Cells receive dozens of signals at different times and in different places. Integrating incoming information and deciding how to respond is no easy task. Signaling molecules on the cell surface pass messages inwards using chemical messengers that interact in complicated networks within the cell. One way to unravel the complexity of these networks is to look at specific groups of signaling molecules in test tubes to see how they interact. But the interior of a living cell is a very different environment. Molecules inside cells are tightly packed and, under certain conditions, they interact with each other by the thousands. They form structures known as ‘supramolecular complexes’, which changes their behavior.

One such supramolecular complex is the ‘central supramolecular activation cluster’, or cSMAC for short. It forms under the surface of immune cells called T cells when they are getting ready to fight an infection. Under the microscope, the cSMAC looks like the bullseye of a dartboard, forming a crowd of signaling molecules at the center of the interface between the T cell and another cell. Its exact role is not clear, but evidence suggests it helps to start and stop the signals that switch T cells on. The cSMAC contains two key protein adaptors called LAT and SLP-76 that help to hold the structure together. So, to find out what the cSMAC does, Clark et al. genetically modified these adaptors to gain control over when the cSMAC forms.

Clark et al. examined mouse T cells using super-resolution microscopy and electron microscopy, watching as other immune cells delivered the signal to switch on. As the T cells started to activate, the composition of the cSMAC changed. In the first two minutes after the cells started activating, the cSMAC included a large number of different components. This made T cell activation more efficient, possibly because the supramolecular complex was helping the network of signals to interact. Later, the cSMAC started to lose many of these components. Separating components may have helped to stop the activation signals.

Understanding how T cells activate could lead to the possibility of turning them on or off in immune-related diseases. But these findings are not just relevant to immune cells. Other cells also use supramolecular complexes to control their signaling. Investigating how these complexes change over time could help us to understand how other cell types make decisions.

T cell activation is governed by spatiotemporal organization of signal transduction across scales. At the nanoscale, receptors form clusters of dozens of molecules that can coalesce into microclusters and are commonly associated with active forms of signaling intermediates (Boyle et al., 2011; Hu et al., 2016; Lillemeier et al., 2006; Schamel et al., 2005; Sherman et al., 2011; Varma et al., 2006; Yokosuka et al., 2005). This association suggests that such receptor clusters mediate efficient T cell signaling. Larger, µm scale assemblies were first described at the center and periphery of T cells activated by antigen-presenting cells (APC) for the TCR, PKCθ and LFA-1, talin, respectively, as central and peripheral supramolecular activation clusters (cSMAC and pSMAC) (Grakoui et al., 1999; Monks et al., 1998; Monks et al., 1997). µm scale of assemblies, in particular in the form of supramolecular protein complexes, provides unique biophysical and signaling properties (Banani et al., 2016; Li et al., 2012; Shin and Brangwynne, 2017). Supramolecular protein complexes play critical roles in viral sensing (Cai et al., 2014), inflammation (Franklin et al., 2014), embryonic development (Brangwynne et al., 2009), protein folding in cancer (Rodina et al., 2016), nuclear ubiquitinoylation (Marzahn et al., 2016), and chromatin compaction (Larson et al., 2017). Such complexes are readily observed by fluorescence microscopy, held together by a network of multivalent protein interactions and often have distinct phase properties (Banani et al., 2016; Li et al., 2012; Shin and Brangwynne, 2017). The cSMAC has many properties of such supramolecular protein complexes: It contains various multivalent signaling intermediates (Balagopalan et al., 2015), prominently LAT (linker of activation of T cells), components of this complex including LAT and PKCθ exchange with the remainder of the cell to a moderate extent and slowly (Roybal et al., 2015), and components of this complex can be assembled into supramolecular structures in vitro (Su et al., 2016). Therefore, understanding biophysical properties of the cSMAC and how it regulates T cell activation is of substantial importance. cSMAC function is controversial despite decades of work. Limitations in investigating the cSMAC are that its properties are largely unresolved and that its composition and/or assembly have not been systematically manipulated inside live T cells. Based on association of cSMAC formation with T cell activation conditions, the cSMAC has been proposed to enhance T cell signaling, terminate it, not be related to signaling or only upon weak stimulation or at late time points (Čemerski et al., 2008; Freiberg et al., 2002; Grakoui et al., 1999; Lee et al., 2002; Monks et al., 1997). cSMAC formation is often associated with efficient T cell activation conditions, fitting with a role in enhancing T cell signaling. Accumulation of signaling intermediates at the T cell:APC interface center is substantially reduced by blockade of the costimulatory receptor CD28 (Singleton et al., 2009; Wülfing et al., 2002), in regulatory T cells (Zanin-Zhorov et al., 2010), during thymic selection (Ebert et al., 2008), or in the absence of the signal amplifying kinase Itk (IL-2 inducible T cell kinase) (Singleton et al., 2011). To determine cSMAC properties, we have used stimulated emission depletion (STED) super-resolution microscopy and correlative light electron microscopy (CLEM). The cSMAC was composed of multiple complexes of supramolecular dimensions and associated with extensive membrane undulations. To investigate cSMAC function, we have systematically manipulated the localization of three adaptor proteins in live primary T cells: LAT (Balagopalan et al., 2015) is an integral component of the cSMAC. SLP-76 (SH2 domain-containing leucocyte protein of 76 kD) (Koretzky et al., 2006) is associated with it only during the first minute of T cell activation (Roybal et al., 2015). Grb2 (growth factor receptor-bound 2) (Jang et al., 2009) association with the cSMAC is less prevalent (Roybal et al., 2015). Interface recruitment of all three adaptors was diminished upon attenuation of T cell activation by costimulation blockade and Itk-deficiency as was IL-2 secretion, a critical T cell effector function. By fusing these adaptors with various protein domains with a strong interface localization preference we brought them back to the interface under the attenuated T cell activation conditions and restored cSMAC formation. Such restoration enhanced IL-2 secretion but only when executed to the extent and with dynamics seen under full stimulus conditions.

The cSMAC was characterized by enhanced membrane undulations and the formation of multiple protein complexes of supramolecular dimensions in the 0.1–0.5 µm³ range. Supramolecular complexes can be driven by the polymerization of a single or few defined proteins (Cai et al., 2014; Franklin et al., 2014; Li et al., 2012; Marzahn et al., 2016) or, likely more common, they can consists of an agglomeration of large numbers of proteins (Rodina et al., 2016; Tarantino et al., 2014). While supramolecular complexes formed by a small number of components are often characterized by defined structures such as lipid droplets or fibers (Shin and Brangwynne, 2017), our understanding of supramolecular complexes built from a large number of components is limited. As close to half of the amount of three components of the central signaling complex, LAT, active Rac, and PKCθ, is immobile and the remainder exchanges only slowly with rest of the cell (Roybal et al., 2015) cSMAC signaling complexes likely have partial solid state properties. T cell signaling has been extensively characterized with super-molecular resolution in T cells activated with planar APC substitutes, supported lipid bilayers and antibody-coated cover slips. Similar to our findings large protein complexes, microclusters, were found. The microclusters are transported to the interface center to form a single µm scale structure (Varma et al., 2006; Yokosuka et al., 2005). However in contrast to T cell activation on planar surfaces, in T cell:APC couples membrane undulations form across the entire interface with F-actin structures perpendicular to the interface plane (Roybal et al., 2015) and thus impair rather than enhance centripetal transport. A large central invagination may remove proteins from the interface center (Singleton et al., 2006) and the central membrane undulations characterized here increase the number of transmembrane proteins in contact with a given cytoplasmic volume. Therefore, the cSMAC in T cell:APC couples will likely have similar molecular constituents as the central complex in T cells activated on planar substrates but distinct biophysical properties. It is the signaling functionality arising from these unique properties that we have investigated here. Nevertheless, the more straightforward access to high resolution imaging and synthetic system engineering using planar APC substitutes provides intriguing data that complement our findings. Supramolecular signaling complexes including LAT, SLP-76 and Grb2 have been reconstituted on model membrane and shown to trigger signaling and F-actin assembly (Su et al., 2016). Large numbers of exosomes at the center of the interface between T cells and supported lipid bilayers could be related to the cSMAC membrane undulations and contribute to signal attenuation (Choudhuri et al., 2014). Highly localized F-actin structures discovered using T cell activation on planar surfaces could mediate transport into supramolecular signaling complexes (Kumari et al., 2015) with the F-actin uncapping protein RLTPR identified as a regulator of the composition of signaling complexes (Liang et al., 2013).

