TCR microclusters form spatially segregated domains and sequentially assemble in calcium-dependent kinetic steps

Abstract

Engagement of the T cell receptor (TCR) by stimulatory ligand results in the rapid formation of microclusters at sites of T cell activation. Whereas microclusters have been studied extensively using confocal microscopy, the spatial and kinetic relationships of their signaling components have not been well characterized due to limits in image resolution and acquisition speed. Here we show, using TIRF-SIM to examine the organization of microclusters at sub-diffraction resolution, the presence of two spatially distinct domains composed of ZAP70-bound TCR and LAT-associated signaling complex. Kinetic analysis of microcluster assembly reveal surprising delays between the stepwise recruitment of ZAP70 and signaling proteins to the TCR, as well as distinct patterns in their disassociation. These delays are regulated by intracellular calcium flux downstream of T cell activation. Our results reveal novel insights into the spatial and kinetic regulation of TCR microcluster formation and T cell activation.

Fig. 1

TIRF-SIM imaging of microclusters show distinct receptor and signaling domains. TIRF-SIM images of microclusters formed in Jurkat T cells activated on coverslip-bound anti-CD3 antibody were visualized using a TCRζ-Halo-JF646 (red) and ZAP70-Emerald (green), b ZAP70-Emerald (green) and LAT-Halo-JF646 (red), c anti-pTCRζ Y142; red) and anti-pLAT (Y226; green), d ZAP70-Emerald (green) and SLP76-Halo-JF646 (red), e GADS-Emerald (green), ZAP70-Apple (red), and GRB2-Halo-JF646 (blue), and f LAT-Emerald (blue), ZAP70-Apple (red), and SLP76-Halo-JF646 (green). a–f Upper right image was magnified from the region marked by a white box in the left image, and the bottom right graph shows relative (rel.) intensity measured across the width of the white line in the corresponding upper right image. a–f Scale bars in left images, 4 μm

Fig. 2

Microcluster components assemble in distinct sequential steps. TIRF images of microclusters formed in Jurkat T cells activated on coverslip-bound anti-CD3 antibody at 21 ^oC were visualized using a CD3ε-YFP (green) and TCRζ-Halo-JF646 (red), b TCRζ-Emerald (green) and ZAP70-Apple (red), c ZAP70-Ruby (red) and GRB2-Emerald (green), d TCRζ-Halo-JF646 (magenta), ZAP70-Apple (red), and GRB2-Emerald (green), or e ZAP70-Apple (red) and c-Cbl-CFP (green). a–e Time-lapse montage at 3 s/frame (middle) and relative intensity plot (bottom) of the boxed region in the top image shows the kinetic lag between the recruitment of microcluster components. The bottom graph shows a plot of the relative intensity per acquired time frame as a colored circle and the best-fit sigmoidal curve as a colored line. Scales bars, 2 μm. f Average kinetic lag observed between the recruitment of respective microcluster components. Kinetic lag was measured by calculating the difference in half-max of best-fit sigmoidal curves. *p < 0.0001. g Kinetic lags measured between TCRζ-Emerald and ZAP70-Apple (red), or ZAP70-Apple and GRB2-Emerald (blue) in Jurkat cells activated on coverslip-bound antibody at indicated temperatures. h–j TIRF images of microclusters formed in primary human CD3⁺ T cells activated on coverslip-bound anti-CD3 and anti-CD28 antibodies at 37 °C were visualized using h TCRζ-Emerald (green) and ZAP70-Scarlet (red), i ZAP70-Emerald (green) and GRB2-Scarlet (red). Time-lapse montage at 3 s/frame (middle) and relative intensity plot with best fit sigmoidal curve (bottom) of the boxed region in the top image. Scale bars, 2 μm. j Average kinetic lags (measured as above) observed between the recruitment of respective microcluster components

Fig. 3

Signaling cluster components are recruited simultaneously to microclusters. TIRF images of microclusters formed in Jurkat T cells activated on coverslip-bound anti-CD3 antibody at 21^oC were visualized using a LAT-Emerald (green) and ZAP70-Apple (red), b SLP76-Emerald (green) and ZAP70-Apple (red), c PLCγ1-CFP (magenta), GADS-YFP (yellow), and GRB2-Apple (cyan), or d SLP76-CFP (magenta), NCK-YFP (yellow), and GRB2-Apple (cyan). a–d Time-lapse montage (middle) and relative intensity plot (bottom) of the boxed region in the top image shows simultaneous recruitment of signaling cluster components. The bottom graph shows a plot of the relative intensity per acquired time frame as a colored circle and the best-fit sigmoidal curve as a colored line. Scales bars, 2 μm. e Average kinetic lag observed between the recruitment of indicated pair of microcluster components at 21 °C. n.s., not significant. *p < 0.0001 compared to ZAP70-GRB2. **p < 0.0001 compared to kinetic lags from GRB2-PLCγ1 to GRB2-ADAP

