Supplementary Information. Conformational states of Lck regulate clustering in early T cell signaling

Size: px

Start display at page:

Download "Supplementary Information. Conformational states of Lck regulate clustering in early T cell signaling"

Chester Hardy
5 years ago
Views:

1 Supplementary Information Conformational states of Lck regulate clustering in early T cell signaling Jérémie Rossy, Dylan M. Owen, David J. Williamson, Zhengmin Yang and Katharina Gaus Centre for Vascular Research and Australian Centre for Nanomedicine, University of New South Wales, Sydney, Australia Correspondence should be addressed to KG (k.gaus@unsw.edu.au)

2 Supplementary Fig. 1. Quantitative cluster analysis of PALM and dstorm data. (a-b) TIRF (a) and corresponding PALM (b) images of Lck-PS-CFP2 expressed in a JCaM1 cell on activating anti-cd3+anti-cd28 antibody-coated glass coverslips for 10 min. Bar = 5 µm. (c) Histogram of the localisation precision for the image shown in b, with a mean 21 nm. Dotted line indicates threshold for precision (<50 nm). (d-e) Analysis of repeated excitation from the same molecule. Number of molecular localizations (symbols) of Lck-PS-CFP2 expressed in JCaM1 cells (d) and wild-type Lck expressed in JCaM1 cells and stained with primary antibodies and secondary DyLight 649 F(ab')2 fragments (e) plotted against the off-gap. Data was fitted to Equation 1 (solid lines). PS-CFP2 exhibited 0.1 blinks per molecule for 0.5 frames (20 ms) and 1.2 blinks per molecule for 13.7 frames (548 ms). DyLight649 blinked an average of 0.7 times for an average dark-time of 0.5 frames (15 ms) and 1.4 blinks per molecule for 5 frames (150 ms). Based on this analysis, we applied an off-gap of 10 frames for PS-CFP2 and 30 frames for DyLight649, which accounted for over 97% of the molecular over-estimation caused by multiple blinking, to all our data. Hence, clustering is not caused by repeated excitation of the same molecule. Data are from cells obtained from three independent experiments.

3 Supplementary Fig. 2. Comparison of TIRF and PALM images (a-c), activation of JCaM1 cells (d) and clusters of phosphorylated Lck (e-f). (a-c) Lck microclusters (red squares) are visible in the TIRF image (a), that are composed of a higher density of Lck molecules (PALM image, b) that are detected as several discreet nanoclusters based on Getis and Franklin s local point pattern analysis (color-coded cluster map, c). Scale bar = 2 µm. (d) Maxima of Ripley s K-function curves of Lck clustering in JCaM1 cells on glass coverslips coated with anti-cd3 alone or anti-cd3+anti-cd28 antibodies. Points represent one image region. NS; not significant (Student s t-test). Data are from 8-12 cells from four independent experiments. (e) Anti-p-Y394 Lck antibodies recognize Lck in the open and closed conformation. dstorm image of ptyr394-src family kinases (middle) and color-coded cluster maps (right) of highlighted region in JCaM1 cells expressing wild-type Lck on anti-cd3+anticd28-coated glass coverslips. (e-f) Bar = 5 µm (middle); bar = 500 nm (right). (f) Anti-p-Y505 Lck antibodies recognize Lck in the open and closed conformation. dstorm image of ptyr505-lck (middle) and cluster map (right) of indicated region in a JCaM1 cell expressing wild-type Lck on anti-cd3+anti-cd28 antibody-coated glass coverslips. Secondary stain of antibodies was DyLight 649 F(ab')2 fragments.

4 Supplementary Fig. 3. Distribution analysis of clusters. (a-f) Ripley s K-function analysis of the center-of-mass coordinates of wild-type Lck clusters in resting JCaM1 cells (a), wild-type Lck clusters in the periphery (b) and center of the activation zone in activated JCaM1 cells (c), phosphorylated TCRζ (p-tcrζ) in the periphery (d) and center (e) of the activation zone, and CD45 clusters in activated cells (f) obtained from the thresholded cluster maps. If clusters themselves are randomly distributed, the L(r)-r curves (red lines) fall within the 99% confidence intervals (black lines) that were calculated from 100 spatially random simulated distributions. Only Lck clusters in the center of the activation zone on large radial scales and clusters of phosphorylated TCR in the center of the activation zone were clustered themselves.