Using a large-scale live cell imaging approach, we were able to distinguish between inefficient cSMAC formation and inefficient recruitment of a signaling intermediate to an existing cSMAC: As central interface accumulation of numerous signaling intermediates was diminished upon costimulation blockade and Itk-deficiency (Figure 1C,D) cSMAC formation was most likely impaired. Central interface accumulation of spatially targeted LAT and SLP-76 upon the attenuated T cell stimuli thus has to represent cSMAC restoration as confirmed by CLEM (Figure 3D).

Complexes of supramolecular dimensions are part of the cSMAC (Figure 3B). In general, supramolecular complex formation becomes more likely with increasing concentrations and valences of the complex components (Banani et al., 2016; Li et al., 2012). Accordingly, replacement of proline-rich regions in Sos that mediate multivalent LAT/Grb2/Sos interactions leads to reduced LAT clustering and phosphorylation (Kortum et al., 2013). Similarly, when LAT phosphorylation (Figure 2C), which is required for interactions of LAT with SH2 domain-containing signaling intermediates, was reduced upon costimulation blockade and Itk deficiency LAT clustering at the interface center was also diminished (Figure 2B). Compensation for such diminished LAT valence by fusing LAT with additional protein interaction domains restored µm scale central LAT accumulation. The dependence of LAT clustering on the number of functional protein interaction motifs, that is its valence, supports the supramolecular nature of protein complexes within the cSMAC. The importance of valence is further illustrated in the divergent behavior of fusion proteins versus their components. For example, the PKCθ V3 domain in isolation does not localize to the interface center even under full stimulus conditions (Figure 4—figure supplement 1A) nor does LAT upon costimulation blockade (Figure 2B). However, the LAT-V3 fusion protein shows dominant central localization upon costimulation blockade (Figure 4B). How is this possible? By generating a LAT-V3 fusion proteins we didn’t only add the localization preferences of LAT and the PKCθ V3 domain, we also increased the valence of the fusion protein over its components as a key facilitator of recruitment to supramolecular complexes (Li et al., 2012).

Interestingly, fusion domains could not drive spatial features of adaptor localization by themselves but could only enhance intrinsic adaptor localization preferences. We could enhance central LAT clustering across all time points, central SLP-76 clustering only during the first minute of cell coupling and central Grb2 clustering not at all. This mirrored the accumulation patterns of the non-targeted adaptors under full stimulus conditions (Figures 4, 7 and 8) and therefore strongly suggests that localization of the spatially targeted adaptors was driven by a combination of intrinsic localization motifs and the spatial information provided by the fused domains. In support, while both PKCθ V3 and PLCδ PH did not display central localization in isolation, they could enhance central localization of LAT and SLP-76 (Figures 4 and 7). Similarly, artificially enhanced cSMAC formation through expression of LAT V3 did not lead to the recruitment of signaling intermediates with intrinsically weak cSMAC preference such as Vav1 (Figure 4C). Our inability to force cSMAC localization implies that even upon addition of new protein interaction domains the overall molecular composition of supramolecular complexes in the cSMAC is fairly well conserved: A core of protein-protein interactions may be required for complex formation such that addition of new protein interactions can only enhance complex stability but not generate complexes of fundamentally different composition. Supramolecular assemblies in the cSMAC thus likely exist in a delicate balance between compositional conservation for complex identity and some redundancy in protein interaction motifs and stoichiometry to allow flexible regulation of stability. As a possible exception to this rule it was only upon the strongest attenuation of T cell activation, Itk-deficiency combined with costimulation blockade, that Grb2 as one of three signaling intermediates investigated was recruited to a LAT V3-based cMSAC to an extent not seem for Grb2 interface accumulation under other T cell activation conditions (Figure 6). We suggest the following scenario. Grb2 doesn’t effectively compete for binding to the cSMAC under strong and even moderate T cell activation conditions. Likely, stronger binding signaling intermediates are activated to a sufficient extent under those conditions to outcompete Grb2. However, with the strong signaling attenuation provided by the combination of costimulation blockade and Itk-deficiency activation of more competitive signaling intermediates becomes sufficiently inefficient to allow Grb2 recruitment to the LAT V3-based signaling complexes.

Reconstitution of central LAT and SLP-76 clustering under attenuated T cell activation conditions could restore IL-2 mRNA generation. The association of lack of Grb2 centrality with lack of an effect on IL-2 mRNA generation provides a specificity control. Protein clustering in the cSMAC thus was a critical component of efficient T cell signaling. For the cSMAC to enhance T cell function experimental reconstitution of cSMAC formation needed to closely mimic cSMAC features observed with non-targeted constructs under full stimulus conditions. Time-dependent roles of the cSMAC have been proposed before (Freiberg et al., 2002) yet are difficult to compare to the work here as the molecules investigated differed. In our work under full stimulus conditions LAT-GFP and SLP-76-GFP were efficiently recruited to the interface center within the first two minutes of T cell activation with diminished recruitment thereafter. Recruitment of LAT and SLP-76 to the interface center upon attenuation of T cell activation by fusion with the Vav1 SH3SH2SH3 and PLCδ PH domains closely reproduced these dynamics (Figures 4B and 7) and largely restored IL-2 mRNA generation (Figures 5A and 7). However, recruitment of LAT to the interface center to a greater extent and duration by fusion with PKCθ V3 (Figure 4B) could not enhance IL-2 mRNA generation (Figure 5A). We therefore suggest that the cSMAC displays two different time-dependent roles. Within the first one to two minutes of T cell activation it efficiently brings together the large number of proximal T cell signaling intermediates required for efficient T cell activation. Subsequently, a substantial number of key signaling intermediates including SLP-76, Itk, PLCγ, and Vav1 leave the cSMAC and move to smaller signaling complexes supported by an interface-wide lamellal actin network (Roybal et al., 2015). Retention of a more limited subset of signaling intermediates in the cSMAC after this time thus may render them less accessible to their interaction partners and therefore diminish sustained signal transduction. Signal enhancing and attenuating roles of the cSMAC thus may both occur as regulated by the specific time-dependent composition of the complex.

Itk and costimulation likely contribute to cSMAC formation by overlapping yet partially distinct means. Recruitment to the interface center peaks within the first two minutes of cell coupling for both ligand-engaged CD28 (Purtic et al., 2005; Singleton et al., 2009) and full length Itk (Roybal et al., 2015). Both thus can be expected to provide protein-protein interactions during the early signal amplifying stage of the cSMAC. Itk also has enzymatic activity to directly modify cSMAC components. Both costimulation and Itk regulate actin dynamics. However, while costimulation controls core actin turnover through the Arp2/3 complex and Cofilin (Roybal et al., 2016), Itk only regulates a SLAT-dependent subset of actin dynamics (Singleton et al., 2011). CD28 thus can be expected to contribute more strongly to cSMAC directed transport in complex assembly (Roybal et al., 2016)(Figure 2C).

In summary, our work establishes that the cSMAC formed in the activation of T cells by APCs consists of multiple supramolecular complexes driven by extensive membrane undulations with a dynamically changing composition. Compositionally rich complexes in the first two minutes of cell coupling enhanced T cell activation by facilitating efficient signaling interactions whereas thereafter compositionally poorer complexes may sequester signaling intermediates.

All imaging data are openly accessible via figshare (http://doi.org/10.1184/R1/9963566) and LAT phosphorylation data that support the findings of this study are available at the University of Bristol data repository (https://doi.org/10.5523/bris.2uoex1k196c4o2c80eddeekf04).

