Ligand-Mediated Friction Determines Morphodynamics of Spreading T Cells

      Abstract

      Spreading of T cells on antigen presenting cells is a crucial initial step in immune response. Spreading occurs through rapid morphological changes concomitant with the reorganization of surface receptors and of the cytoskeleton. Ligand mobility and frictional coupling of receptors to the cytoskeleton were separately recognized as important factors but a systematic study to explore their biophysical role in spreading was hitherto missing. To explore the impact of ligand mobility, we prepared chemically identical substrates on which molecules of anti-CD3 (capable of binding and activating the T cell receptor complex), were either immobilized or able to diffuse. We quantified the T cell spreading area and cell edge dynamics using quantitative reflection interference contrast microscopy, and imaged the actin distribution. On mobile ligands, as compared to fixed ligands, the cells spread much less, the actin is centrally, rather than peripherally distributed and the edge dynamics is largely altered. Blocking myosin-II or adding molecules of ICAM1 on the substrate largely abrogates these differences. We explain these observations by building a model based on the balance of forces between activation-dependent actin polymerization and actomyosin-generated tension on one hand, and on the frictional coupling of the ligand-receptor complexes with the actin cytoskeleton, the membrane and the substrate, on the other hand. Introducing the measured edge velocities in the model, we estimate the coefficient of frictional coupling between T Cell receptors or LFA-1 and the actin cytoskeleton. Our results provide for the first time, to our knowledge, a quantitative framework bridging T cell-specific biology with concepts developed for integrin-based mechanisms of spreading.