5 Supplementary Fig. 4. Lck10 clustering in resting and activated Jurkat E6.1 cells. (a) Ripley s K-functions of Lck10 expressed in Jurkat cells on PLL-coated glass coverslips (Rest) or on activating anti-cd3+anti-cd28 antibody-coated glass coverslips (Act) plotted against radius, r, of concentric circles centered on each molecule. 99% CI represents 99% confidence intervals based on simulated data. (b) Maxima of Ripley s K-function curves, as plotted in a. (c) Cluster radius in nm, obtained from thresholded cluster maps. (d) Circularity of clusters obtained from thresholded cluster maps. Circularity of clusters was calculated as 4π(area/perimeter 2 ). Lck10 displayed a different clustering behavior in WT Jurkat E6.1 cells compared to JCaM1 cells that lacked endogenous expression of Lck. In JCaM1 cells, TCR activation reorganized Lck from a near random distribution to a low level of clustering, whereas in Jurkat E6.1 cells, Lck10 distribution remained near random. We propose that in Jurkat E6.1 cells, endogenous Lck competes with Lck10 for microdomains residency, preventing Lck10 to be re-organized upon TCR activation. In b-d, points represent one image region, horizontal bars and error bars show mean and SEM, respectively. * P < (Student s t-test). Data are from cells from three independent experiments.

6 Supplementary Fig 5. Inhibition of membrane condensation by incorporation of the oxysterol 7-ketocholesterol (7KC) does not affect Lck clustering. To inhibit the membrane condensation upon TCR activation, JCaM1 cells were activated on anti- CD3+anti-CD28 antibody-coated glass coverslips for 10 min and treated with 5 µm 7KC for the last 1.5 min of the activation period before fixation. (a) Ripley s K-function curves of control (Ctrl) and 7KC-treated cells calculated for individual image regions and plotted against radius, r, of concentric circles centered on each molecule. 99% CI represents 99% confidence intervals based on simulated data. (b) Maxima of Ripley s K-function curves, as plotted in a. (c-d) Percentages of Lck molecules in clusters (c), ratio of the molecular densities inside clusters to molecular densities outside clusters (d), clusters radii in nm (e), and cluster numbers per µm 2 (f) obtained from thresholded cluster maps. Points represent one image region, horizontal bars and error bars show mean and SEM, respectively. All data are not statistically significant (P > 0.05, Student s t-test). Data are from 13 cells from three independent experiments.

7 Supplementary Fig. 6. Kinase activity does not influence Lck clustering. (a) Ripley s K-functions of wild-type Lck and the kinase dead Lck mutant, R273A Lck, expressed in JCaM1 cells on anti-cd3+anti-cd28 antibody-coated glass coverslips plotted against radius, r, of concentric circles centered on each molecule. 99% CI represents 99% confidence intervals based on simulated data. (b) Maxima of Ripley s K-function curves, as plotted in a. (c-d) Percentages of Lck molecules in clusters (c), ratio of the molecular densities inside clusters to molecular densities outside clusters (d), clusters radii in nm (e), and cluster numbers per µm 2 (f) obtained from thresholded cluster maps. Points represent one image region, horizontal bars and error bars show mean and SEM, respectively. Wild-type Lck data are identical to Fig. 1. No significant differences were found (P > 0.05, one-way ANOVA test). Data are from 14 cells from three independent experiments.

8 Supplementary Fig. 7. TCR activation has no impact on CD45 clustering. (a-c) JCaM1 cells expressing wild-type Lck on non-activating PLL surfaces (Rest) or on anti- CD3+anti-CD28 antibody-coated glass coverslips (Act) were stained with anti-cd45 antibodies. CD45 localizations were imaged with dstorm and analyzed as described in Online Methods. Percentage of CD45 molecules in clusters (a), ratios of the molecular densities inside clusters to molecular densities outside clusters (b), and clusters radii in nm (c) obtained from thresholded cluster maps. Points represent one image region; horizontal bars and error bars show mean and SEM, respectively. No significant differences were found (P > 0.05; Student s t-test). Data are from 10 cells from two independent experiments.