1. 1. Danielle J Clark
   2. Laura E McMillan
   3. Sin Lih Tan
   4. Gaia Bellomo
   5. Clémentine Massoué
   6. Harry Thompson
   7. Lidiya Mykhaylechko
   8. Dominic Alibhai
   9. Xiongtao Ruan
   10. Kentner L Singleton
   11. Minna Du
   12. Alan J Hedges
   13. Pamela L Schwartzberg
   14. Paul Verkade
   15. Robert F Murphy
   16. Christoph Wülfing

   (2019) University of Bristol data repository

   Data in support of Clark et al.

   https://doi.org/10.5523/bris.2uoex1k196c4o2c80eddeekf04
2. 1. Danielle J Clark
   2. Laura E McMillan
   3. Sin Lih Tan
   4. Gaia Bellomo
   5. Clementine Massoue
   6. Harry Thompson
   7. Lidiya Mykhaylechko
   8. Dominic Alibhai
   9. Xiongtao Ruan
   10. Kentner L Singleton
   11. Minna Du
   12. Alan Hedges
   13. Pamela L Schwartzberg
   14. Paul Verkade
   15. Robert F Murphy
   16. Christoph Wülfing

   (2019) figshare

   Image data from Transient protein accumulation at thecenter of the T cell antigen-presenting cellinterface drives efficient IL-2 secretion.

   https://doi.org/10.1184/R1/9963566

1. 1. Ambler R
   2. Ruan X
   3. Murphy RF
   4. Wülfing C

   (2017) Systems imaging of the immune synapse

   Methods in Molecular Biology 1584:409–421.

   https://doi.org/10.1007/978-1-4939-6881-7_25

   • PubMed
   • Google Scholar
2. 1. Balagopalan L
   2. Kortum RL
   3. Coussens NP
   4. Barr VA
   5. Samelson LE

   (2015) The Linker for activation of T cells (LAT) signaling hub: from signaling complexes to microclusters

   Journal of Biological Chemistry 290:26422–26429.

   https://doi.org/10.1074/jbc.R115.665869

   • PubMed
   • Google Scholar
3. 1. Banani SF
   2. Rice AM
   3. Peeples WB
   4. Lin Y
   5. Jain S
   6. Parker R
   7. Rosen MK

   (2016) Compositional control of Phase-Separated cellular bodies

   Cell 166:651–663.

   https://doi.org/10.1016/j.cell.2016.06.010

   • PubMed
   • Google Scholar
4. 1. Boyle S
   2. Kolin DL
   3. Bieler JG
   4. Schneck JP
   5. Wiseman PW
   6. Edidin M

   (2011) Quantum dot fluorescence characterizes the nanoscale organization of T cell receptors for antigen

   Biophysical Journal 101:L57–L59.

   https://doi.org/10.1016/j.bpj.2011.10.028

   • PubMed
   • Google Scholar
5. 1. Brangwynne CP
   2. Eckmann CR
   3. Courson DS
   4. Rybarska A
   5. Hoege C
   6. Gharakhani J
   7. Jülicher F
   8. Hyman AA

   (2009) Germline P granules are liquid droplets that localize by controlled dissolution/condensation

   Science 324:1729–1732.

   https://doi.org/10.1126/science.1172046

   • PubMed
   • Google Scholar
6. 1. Cai X
   2. Chen J
   3. Xu H
   4. Liu S
   5. Jiang QX
   6. Halfmann R
   7. Chen ZJ

   (2014) Prion-like polymerization underlies signal transduction in antiviral immune defense and inflammasome activation

   Cell 156:1207–1222.

   https://doi.org/10.1016/j.cell.2014.01.063

   • PubMed
   • Google Scholar
7. 1. Čemerski S
   2. Das J
   3. Giurisato E
   4. Markiewicz MA
   5. Allen PM
   6. Chakraborty AK
   7. Shaw AS

   (2008) The balance between T cell receptor signaling and degradation at the center of the immunological synapse is determined by antigen quality

   Immunity 29:414–422.

   https://doi.org/10.1016/j.immuni.2008.06.014

   • PubMed
   • Google Scholar
8. 1. Choudhuri K
   2. Llodrá J
   3. Roth EW
   4. Tsai J
   5. Gordo S
   6. Wucherpfennig KW
   7. Kam LC
   8. Stokes DL
   9. Dustin ML

   (2014) Polarized release of T-cell-receptor-enriched microvesicles at the immunological synapse

   Nature 507:118–123.

   https://doi.org/10.1038/nature12951

   • PubMed
   • Google Scholar
9. 1. Ebert PJ
   2. Ehrlich LI
   3. Davis MM

   (2008) Low ligand requirement for deletion and lack of synapses in positive selection enforce the gauntlet of thymic T cell maturation

   Immunity 29:734–745.

   https://doi.org/10.1016/j.immuni.2008.09.014

   • PubMed
   • Google Scholar
10. 1. Franklin BS
    2. Bossaller L
    3. De Nardo D
    4. Ratter JM
    5. Stutz A
    6. Engels G
    7. Brenker C
    8. Nordhoff M
    9. Mirandola SR
    10. Al-Amoudi A
    11. Mangan MS
    12. Zimmer S
    13. Monks BG
    14. Fricke M
    15. Schmidt RE
    16. Espevik T
    17. Jones B
    18. Jarnicki AG
    19. Hansbro PM
    20. Busto P
    21. Marshak-Rothstein A
    22. Hornemann S
    23. Aguzzi A
    24. Kastenmüller W
    25. Latz E

    (2014) The adaptor ASC has extracellular and 'prionoid' activities that propagate inflammation

    Nature Immunology 15:727–737.

    https://doi.org/10.1038/ni.2913

    • Google Scholar
11. 1. Freiberg BA
    2. Kupfer H
    3. Maslanik W
    4. Delli J
    5. Kappler J
    6. Zaller DM
    7. Kupfer A

    (2002) Staging and resetting T cell activation in SMACs

    Nature Immunology 3:911–917.

    https://doi.org/10.1038/ni836

    • PubMed
    • Google Scholar
12. 1. Grakoui A
    2. Bromley SK
    3. Sumen C
    4. Davis MM
    5. Shaw AS
    6. Allen PM
    7. Dustin ML

    (1999) The immunological synapse: a molecular machine controlling T cell activation

    Science 285:221–227.

    https://doi.org/10.1126/science.285.5425.221

    • PubMed
    • Google Scholar
13. 1. Hu YS
    2. Cang H
    3. Lillemeier BF

    (2016) Superresolution imaging reveals nanometer- and micrometer-scale spatial distributions of T-cell receptors in lymph nodes

    PNAS 113:7201–7206.

    https://doi.org/10.1073/pnas.1512331113

    • PubMed
    • Google Scholar
14. 1. Iezzi G
    2. Karjalainen K
    3. Lanzavecchia A

    (1998) The duration of antigenic stimulation determines the fate of naive and effector T cells

    Immunity 8:89–95.

    https://doi.org/10.1016/S1074-7613(00)80461-6

    • PubMed
    • Google Scholar
15. 1. Jang IK
    2. Zhang J
    3. Gu H

    (2009) Grb2, a simple adapter with complex roles in lymphocyte development, function, and signaling

    Immunological Reviews 232:150–159.

    https://doi.org/10.1111/j.1600-065X.2009.00842.x

    • PubMed
    • Google Scholar
16. 1. Jenkinson SR
    2. Williams NA
    3. Morgan DJ

    (2005) The role of intercellular adhesion molecule-1/LFA-1 interactions in the generation of tumor-specific CD8+ T cell responses

    The Journal of Immunology 174:3401–3407.

    https://doi.org/10.4049/jimmunol.174.6.3401

    • PubMed
    • Google Scholar
17. 1. Kong KF
    2. Yokosuka T
    3. Canonigo-Balancio AJ
    4. Isakov N
    5. Saito T
    6. Altman A

    (2011) A motif in the V3 domain of the kinase PKC-θ determines its localization in the immunological synapse and functions in T cells via association with CD28

    Nature Immunology 12:1105–1112.

    https://doi.org/10.1038/ni.2120

    • PubMed
    • Google Scholar
18. 1. Koretzky GA
    2. Abtahian F
    3. Silverman MA

    (2006) SLP76 and SLP65: complex regulation of signalling in lymphocytes and beyond

    Nature Reviews Immunology 6:67–78.

    https://doi.org/10.1038/nri1750

    • Google Scholar
19. 1. Kortum RL
    2. Balagopalan L
    3. Alexander CP
    4. Garcia J
    5. Pinski JM
    6. Merrill RK
    7. Nguyen PH
    8. Li W
    9. Agarwal I
    10. Akpan IO
    11. Sommers CL
    12. Samelson LE

    (2013) The ability of Sos1 to oligomerize the adaptor protein LAT is separable from its guanine nucleotide exchange activity in vivo

    Science Signaling 6:ra99.

    https://doi.org/10.1126/scisignal.2004494

    • PubMed
    • Google Scholar
20. 1. Kumari S
    2. Depoil D
    3. Martinelli R
    4. Judokusumo E
    5. Carmona G
    6. Gertler FB
    7. Kam LC
    8. Carman CV
    9. Burkhardt JK
    10. Irvine DJ
    11. Dustin ML

    (2015) Actin foci facilitate activation of the phospholipase C-γ in primary T lymphocytes via the WASP pathway

    eLife 4:e04953.