Fig. 4

Signaling domain disassociation coincides with c-Cbl recruitment and occurs at two distinct rates. a TIRF image of microclusters formed in Jurkat T cells activated on coverslip-bound anti-CD3 antibody at 21 °C were visualized using c-Cbl-YFP and GRB2-Apple. Time-lapse montage (middle) and relative intensity plot (bottom) of the boxed region in the top image shows decrease in GRB2 and c-Cbl signal coincident with peak c-Cbl recruitment (black arrowhead). Scale bar, 2 μm. b–d Representative relative intensity plots of b PLCγ1-CFP (blue), GADS-YFP (green), and GRB2-Apple (red), or c SLP76-CFP (blue), NCK-YFP (green), and GRB2-Apple (red), or d SLP76-CFP (blue), Vav1-YFP (green), and GRB2-Apple (red) at microclusters. The colored arrowheads indicate the final intensity levels after disassociation of each corresponding signaling cluster component

Fig. 5

Signaling cluster recruitment is required for progression to downstream T cell activation events. a Time-lapse montage (left) of a Jurkat T cell becoming activated on coverslip-bound anti-CD3 antibody at 21 °C expressing ZAP70-Apple (yellow), GRB2-Halo-JF646 (magenta), and Fluo-4 (cyan), and the corresponding intensity plot over time (right) shows that downstream calcium flux follows recruitment of signaling cluster as reported by GRB2-Halo-JF646. b Right, kymograph of TCRζ-Halo-JF646 and GRB2-Emerald in an activating Jurkat T cell corresponding to the white line in the left TIRF image shows that cell spreading initiates after recruitment of signaling cluster (white arrowhead in merged image). c Representative TIRF images of TCRζ-Emerald, ADAP-GFP, SLP76-CFP, GADS-YFP, VAV1-YFP, PLCγ1-CFP, NCK-YFP, GRB2-Emerald, c-Cbl-CFP localized in activated E6.1 Jurkat cells (top) or with either ZAP70-Apple or GRB2-Apple in activated J.LAT KO (LAT^−/−) Jurkat cells. Representative relative intensity plots of d TCRζ-Emerald (green) and ZAP70-Apple (red), e ZAP70-Apple (red) and GRB2-Emerald (green), or f c-Cbl-YFP and GRB2-Apple at microclusters and g average kinetic lag measured between indicated pair of microcluster components in J.LAT KO cells. h Activated E6.1 and J.LAT KO Jurkat cells co-expressing TCRζ-Halo-JF646 and mEmerald-WAVE1. i Representative kymographs of TCRζ-Emerald in an activated E6.1 and J.LAT KO Jurkat cell. Scale bars, 2 μm

Fig. 6

Kinetic lags between recruitment of microcluster components are dependent on intracellular calcium flux. a Plot of kinetic lags measured between TCRζ-Emerald and ZAP70-Apple in microclusters formed at indicated image acquisition time in the same Jurkat T cell activated at 21 °C (left). Representative intensity plots of TCRζ-Halo-JF646 (green) and ZAP70-Emerald (red) measured in a microcluster at 75 s and 150 s after T cell activation (right). b Plot of kinetic lags measured between ZAP70-Emerald and GRB2-Halo-JF646 in microclusters formed at indicated image acquisition time in the same Jurkat T cell activated at 21 °C (left). Representative intensity plots of ZAP70-Emerald (red) and GRB2-Halo-JF646 (green) during early (34 s, middle) and late (110 s, right) temporal phases of T cell activation (right). c Representative intensity plots of TCRζ-Emerald (green) and ZAP70-Scarlet (red) (top row), or ZAP70-Scarlet (red) and GRB2-Emerald (green) (bottom row) at a microcluster in a Jurkat cell treated with BAPTA and EGTA (left column), or between TCRζ-Halo-JF646 (green) and ZAP-Scarlet (red) (top row), or ZAP70-Scarlet (red) and GRB2-Halo-JF646 (green) (bottom row) in GCaMP6-expressing cells in an imaging buffer with 2 mM CaCl₂ (middle column), or 5 mM CaCl₂ (right column). d Average kinetic lags measured between TCRζ and ZAP70, or ZAP70 and GRB2 at microclusters under the indicated calcium chelation or elevation conditions. *p < 0.0001 compared to DMSO. n.s., not significant compared to DMSO. Average kinetic lags measured between e TCRζ-Emerald and GRB2-Scarlet, or f ZAP70-Emerald and GRB2-Scarlet at microclusters at indicated extracellular CaCl₂ concentrations