Supplementary Fig. 8. An Lck-centric model of early T cell signaling links intra-molecular rearrangements to surface patterning of signaling molecules.

The conformational flexibility of Lck continuously redistributes the kinase in and out of clusters and some Lck cluster co-localize with clusters of the phosphatase CD45.

9 Supplementary Fig. 8. An Lck-centric model of early T cell signaling links intra-molecular rearrangements to surface patterning of signaling molecules. (a) In resting T cells, Lck is present in the active, open and inactive, closed conformations and selfassociates to form nano-scaled clusters containing 26 ± 7% of Lck molecules. The conformational flexibility of Lck continuously redistributes the kinase in and out of clusters and some Lck cluster co-localize with clusters of the phosphatase CD45. (b) After TCR triggering, segregation of CD45 clusters from Lck clusters increases the proportion of Lck in open conformation in and around the pre-existing Lck nano-clusters. This local increase in open conformation leads to the expansion and densification of Lck clusters to a radius of 51 ± 5 nm incorporating 43 ± 8% of Lck molecules (c) From the nano-scaled re-arrangement upon TCR triggering, a micro-scaled organization emerges in activated T cells, where the phosphorylated receptor and Lck co-localize in common domains to promote phosphorylation of the ITAM domains in the TCRζ chains. The reduced clustering dynamics of Lck in the open conformation may further stabilize TCR microclusters for sustained signaling.

10 Supplementary Movie 1. High clustering dynamics of wild-type Lck during TCR activation. Live cell cluster maps of 1,500-frame windows (equivalent to 45 s) shifted by 250 frames (7.5 s) of a JCaM1 cell expressing wild-type Lck-mEos2 on an anti-cd3+anti-cd28 antibody-coated glass coverslip imaged for 15 min at 37 C. Color-coded cluster maps were generated from local point pattern analysis and color-coded as indicated in Fig. 5. Image size is 4 µm x 4 µm. Supplementary Movie 2. Reorganization of a wild-type Lck cluster during TCR activation. Live cell cluster map movie (Supplementary Movie 1) of a JCaM1 cell expressing wild-type LckmEos2 on an anti-cd3+anti-cd28 antibody-coated glass coverslip was thresholded and cropped to generate a time series of a hot spot in which an Lck cluster formed, moved laterally, fragmented and reformed. Points are individual wild-type Lck molecules associated with this cluster. Each frame in the movie corresponds to 45s with 7.5s time intervals between frames. Image size is 750 nm x 750 nm. Supplementary Movie 3. Clustering dynamics of Y505F Lck during TCR activation. Live cell cluster maps of 1,500-frame windows (equivalent to 45 s) shifted by 250 frames (7.5 s) of a JCaM1 cell expressing Y505F Lck-mEos2 on an anti-cd3+anti-cd28 antibody-coated glass coverslip imaged for 15 min at 37 C. Color-coded cluster maps were generated from local point pattern analysis and color-coded as indicated in Fig. 5. Image size is 4 µm x 4 µm. Supplementary Movie 4. Stability of a cluster of Y505F Lck during TCR activation. Live cell cluster map movie (Supplementary Movie 3) of a JCaM1 cell expressing Y505F LckmEos2 on an anti-cd3+anti-cd28 antibody-coated glass coverslip was thresholded and cropped to generate a time series of a single Lck cluster those morphology changed over time but the cluster itself was stable. Points are individual Y505F Lck molecules associated with this cluster. Each frame in the movie corresponds to 45s with 7.5s time intervals between frames. Image size is 750 nm x 750 nm.

Nature Immunology: doi: /ni.3631

Nature Immunology: doi: /ni.3631 Supplementary Figure 1 SPT analyses of Zap70 at the T cell plasma membrane. (a) Total internal reflection fluorescent (TIRF) excitation at 64-68 degrees limits single molecule detection to 100-150 nm above