    https://doi.org/10.7554/eLife.04953

    • Google Scholar
21. 1. Larson AG
    2. Elnatan D
    3. Keenen MM
    4. Trnka MJ
    5. Johnston JB
    6. Burlingame AL
    7. Agard DA
    8. Redding S
    9. Narlikar GJ

    (2017) Liquid droplet formation by HP1α suggests a role for phase separation in heterochromatin

    Nature 547:236–240.

    https://doi.org/10.1038/nature22822

    • PubMed
    • Google Scholar
22. 1. Lee KH
    2. Holdorf AD
    3. Dustin ML
    4. Chan AC
    5. Allen PM
    6. Shaw AS

    (2002) T cell receptor signaling precedes immunological synapse formation

    Science 295:1539–1542.

    https://doi.org/10.1126/science.1067710

    • PubMed
    • Google Scholar
23. 1. Li P
    2. Banjade S
    3. Cheng HC
    4. Kim S
    5. Chen B
    6. Guo L
    7. Llaguno M
    8. Hollingsworth JV
    9. King DS
    10. Banani SF
    11. Russo PS
    12. Jiang QX
    13. Nixon BT
    14. Rosen MK

    (2012) Phase transitions in the assembly of multivalent signalling proteins

    Nature 483:336–340.

    https://doi.org/10.1038/nature10879

    • PubMed
    • Google Scholar
24. 1. Liang Y
    2. Cucchetti M
    3. Roncagalli R
    4. Yokosuka T
    5. Malzac A
    6. Bertosio E
    7. Imbert J
    8. Nijman IJ
    9. Suchanek M
    10. Saito T
    11. Wülfing C
    12. Malissen B
    13. Malissen M

    (2013) The lymphoid lineage-specific actin-uncapping protein rltpr is essential for costimulation via CD28 and the development of regulatory T cells

    Nature Immunology 14:858–866.

    https://doi.org/10.1038/ni.2634

    • PubMed
    • Google Scholar
25. 1. Lillemeier BF
    2. Pfeiffer JR
    3. Surviladze Z
    4. Wilson BS
    5. Davis MM

    (2006) Plasma membrane-associated proteins are clustered into islands attached to the cytoskeleton

    PNAS 103:18992–18997.

    https://doi.org/10.1073/pnas.0609009103

    • PubMed
    • Google Scholar
26. 1. Marzahn MR
    2. Marada S
    3. Lee J
    4. Nourse A
    5. Kenrick S
    6. Zhao H
    7. Ben-Nissan G
    8. Kolaitis RM
    9. Peters JL
    10. Pounds S
    11. Errington WJ
    12. Privé GG
    13. Taylor JP
    14. Sharon M
    15. Schuck P
    16. Ogden SK
    17. Mittag T

    (2016) Higher-order oligomerization promotes localization of SPOP to liquid nuclear speckles

    The EMBO Journal 35:1254–1275.

    https://doi.org/10.15252/embj.201593169

    • PubMed
    • Google Scholar
27. 1. Monks CR
    2. Kupfer H
    3. Tamir I
    4. Barlow A
    5. Kupfer A

    (1997) Selective modulation of protein kinase C-theta during T-cell activation

    Nature 385:83–86.

    https://doi.org/10.1038/385083a0

    • PubMed
    • Google Scholar
28. 1. Monks CRF
    2. Freiberg BA
    3. Kupfer H
    4. Sciaky N
    5. Kupfer A

    (1998) Three-dimensional segregation of supramolecular activation clusters in T cells

    Nature 395:82–86.

    https://doi.org/10.1038/25764

    • Google Scholar
29. 1. Otsu N

    (1979) A threshold selection method from Gray-Level histograms

    IEEE Transactions on Systems, Man, and Cybernetics 9:62–66.

    https://doi.org/10.1109/TSMC.1979.4310076

    • Google Scholar
30. 1. Purtic B
    2. Pitcher LA
    3. van Oers NS
    4. Wülfing C

    (2005) T cell receptor (TCR) clustering in the immunological synapse integrates TCR and costimulatory signaling in selected T cells

    PNAS 102:2904–2909.

    https://doi.org/10.1073/pnas.0406867102

    • PubMed
    • Google Scholar
31. 1. Rodina A
    2. Wang T
    3. Yan P
    4. Gomes ED
    5. Dunphy MP
    6. Pillarsetty N
    7. Koren J
    8. Gerecitano JF
    9. Taldone T
    10. Zong H
    11. Caldas-Lopes E
    12. Alpaugh M
    13. Corben A
    14. Riolo M
    15. Beattie B
    16. Pressl C
    17. Peter RI
    18. Xu C
    19. Trondl R
    20. Patel HJ
    21. Shimizu F
    22. Bolaender A
    23. Yang C
    24. Panchal P
    25. Farooq MF
    26. Kishinevsky S
    27. Modi S
    28. Lin O
    29. Chu F
    30. Patil S
    31. Erdjument-Bromage H
    32. Zanzonico P
    33. Hudis C
    34. Studer L
    35. Roboz GJ
    36. Cesarman E
    37. Cerchietti L
    38. Levine R
    39. Melnick A
    40. Larson SM
    41. Lewis JS
    42. Guzman ML
    43. Chiosis G

    (2016) The epichaperome is an integrated chaperome network that facilitates tumour survival

    Nature 538:397–401.

    https://doi.org/10.1038/nature19807

    • PubMed
    • Google Scholar
32. 1. Roybal KT
    2. Sinai P
    3. Verkade P
    4. Murphy RF
    5. Wülfing C

    (2013) The actin-driven spatiotemporal organization of T-cell signaling at the system scale

    Immunological Reviews 256:133–147.

    https://doi.org/10.1111/imr.12103

    • PubMed
    • Google Scholar
33. 1. Roybal KT
    2. Mace EM
    3. Mantell JM
    4. Verkade P
    5. Orange JS
    6. Wülfing C

    (2015) Early signaling in primary T cells activated by antigen presenting cells is associated with a deep and transient lamellal actin network

    PLOS ONE 10:e0133299.

    https://doi.org/10.1371/journal.pone.0133299

    • PubMed
    • Google Scholar
34. 1. Roybal KT
    2. Buck TE
    3. Ruan X
    4. Cho BH
    5. Clark DJ
    6. Ambler R
    7. Tunbridge HM
    8. Zhang J
    9. Verkade P
    10. Wülfing C
    11. Murphy RF

    (2016) Computational spatiotemporal analysis identifies WAVE2 and cofilin as joint regulators of costimulation-mediated T cell actin dynamics

    Science Signaling 9:rs3.

    https://doi.org/10.1126/scisignal.aad4149

    • PubMed
    • Google Scholar
35. 1. Schaeffer EM
    2. Debnath J
    3. Yap G
    4. McVicar D
    5. Liao XC
    6. Littman DR
    7. Sher A
    8. Varmus HE
    9. Lenardo MJ
    10. Schwartzberg PL

    (1999) Requirement for tec kinases rlk and itk in T cell receptor signaling and immunity

    Science 284:638–641.

    https://doi.org/10.1126/science.284.5414.638

    • PubMed
    • Google Scholar
36. 1. Schamel WW
    2. Arechaga I
    3. Risueño RM
    4. van Santen HM
    5. Cabezas P
    6. Risco C
    7. Valpuesta JM
    8. Alarcón B

    (2005) Coexistence of multivalent and monovalent TCRs explains high sensitivity and wide range of response

    The Journal of Experimental Medicine 202:493–503.

    https://doi.org/10.1084/jem.20042155

    • PubMed
    • Google Scholar
37. 1. Seder RA
    2. Paul WE
    3. Davis MM
    4. Fazekas de St Groth B

    (1992) The presence of interleukin 4 during in vitro priming determines the lymphokine-producing potential of CD4+ T cells from T cell receptor transgenic mice

    Journal of Experimental Medicine 176:1091–1098.

    https://doi.org/10.1084/jem.176.4.1091

    • PubMed
    • Google Scholar
38. 1. Sherman E
    2. Barr V
    3. Manley S
    4. Patterson G
    5. Balagopalan L
    6. Akpan I
    7. Regan CK
    8. Merrill RK
    9. Sommers CL
    10. Lippincott-Schwartz J
    11. Samelson LE

    (2011) Functional nanoscale organization of signaling molecules downstream of the T cell antigen receptor

    Immunity 35:705–720.

    https://doi.org/10.1016/j.immuni.2011.10.004

    • PubMed
    • Google Scholar
39. 1. Shin Y
    2. Brangwynne CP

    (2017) Liquid phase condensation in cell physiology and disease

    Science 357:eaaf4382.

