm or (e) 32 nm dextran. Panels show interference reflection microscopy (IRM) image and corresponding fluorescence image of synapse formation after 500 s contact with the coated slides. Graphs show the mean fluorescence intensity (MFI) of 4 nm (red) and 32 nm (blue) dextran along the dotted lines indicated in fluorescence images. Scale bar, 10 μm. (f) Graph shows the mean relative intensity of dextran over time within the cell–slide interface during synapse assembly in the presence of 4 nm (−) and 32 nm (--) dextran. Red lines represent 95% CI. n=5 for 4 nm dextran and n=5 for 32 nm dextran. (g) pNK cells were plated with Texas Red-labelled 4 nm dextran and fluorescein-labelled 32 nm dextran on MICA-Fc- and ICAM1-Fc-coated slides. Cell–slide contact and synapse formation was then imaged by TIRF microscopy for 5 min, taking images every 20 s. Panels show representative images of initial contact (time=0) and synapse formation at the time points indicated. A HiLo threshold was used to identify pixel intensities below 32 a.u. (blue) and above 223 a.u. (red). Scale bars, 10 μm. Images are representative of six cells from five independent experiments.

Figure 5. The actin cytoskeleton is important for the exclusion of extracellular molecules from the synapse.

pNK cells were co-incubated with 221 target cells and fluorescein-labelled 32-nm-sized dextran. Conjugates were imaged before and following the addition of DMSO, 1 μM Lat A or 0.5 μM Jasp. Images show bright-field (BF) images overlaid with target cell nuclear staining and corresponding fluorescence images of 32 nm dextran before and following the addition of (a) DMSO, (c) 1 μM Lat A and (e) 0.5 μM Jasp. Scale bars, 10 μm. Graphs show the relative intensity of dextran within the same synapse at each time point (connecting lines) prior and after addition of (b) DMSO, (d) 1 μM Lat A and (f) 0.5 μM Jasp, as indicated. Data are from 7, 6 and 6 conjugates for DMSO, Lat A and Jasp, respectively. Data were analysed using a one-way analysis of variance with Tukey corrections. *P<0.1, ***P<0.001, ****P<0.0001.

Figure 6. Proteins are excluded from the IS in a size-dependent manner.

Representative images of (a) YTS and (c) pNK cells in contact with 221 target cells co-incubated with differently sized fluorescently labelled proteins as indicated. Panels show bright-field (BF) images overlaid with a nuclear stain marking target cells and corresponding fluorescence images of labelled proteins. Scale bars, 10 μm. (b) Graph shows the mean relative fluorescence intensities of each protein within synapses formed by YTS cells. Bars show mean of all data points. n=41, 59 and 49 from three independent experiments. Data were analysed using a one-way analysis of variance (ANOVA) with Tukey corrections. **P<0.01, ****P<0.0001. (d) Graph shows the mean relative fluorescence intensities of each protein within synapses formed by pNK cells. Bars show mean of all data points. n=38, 35, 39 and 54 from three independent experiments. Data were analysed using a one-way ANOVA with Tukey corrections. **P<0.01, ***P<0.001, ****P<0.0001.