    https://doi.org/10.1126/science.aaf4382

    • PubMed
    • Google Scholar
40. 1. Singleton K
    2. Parvaze N
    3. Dama KR
    4. Chen KS
    5. Jennings P
    6. Purtic B
    7. Sjaastad MD
    8. Gilpin C
    9. Davis MM
    10. Wülfing C

    (2006) A large T cell invagination with CD2 enrichment resets receptor engagement in the immunological synapse

    The Journal of Immunology 177:4402–4413.

    https://doi.org/10.4049/jimmunol.177.7.4402

    • PubMed
    • Google Scholar
41. 1. Singleton KL
    2. Roybal KT
    3. Sun Y
    4. Fu G
    5. Gascoigne NR
    6. van Oers NS
    7. Wülfing C

    (2009) Spatiotemporal patterning during T cell activation is highly diverse

    Science Signaling 2:ra15.

    https://doi.org/10.1126/scisignal.2000199

    • PubMed
    • Google Scholar
42. 1. Singleton KL
    2. Gosh M
    3. Dandekar RD
    4. Au-Yeung BB
    5. Ksionda O
    6. Tybulewicz VLJ
    7. Altman A
    8. Fowell DJ
    9. Wulfing C

    (2011) Itk controls the spatiotemporal organization of T cell activation

    Science Signaling 4:ra66.

    https://doi.org/10.1126/scisignal.2001821

    • Google Scholar
43. 1. Storey JD
    2. Taylor JE
    3. Siegmund D

    (2004) Strong control, conservative point estimation and simultaneous conservative consistency of false discovery rates: a unified approach

    Journal of the Royal Statistical Society: Series B 66:187–205.

    https://doi.org/10.1111/j.1467-9868.2004.00439.x

    • Google Scholar
44. 1. Su X
    2. Ditlev JA
    3. Hui E
    4. Xing W
    5. Banjade S
    6. Okrut J
    7. King DS
    8. Taunton J
    9. Rosen MK
    10. Vale RD

    (2016) Phase separation of signaling molecules promotes T cell receptor signal transduction

    Science 352:595–599.

    https://doi.org/10.1126/science.aad9964

    • Google Scholar
45. 1. Tarantino N
    2. Tinevez JY
    3. Crowell EF
    4. Boisson B
    5. Henriques R
    6. Mhlanga M
    7. Agou F
    8. Israël A
    9. Laplantine E

    (2014) TNF and IL-1 exhibit distinct ubiquitin requirements for inducing NEMO-IKK supramolecular structures

    The Journal of Cell Biology 204:231–245.

    https://doi.org/10.1083/jcb.201307172

    • PubMed
    • Google Scholar
46. 1. Varma R
    2. Campi G
    3. Yokosuka T
    4. Saito T
    5. Dustin ML

    (2006) T cell receptor-proximal signals are sustained in peripheral microclusters and terminated in the central supramolecular activation cluster

    Immunity 25:117–127.

    https://doi.org/10.1016/j.immuni.2006.04.010

    • PubMed
    • Google Scholar
47. 1. Wülfing C
    2. Sumen C
    3. Sjaastad MD
    4. Wu LC
    5. Dustin ML
    6. Davis MM

    (2002) Costimulation and endogenous MHC ligands contribute to T cell recognition

    Nature Immunology 3:42–47.

    https://doi.org/10.1038/ni741

    • PubMed
    • Google Scholar
48. 1. Yokosuka T
    2. Sakata-Sogawa K
    3. Kobayashi W
    4. Hiroshima M
    5. Hashimoto-Tane A
    6. Tokunaga M
    7. Dustin ML
    8. Saito T

    (2005) Newly generated T cell receptor microclusters initiate and sustain T cell activation by recruitment of Zap70 and SLP-76

    Nature Immunology 6:1253–1262.

    https://doi.org/10.1038/ni1272

    • PubMed
    • Google Scholar
49. 1. Zanin-Zhorov A
    2. Ding Y
    3. Kumari S
    4. Attur M
    5. Hippen KL
    6. Brown M
    7. Blazar BR
    8. Abramson SB
    9. Lafaille JJ
    10. Dustin ML

    (2010) Protein kinase C-theta mediates negative feedback on regulatory T cell function

    Science 328:372–376.

    https://doi.org/10.1126/science.1186068

    • PubMed
    • Google Scholar

Thank you for submitting your article "Transient protein accumulation at the center of the T cell antigen presenting cell interface drives IL-2 secretion" for consideration by eLife. Your article has been reviewed by three peer reviewers, including Michael L Dustin as the Reviewing Editor and Reviewer #1, and the evaluation has been overseen by Arup Chakraborty as the Senior Editor.

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

Summary:

Your work is commended for quantitative analysis of supramolecular structures with a focus on the SMACs including the use of chimeric signaling adapter to better understand the role of costimulation and the Itk kinase in supramolecular structures and function. However, there were a number of concerns raised. While there are many issues, most can be addressed without new data, but relate significantly to the organisation of information. The comments where new data appear to be necessary to address the concern are marked with asterisks (*).

1) Regarding inconsistency in IL-2 protein and message:

a) The data presented in Figure 1A and B are inconsistent. On the one hand, it is demonstrated that IL-2 amounts secreted to the supernatant of full stimulus T cells vs. full stimulus Itk KO at 10 µM is relatively unchanged, but strangely, on the other hand, the IL-2 mRNA is drastically lower in the Itk KO. No explanations are provided for this inconsistency between the protein and the mRNA levels. This issue should be addressed by the authors.

* Could you provide a time course of IL-2 accumulation in comparison of IL-2 mRNA to provide a clear explanation and bolster the physiological significant of the subsequent investigation.

b) Figure 1C and D. The authors profile the signaling molecules at the different compartments of the IS at different settings, i.e., full stimulation in comparison with co-stimulation blockade (C) and full stimulation vs. full stimulus Itk deficiency (D). However, it is difficult to deduce clear-cut conclusions since the profile of signaling molecules in C is much wider than in D. The authors should explain why they neglect specific molecules in D that were analyzed in C. In addition, the authors should include the profile of signaling molecule localization for the combined treatment of co-stimulation blockade and Itk knockout, as these appear to create an additive effect. Similarly, for Figure 1—figure supplement 2B, Figure 1—figure supplement 3 and Figure 1—figure supplement 4, the graphs of full stimulation vs. co-stimulation blockade/Itk deficiency should compare similar signaling molecular kinetics. Also, Figure 1—figure supplement 3 shows the kinetics of Akt (co-stimulation blockade) that is missing in Figures 1—figure supplements 2 and 4 (Itk deficiency), and Figure 1—figure supplement 4 shows the kinetics of PKC that is missing in Figure 1—figure supplements 2 and 3.

c) In Figure 5, the authors should show IL-2 concentration in addition to IL-2 mRNA amounts due to the inconsistency between them in Figure 1. In addition, the authors show that LAT-V3 restored LAT centralization under co-stimulation inhibition/Itk deficiency, and that LAT-V3 partially restores Grb-2 and Lck centralization under Itk deficiency, yet it had no effect on IL-2 mRNA. This should be discussed by the authors, as it suggests that it is more than LAT centralization that affects IL-2 production.

d) In Figure 5, the authors showed Grb-2, Lck, and VAV-1 centralization as a function of LAT-V3 presence upon stimulation, but omitted the LAT-VAV and LAT- PLCδPH constructs and how they affect signaling molecule centralization.

* These should be included in the data as they also rescued LAT centralization. In addition, the authors should show how the different LAT constructs under the combined treatment of Itk and the co-stimulation blockade influence Grb-2, Lck, and VAV-1, as this may have an additive effect that can be noticeable.

2) Missing information: a) * In the WBs of LAT phosphorylation (Figure 2—figure supplement 2 and Figure 4—figure supplement 1F) the unstimulated time point (often called 0 min stimulation) is missing. Thus, the fold increase upon stimulation remains unknown. Similarly, the total amount of LAT needs to be shown; e.g. in S5D GFP-LATV3 is less phosphorylated than wt GFP-LAT. Was this due to lower expression levels of LATV3 and per molecule there was actually more LATV3 phosphorylation than wt LAT phosphorylation?

b) * Relative expression levels of wt to mutant molecules need to be shown for the most important experiments.

c) It’s not exactly clear how the authors go from the STED results to the volumes. In order to be able to do this the authors would need to report the resolution of the STED system. Is resolution with STED will have a tradeoff between lateral and axial resolution the best isotropic resolution for 3D step might be on the order of 80 nm xy and 100 nm z. Higher xy resolution might be achieved, but only at cost of degraded z resolution.

d) The correlative fluorescence-EM seems to be focused on temporal correlation but it’s not clear if they are look at the same call imaged by fluorescence and if they can register the fluorescence and EM data? Can they really say what structure is the fluorescence reported cSMAC in a given EM view. * A supplementary figure could be constructed to better illustrate how and to what level they are look at correlations.

3) Chimeric receptors:

a) Figure 3. The authors should introduce the fused V3 domain to LAT when it is first used, and present the goal of this experiment. I assume that the authors used LAT-V3 under co-stimulation blockade to demonstrate that it rescue membrane undulation; however, this is not mentioned. Did co-stimulation blockade abrogate membrane undulation in Figure 3? Figure 3D does not show these data. Although LAT-V3 is mentioned in Figure 3, it is explained only in Figure 4.

b) * Equipping LAT (but also SLP76 and Grb2) with new interaction domains is supposed to generate chimeric molecules that bind to new binding partners. For example, LATVav should also bind to proteins that normally would bind to Vav. This should be tested by classical immunoprecipitation experiments (and if this is difficult due to low cell numbers in the primary cells, one could do it with stable cell lines expressing the chimeric molecules). Without these experiments the new interactions remain hypothetical.

c) The chimeras should have shown an intermediate spatio-temporal localization compared to the individual proteins, i.e. LATVav should have shown a pattern between the ones of LAT and Vav. Was this the case for all of the chimeras?

[…] The comments where new data appear to be necessary to address the concern are marked with asterisks (*).

1) Regarding inconsistency in IL-2 protein and message:

a) The data presented in Figure 1A and B are inconsistent. On the one hand, it is demonstrated that IL-2 amounts secreted to the supernatant of full stimulus T cells vs. full stimulus Itk KO at 10 µM is relatively unchanged, but strangely, on the other hand, the IL-2 mRNA is drastically lower in the Itk KO. No explanations are provided for this inconsistency between the protein and the mRNA levels. This issue should be addressed by the authors.

* Could you provide a time course of IL-2 accumulation in comparison of IL-2 mRNA to provide a clear explanation and bolster the physiological significant of the subsequent investigation.

IL-2 mRNA and protein levels don’t necessarily align as IL-2 mRNA levels only measure IL-2 generation whereas the protein levels measure the difference between IL-2 generation and consumption. Concurrent diminished IL-2 generation and consumption can leave IL-2 protein levels largely unchanged while resulting in substantially reduced IL-2 mRNA levels.

We have provided a time course of IL-2 mRNA versus protein levels in Figure 1—figure supplement 1 showing that IL-2 generation is limited to the first few hours of the reactivation of primed T cells. Providing a molecular explanation for these dynamics we show that the transcription factor NFκB that is required for IL-2 gene transcription translocates to the nucleus with a comparable dynamics.

b) Figure 1C and D. The authors profile the signaling molecules at the different compartments of the IS at different settings, i.e., full stimulation in comparison with co-stimulation blockade (C) and full stimulation vs. full stimulus Itk deficiency (D). However, it is difficult to deduce clear-cut conclusions since the profile of signaling molecules in C is much wider than in D. The authors should explain why they neglect specific molecules in D that were analyzed in C. In addition, the authors should include the profile of signaling molecule localization for the combined treatment of co-stimulation blockade and Itk knockout, as these appear to create an additive effect. Similarly, for Figure 1—figure supplement 2B, Figure 1—figure supplement 3 and Figure 1—figure supplement 4, the graphs of full stimulation vs. co-stimulation blockade/Itk deficiency should compare similar signaling molecular kinetics. Also, Figure 1—figure supplement 3 shows the kinetics of Akt (co-stimulation blockade) that is missing in Figures 1—figure supplements 2 and 4 (Itk deficiency), and Figure 1—figure supplement 4 shows the kinetics of PKC that is missing in Figure 1—figure supplements 2 and 3.

The reviewers are correct in pointing out that we have not imaged all sensors equally across all experimental conditions. This has practical reasons. Our initial submission is based on the time-resolved image acquisition and analysis of 4781 cell couples in addition to previous data on more than a thousand of cell couples also included in Figure 1C/D. Extending the data to an equal coverage of sensors across all experimental conditions would have required more than a thousand additional cell couples. While we can efficiently handle large amounts of imaging data such extension still would require months of work. The imaging data as they are unambiguously support impaired formation of the cSMAC upon costimulation blockade and Itk deficiency. We agree with the reviewers that consistent sensor coverage across all experimental conditions is desirable and we are extending the Itk ko data set accordingly for future work. However, we believe that in the context of this manuscript the months of work required to harmonize the data sets doesn’t stand in a reasonable relation to the resulting gain in scientific insight.

In the supplementary figures we only show data that hasn’t been previously published with a reference to the relevant previous work in Figure 1—figure supplement 5. We will gladly add previously published work if requested.

c) In Figure 5, the authors should show IL-2 concentration in addition to IL-2 mRNA amounts due to the inconsistency between them in Figure 1. In addition, the authors show that LAT-V3 restored LAT centralization under co-stimulation inhibition/Itk deficiency, and that LAT-V3 partially restores Grb-2 and Lck centralization under Itk deficiency, yet it had no effect on IL-2 mRNA. This should be discussed by the authors, as it suggests that it is more than LAT centralization that affects IL-2 production.

We agree with the reviewers that it is interesting that the drastically enhanced central LAT-V3 localization with the accompanying moderate enhancement of Grb2 and Lck centrality in Itk-deficient T cells didn’t lead to restoration of IL-2 secretion. We had already discussed that our data strongly suggest that extent and dynamics of cSMAC formation, moderate and an emphasis on the first two minutes, respectively, are critical for its function in promotion IL-2 secretion with LAT-V3 getting both wrong. Very briefly, we suggest that excessive cSMAC formation as seen with LAT-V3 switches cSMAC function from a reaction crucible to sequestration of signaling intermediates. We now point out the delayed timing of the moderately enhanced Grb2 and Lck localization, have added data on enhanced Grb2 interface recruitment upon LAT-V3 expression in Itk-deficient T cells upon costimulation blockade and refer to an expanded Discussion section when showing the data.

As discussed in our response point 1a, mRNA generation is a more direct and more sensitive readout of T cell activation. Given the moderate changes in IL-2 mRNA levels shown, it seems highly unlikely that a determination of IL-2 protein levels by ELISA can contribute.

d) In Figure 5, the authors showed Grb-2, Lck, and VAV-1 centralization as a function of LAT-V3 presence upon stimulation, but omitted the LAT-VAV and LAT- PLCδPH constructs and how they affect signaling molecule centralization.

* These should be included in the data as they also rescued LAT centralization. In addition, the authors should show how the different LAT constructs under the combined treatment of Itk and the co-stimulation blockade influence Grb-2, Lck, and VAV-1, as this may have an additive effect that can be noticeable.

We have exchanged a set of emails with Prof. Dustin regarding this concern.

Our question: We appreciate the importance of the question of how spatiotemporal signaling organization changes when we force central LAT localization. For that reason, we have included the Grb-2, Lck, Vav-1 localization data upon LAT-V3 expression. We will gladly extend these data to costimulation blockade in Itk-deficient T cells as suggested. However, recapitulating the entire LAT-V3 data with LAT-Vav and LAT- PLCδPH requires substantial effort well beyond two months. We first would have to clone the necessary six dual expression constructs and then image 24 experimental conditions (two times (LAT-Vav and LAT- PLCδPH) three (Grb-2, Vav-1, Lck) constructs times four stimuli) that with the somewhat lower efficiency of the dual expression constructs will require around 100 imaging runs. With work required to address other review concerns 4 imaging runs per week dedicated to this question seems reasonable. The LAT-V3 data already make the point that overall spatiotemporal changes upon forced LAT centralization are rather modest. While experiments obviously can yield intriguing surprises, it seems most likely that we will see similarly modest changes in spatiotemporal signaling organization upon expression of LAT-Vav and LAT-PLCδPH. I our opinion only investigating a larger number of sensors has a substantial chance of discovering meaningful differences in signaling organization upon expression of LAT-V3 versus LAT-Vav versus LAT-PLCδPH, something that is essentially a study in itself. Therefore, we are uncertain whether in an investigation of the localization of Grb-2, Lck, and Vav-1 upon expression of LAT-Vav and LAT-PLCδPH the effort required stands in a reasonable relation to the likely outcomes in the context of this manuscript. I would very much appreciate if you could suggest how you would like us to proceed.

The answer: “Thank you for calling our attention to this issue and please try address the reviewer concern in a manner that is feasible within the 2 month window. For example, an acceptable solution may be to include a fix and stain experiment to support the localisation of some of the endogenous players requested by the reviewer at key times predicted by your model. Please outline the difficulties in addressing the concern in your point-by-point response to the reviewer comments and this will allow us to take this into account when evaluating your revision.”

As requested by the reviewers, we have now imaged Grb2, Lck and Vav-1 in the presence of LAT V3 in Itk-deficient 5C.C7 T cells upon costimulation blockade. Interestingly, we found that expression of LAT V3 in Itk-deficient 5C.C7 T cells upon costimulation blockade triggered Grb2 interface accumulation to an extent not seen under any of the other experimental conditions. We suggest the following scenario to explain these findings. Grb2 doesn’t effectively compete for binding to the cSMAC under strong and even moderate T cell activation conditions. Likely, stronger binding signaling intermediates are activated to a sufficient extent under those conditions to outcompete Grb2. However, with the strong signaling attenuation provided by the combination of costimulation blockade and Itk-deficiency activation of more competitive signaling intermediates becomes sufficiently inefficient to allow Grb2 recruitment to the LAT V3-based signaling complexes.

We agree with the reviewers that an analysis of the spatiotemporal organization of T cell signaling upon expression of LAT-Vav and LAT-PLCδPH is of interest. We have, therefore, extensively thought about the feasibility of the staining experiments suggested in Prof. Dustin’s reply. Staining experiments actually require substantially larger numbers of T cells than our imaging experiments where 30,000 T cells are generally enough to complete an entire data set. For the experiments suggested here the T cells still need to be FACS sorted for expression of LAT-Vav or LAT-PLCδPH and T cell numbers thus can become limiting. Moreover, we know from the careful analysis of T cell morphology in electron microscopy experiments (Roybal et al., 2015) that even when using centrifugation to force cell coupling, fixed T cell/APC samples invariably contain a range of time points relative to tight cell couple formation that can only be approximated. Given the importance of cSMAC dynamics established in our manuscript imprecise determination of time of cell coupling is highly problematic. Looking at the 12 conditions imaged with LAT-V3, there is no obvious condition that when determined in isolation should provide substantial insight into how LAT-Vav or LAT-PLCδPH shape the spatiotemporal organization of T cell signaling. This leaves us with having to use staining experiments with the constraints just discussed across a range of T cell activation conditions, something likely to take months of work. We thus believe that any meaningful insight into the spatiotemporal organization of T cell signaling upon expression of LAT-Vav and LAT-PLCδPH remains well beyond the (already extended) time scale of this revision. We don’t believe that lack of such data weakens the manuscript: As overexpression of the more dominantly localized LAT-V3 had largely minor impacts on the localization of Grb2, Lck and Vav-1, is seems unlikely that the less dominantly localized LAT-Vav and LAT-PLCδPH will yield changes that are sufficientlydramatic to be readily detectable.

Despite these concerns we have executed a first staining experiment as suggested. We have sorted T cells transduced to express LAT-Vav-GFP, activated them with APC and peptide for 1min under full stimulus and costimulation-blocked conditions, fixed and stained the cell couples with an antibody against active Src kinase family members as shown in Author response image 1.

Author response image 1

Download asset Open asset

Limiting the biological information that could be gained from this experiment, we, first, lost part of the GFP-fluorescence during cell couple fixation. A reliable determination of the formation of a LAT-Vav-based cSMAC became questionable. Second, as Src family kinases are expressed in both the T cell and the APC, unambiguous assignment of interface fluorescence to either cell type was not feasible. This limitation will similarly extend to a determination of interface accumulation of Vav-1 or Grb2. Even assuming dominant accumulation of active Src kinase family members in the T cell, substantial central Src kinase family activation was not found. With all technical caveats to be considered, these data seem consistent with the limited central accumulation of Lck upon expression of LAT-V3 (Figure 6). We believe that these experiments illustrate our willingness to pursue reviewers’ suggestions and support our notion that fixed cell staining experiments are unlikely to generate insight into the organization of T cell signaling upon cSMAC reconstitution any more efficiently than the live cell imaging experiments do.

2) Missing information: a) * In the WBs of LAT phosphorylation (Figure 2—figure supplement 2 and Figure 4—figure supplement 1F) the unstimulated time point (often called 0 min stimulation) is missing. Thus, the fold increase upon stimulation remains unknown. Similarly, the total amount of LAT needs to be shown; e.g. in S5D GFP-LATV3 is less phosphorylated than wt GFP-LAT. Was this due to lower expression levels of LATV3 and per molecule there was actually more LATV3 phosphorylation than wt LAT phosphorylation?

We have included a no peptide control in the first three of the seven experimental replicates of the determination of LAT phosphorylation as a function of time. Having established minimal LAT phosphorylation as already published twice before using the same experimental system (Purtic et al., 2005, Supplementary Figure 1 and Singleton et al., 2006,, Figure 8B), we then dropped this control for the remainder of the experiments. We apologize that the level of LAT phosphorylation upon no peptide stimulation was shown in a rather misleading fashion in our original submission. We have now revised Figure 2D to more clearly identify this important control. The representative Western blots shown in the supplement came from repeats without the no peptide control. Author response image 2 shows one of the earlier blots (lacking the 10 min time point, without Itk-deficient T cells, and with an additional experimental condition not further pursued) demonstrating the lack of LAT phosphorylation in the absence of agonist peptide (lane labelled ‘-‘). We will gladly add the blot to the supplement if requested.

Author response image 2

Download asset Open asset

The expression level of all wild type and mutant molecules was fixed by FACS sorting on the GFP tag and is, therefore, the same across the entire manuscript with a concentration of about 2µM (Sci. Signal. 2009, 2, ra15). This is now stated consistently. For further quantitative insight, we have now twice measured the amount of GFP-tagged LAT relative to the endogenous protein and report this quantification in the Results section and Figure 4—figure supplement 1B, C. Briefly, GFP fusion proteins were expressed at 2.1 ± 0.7-fold the endogenous level of LAT in non-transduced T cells (Figure 4—figure supplement 1B, C) with little change to endogenous LAT levels in the transduced T cells. We will gladly add a third replicate if requested. This quantification extends equally to all LAT fusion proteins because of the use of GFP-based FACS sorting.

b) * Relative expression levels of wt to mutant molecules need to be shown for the most important experiments.

The expression of all wild type and mutant molecules was fixed by FACS sorting on the GFP tag. This is now stated consistently.

c) It’s not exactly clear how the authors go from the STED results to the volumes. In order to be able to do this the authors would need to report the resolution of the STED system. Is resolution with STED will have a tradeoff between lateral and axial resolution the best isotropic resolution for 3D step might be on the order of 80 nm xy and 100 nm z. Higher xy resolution might be achieved, but only at cost of degraded z resolution.

We have measured the closest detectable distance between adjacent puncta as the experimentally achieved resolution as 100nm in all dimensions. This results in a smallest theoretically detectable cluster size of 0.001µm³, well below the experimental cut-off of 0.04/0.06µm³ that distinguishes non-specific from LAT/phospho-LAT-specific signals, as detailed in the Materials and methods section. We now report the experimentally determined STED resolution in the Materials and methods.

d) The correlative fluorescence-EM seems to be focused on temporal correlation but it’s not clear if they are look at the same call imaged by fluorescence and if they can register the fluorescence and EM data? Can they really say what structure is the fluorescence reported cSMAC in a given EM view. * A supplementary figure could be constructed to better illustrate how and to what level they are look at correlations.

We have now generated a figure to illustrate the correlative light electron microscopy work flow, Figure 3—figure supplement 1. Because of the use of finder grids we can readily match cells in the bright field and EM images. Because of the resolution limits of spinning disk confocal microscopy, a sub µm-scale registration of the fluorescence and electron microscopy data is difficult. To account for this difficulty, we have chosen a conservative image analysis approach that biases the outcome against the hypothesis to be tested, i.e. that enhanced membrane undulations are associated with the cSMAC: To determine membrane undulations in the cSMAC region, the interface diameter in the electron micrograph was divided into four equal sections with the central two sections defined as the cSMAC. This is a conservative assumption as cSMACs were by definition contained within this region, yet commonly smaller than the entire half of the interface diameter. This is now detailed in the Materials and methods section.

3) Chimeric receptors:

a) Figure 3. The authors should introduce the fused V3 domain to LAT when it is first used, and present the goal of this experiment. I assume that the authors used LAT-V3 under co-stimulation blockade to demonstrate that it rescue membrane undulation; however, this is not mentioned. Did co-stimulation blockade abrogate membrane undulation in Figure 3? Figure 3D does not show these data. Although LAT-V3 is mentioned in Figure 3, it is explained only in Figure 4.

We now briefly introduce the LAT-V3 construct in the description of the CLEM experiments and refer to the LAT-V3 schematic in the legend of Figure 3C. However, to avoid duplication, we leave a full justification of the use LAT-V3 to the detailed Introduction in the subsequent paragraph as referred to now.

We have earlier published reduced membrane undulations upon costimulation blockade and now refer to these data. That said, the aim of the CLEM experiments was to determine whether cSMAC formation is associated with enhanced membrane undulations. Costimulation blockade substantially impairs cSMAC formation, making the determination of an association of cSMAC formation with membrane undulation rather difficult. Yet published data are now referred to. Activation of 5C.C7 T cells expressing LAT-V3 under full stimulus conditions as another possible comparator leads to enhancement of cSMAC formation very similar to the LAT-V3/costimulation blocked condition. We are not sure what additional insight would be gained including both conditions. Having to decide between investigating LAT-V3 under full stimulus or costimulation-blocked conditions we felt that a restorative condition, i.e. a condition with cSMAC formation forced when normally absent, would be more informative.

b) * Equipping LAT (but also SLP76 and Grb2) with new interaction domains is supposed to generate chimeric molecules that bind to new binding partners. For example, LATVav should also bind to proteins that normally would bind to Vav. This should be tested by classical immunoprecipitation experiments (and if this is difficult due to low cell numbers in the primary cells, one could do it with stable cell lines expressing the chimeric molecules). Without these experiments the new interactions remain hypothetical.

We agree with the reviewers that a determination of the changes in the LAT interactome upon addition of new protein interaction domains is of interest. We have tried to address this question using pulldowns with anti-GFP beads. These experiments proved complex in having to consider number of T cells required, T cell activation conditions (resting, pervandate, APC plus peptide) and means of analysis of LAT binders (hypothesis-driven or unbiased, background of unspecific bead-binding proteins). We therefore exchanged emails with the editor regarding these challenges and received the following answer:

“If the authors have tried the biochemistry and failed on this, the best option may just be to acknowledge this in the paper and that the interactions are predicted, but are difficult to verify. We know that all of the individual SH2 and SH3 interactions are very weak and could not lead to immunoprecipitation on their own. Extracting complexes that are stable enough for indirect immunoprecipitation depends upon multi protein interactions that may be hard to predict..… So if the manuscript is otherwise ready I think they should acknowledge that these interactions are predicted, but were not sufficiently stable to be captured by immunoprecipitation and resubmit.”

We have now acknowledged in the Results section that ‘while the addition of protein interaction domains to LAT is predicted to alter their interactome, initial experiments to support this notion remained inconclusive’.

c) The chimeras should have shown an intermediate spatio-temporal localization compared to the individual proteins, i.e. LATVav should have shown a pattern between the ones of LAT and Vav. Was this the case for all of the chimeras?

In the context of supramolecular complex formation the localization of a fusion protein is not the ‘average’ of the localization of its components. For example, the PKCθ V3 domain in isolation does not localize to the interface center even under full stimulus conditions (Figure 5—figure supplement 1A) nor does LAT upon costimulation blockade (Figure 2B). However, the LAT-V3 fusion protein shows dominant central localization upon costimulation blockade (Figure 4B). How is this possible? By generating a LAT-V3 fusion proteins one doesn’t only add the localization preferences of LAT and the PKC V3 domain, one also increases the valence of the fusion protein over its components. As increased valence is a key facilitator of recruitment to supramolecular complexes localization properties of fusion proteins can become qualitatively different from those of their components. This exciting argument has now been added to the Discussion section.

Author details

1. Danielle J Clark

   School of Cellular and Molecular Medicine, University of Bristol, Bristol, United Kingdom

   Contribution

   Formal analysis, Investigation

   Contributed equally with

   Laura E McMillan

   Competing interests

   No competing interests declared

2. Laura E McMillan

   School of Cellular and Molecular Medicine, University of Bristol, Bristol, United Kingdom

   Contribution

   Formal analysis, Investigation

   Contributed equally with

   Danielle J Clark

   Competing interests

   No competing interests declared

3. Sin Lih Tan

   School of Cellular and Molecular Medicine, University of Bristol, Bristol, United Kingdom

   Contribution

   Formal analysis, Investigation

   Competing interests

   No competing interests declared

4. Gaia Bellomo

   School of Cellular and Molecular Medicine, University of Bristol, Bristol, United Kingdom

   Contribution

   Formal analysis, Investigation

   Competing interests

   No competing interests declared

5. Clementine Massoue

   School of Cellular and Molecular Medicine, University of Bristol, Bristol, United Kingdom

   Contribution

   Formal analysis, Investigation

   Competing interests

   No competing interests declared

6. Harry Thompson

   School of Cellular and Molecular Medicine, University of Bristol, Bristol, United Kingdom

   Contribution

   Investigation

   Competing interests

   No competing interests declared

7. Lidiya Mykhaylechko

   School of Cellular and Molecular Medicine, University of Bristol, Bristol, United Kingdom

   Contribution

   Investigation

   Competing interests

   No competing interests declared

8. Dominic Alibhai

   School of Biochemistry, University of Bristol, Bristol, United Kingdom

   Contribution

   Formal analysis

   Competing interests

   No competing interests declared

9. Xiongtao Ruan

   Computational Biology Department, School of Computer Science, Carnegie Mellon University, Pittsburgh, United States

   Contribution

   Formal analysis

   Competing interests

   No competing interests declared

10. Kentner L Singleton

    Department of Immunology, University of Texas Southwestern Medical Center, Dallas, United States

    Contribution

    Formal analysis, Investigation

    Competing interests

    No competing interests declared

11. Minna Du

    Department of Immunology, University of Texas Southwestern Medical Center, Dallas, United States

    Contribution

    Formal analysis, Investigation

    Competing interests

    No competing interests declared

12. Alan Hedges

    School of Cellular and Molecular Medicine, University of Bristol, Bristol, United Kingdom

    Contribution

    Formal analysis

    Competing interests

    No competing interests declared

13. Pamela L Schwartzberg

    Genetic Disease Research Branch, National Human Genome Research Institute, National Institutes of Health, Bethesda, United States

    Contribution

    Resources

    Competing interests

    No competing interests declared

14. Paul Verkade

    School of Biochemistry, University of Bristol, Bristol, United Kingdom

    Contribution

    Conceptualization, Formal analysis, Investigation

    Competing interests

    No competing interests declared

15. Robert F Murphy

    1. Computational Biology Department, School of Computer Science, Carnegie Mellon University, Pittsburgh, United States
    2. Department of Biological Sciences, Carnegie Mellon University, Pittsburgh, United States
    3. Department of Biomedical Engineering, Carnegie Mellon University, Pittsburgh, United States
    4. Department of Machine Learning, Carnegie Mellon University, Pittsburgh, United States
    5. Freiburg Institute for Advanced Studies, Albert Ludwig University of Freiburg, Freiburg, Germany
    6. Faculty of Biology, Albert Ludwig University of Freiburg, Freiburg, Germany

    Contribution

    Formal analysis, Supervision, Funding acquisition

    Competing interests

    No competing interests declared

    "This ORCID iD identifies the author of this article:" 0000-0003-0358-901X

16. Christoph Wülfing

    1. School of Cellular and Molecular Medicine, University of Bristol, Bristol, United Kingdom
    2. Department of Immunology, University of Texas Southwestern Medical Center, Dallas, United States

    Contribution

    Conceptualization, Data curation, Formal analysis, Supervision, Funding acquisition, Investigation, Writing—original draft, Project administration, Writing—review and editing

    For correspondence

    Christoph.Wuelfing@bristol.ac.uk

    Competing interests

    No competing interests declared

    "This ORCID iD identifies the author of this article:" 0000-0002-6156-9861

• 1,680

  Page views

• 213

  Downloads

• 4

  Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.