      Introduction

      Spreading is the key to the T cell’s physiological role of recognizing rare and low abundance antigenic ligands on the surface of antigen presenting cells (APCs) (
      • Fooksman D.R.
      • Vardhana S.
      • Dustin M.L.
      • et al.
      Functional anatomy of T cell activation and synapse formation.
      ). The extent of T cell spreading, while interacting with physiological ligands, is correlated with signal strength (
      • Crites T.J.
      • Padhan K.
      • Varma R.
      • et al.
      TCR Microclusters pre-exist and contain molecules necessary for TCR signal transduction.
      ), and is also an early marker of T cell proliferation (
      • Cretel E.
      • Touchard D.
      • Pierres A.
      • et al.
      A new method for rapid detection of T lymphocyte decision to proliferate after encountering activating surfaces.
      ). T cells undergo repeated spreading events punctuated by migration episodes to search for agonist antigens, resulting in dynamical changes in cellular morphology, accompanied by molecular reorganization at the T cell/APC interface (
      • Dustin M.L.
      Hunter to gatherer and back: immunological synapses and kinapses as variations on the theme of amoeboid locomotion.
      ).
      Early in vitro studies on T cells adhering to supported lipid bilayers (SLBs) via bonds between antigenic ligands and T cell receptors (TCRs), showed that receptors accumulate in the contact area (
      • Chan P.-Y.
      • Lawrence M.B.
      • Springer T.A.
      • et al.
      Influence of receptor lateral mobility on adhesion strengthening between membranes containing LFA-3 and CD2.
      ,
      • Dustin M.L.
      • Ferguson L.M.
      • Golan D.E.
      • et al.
      Visualization of CD2 interaction with LFA-3 and determination of the two-dimensional dissociation constant for adhesion receptors in a contact area.
      ), a phenomenon shown to also occur purely passively in model systems exhibiting ligand/receptor diffusion (
      • Fenz S.F.
      • Merkel R.
      • Sengupta K.
      Diffusion and intermembrane distance: case study of avidin and E-cadherin mediated adhesion.
      ). Over the last 15 years, numerous studies on SLBs carrying ligands of TCR and the integrin LFA1 (ligand: ICAM1), have revealed drastic receptor reorganization at the T cell/APC interface leading to the formation of the immunological synapse (
      • Grakoui A.
      • Bromley S.K.
      • Dustin M.L.
      • et al.
      The immunological synapse: a molecular machine controlling T cell activation.
      ). This synapse, organized into compartments called supramolecular activation clusters (SMACs), is itself formed by coalescence of microclusters of TCR on one hand (
      • Varma R.
      • Campi G.
      • Dustin M.L.
      • et al.
      T cell receptor-proximal signals are sustained in peripheral microclusters and terminated in the central supramolecular activation cluster.
      ,
      • Kaizuka Y.
      • Douglass A.D.
      • Vale R.D.
      • et al.
      Mechanisms for segregating T cell receptor and adhesion molecules during immunological synapse formation in Jurkat T cells.
      ), and of integrins on the other hand (
      • Hartman N.C.
      • Nye J.A.
      • Groves J.T.
      Cluster size regulates protein sorting in the immunological synapse.
      ), both of which are actively transported along the T cell/APC interface. Experiments confining the ligands within micron-size corrals in SLBs have revealed the role of actin in receptor transport, and have emphasized its importance in signaling (
      • Mossman K.D.
      • Campi G.
      • Dustin M.L.
      • et al.
      Altered TCR signaling from geometrically repatterned immunological synapses.
      ,
      • Yu C.H.
      • Wu H.-J.
      • Groves J.T.
      • et al.
      Altered actin centripetal retrograde flow in physically restricted immunological synapses.
      ).
      Major features of T cell activation and spreading response was also recapitulated on substrates coated with activating anti-CD3 (an antibody directed against the CD3-ε subunit of the TCR complex) lacking lateral mobility (
      • Bunnell S.C.
      • Hong D.I.
      • Samelson L.E.
      • et al.
      T cell receptor ligation induces the formation of dynamically regulated signaling assemblies.
      ), and revealed that spreading is accompanied by dynamic actin polymerization (
      • Bunnell S.C.
      • Kapoor V.
      • Samelson L.E.
      • et al.
      Dynamic actin polymerization drives T cell receptor-induced spreading: a role for the signal transduction adaptor LAT.
      ). The similarity of T cell response to immobilized anti-CD3 in the absence of ICAM, and to SLBs featuring mobile dual ligands, is intriguing (
      • Bunnell S.C.
      Multiple microclusters: diverse compartments within the immune synapse.
      ). A recent work has tried to bridge the gap between these two extreme cases by systematic variation of anti-CD3 mobility on supported lipid bilayers using phase transitions in lipid mixtures to control ligand diffusion (
      • Hsu C.-J.
      • Hsieh W.-T.
      • Baumgart T.
      • et al.
      Ligand mobility modulates immunological synapse formation and T cell activation.
      ). However, the diffusion range considered did not cover the fully immobilized case. T cells interacting with ligands that were more mobile exhibited better signaling, in contrast to several other past studies that indicated that T cells are more sensitive to immobile ligands (
      • Segura J.-M.
      • Guillaume P.
      • Luescher I.F.
      • et al.
      Increased mobility of major histocompatibility complex I-peptide complexes decreases the sensitivity of antigen recognition.
      ,
      • Ma Z.
      • Sharp K.A.
      • Finkel T.H.
      • et al.
      Surface-anchored monomeric agonist pMHCs alone trigger TCR with high sensitivity.
      ,
      • Luxembourg A.T.
      • Brunmark A.
      • Cai Z.
      • et al.
      Requirements for stimulating naive CD8+ T cells via signal 1 alone.
      ). These contradictions underline the importance of ligand mobility but point to the need for further studies.
      Interestingly, T cells have recently been shown to be mechanosensitive (
      • Judokusumo E.
      • Tabdanov E.
      • Kam L.C.
      • et al.
      Mechanosensing in T lymphocyte activation.
      ), and to exert forces through CD3 receptors (
      • Bashour K.T.
      • Gondarenko A.
      • Kam L.C.
      • et al.
      CD28 and CD3 have complementary roles in T-cell traction forces.
      ). In the context of spreading, it can be speculated that the resistance of the TCR-complex, via CD3, to dragging by actin generated forces is the key to understanding the biophysical basis of the impact of ligand mobility. Elegant experiments on cross-linked mobile receptors have led to the hypothesis that the local frictional coupling between actin and receptors is the driving force for directed movement of microclusters of TCR and integrins on SLBs (
      • Hartman N.C.
      • Nye J.A.
      • Groves J.T.
      Cluster size regulates protein sorting in the immunological synapse.
      ). The link between actin and the dynamical changes in cell morphology during spreading was recently emphasized for the case of immobilized ligands (
      • Lam Hui K.
      • Kwak S.I.
      • Upadhyaya A.
      Adhesion-dependent modulation of actin dynamics in Jurkat T cells.
      ), but equivalent experiments on mobile ligands are so far missing.
      In the context of general cell biology, friction associated with the actin cortex is recognized as an important factor in determining global aspects of cell spreading (
      • Cuvelier D.
      • Théry M.
      • Mahadevan L.
      • et al.
      The universal dynamics of cell spreading.
      ), as well as generation of local, receptor-mediated traction forces that are linked to the actin retrograde flow (
      • Jurado C.
      • Haserick J.R.
      • Lee J.
      Slipping or gripping? Fluorescent speckle microscopy in fish keratocytes reveals two different mechanisms for generating a retrograde flow of actin.
      ,
      • Craig E.M.
      • Van Goor D.
      • Mogilner A.
      • et al.
      Membrane tension, myosin force, and actin turnover maintain actin treadmill in the nerve growth cone.
      ,
      • Giannone G.
      • Dubin-Thaler B.J.
      • Sheetz M.P.
      • et al.
      Lamellipodial actin mechanically links myosin activity with adhesion-site formation.
      ). These studies were often focused on adherent, focal adhesion-forming cells and integrin-mediated adhesion—now recognized to be dependent on substrate rigidity (
      • Schwartz M.A.
      Integrins and extracellular matrix in mechanotransduction.
      ), and more recently, on substrate-generated friction (
      • Yu C.H.
      • Rafiq N.B.
      • Sheetz M.P.
      • et al.
      Integrin-matrix clusters form podosome-like adhesions in the absence of traction forces.
      ). Concepts and models developed in this context can be linked to T cell morphodynamics via the proposed analogy between receptor organization in T cells into SMACs and actin organization into lamella and lamellipodium (
      • Dustin M.L.
      Hunter to gatherer and back: immunological synapses and kinapses as variations on the theme of amoeboid locomotion.
      ,
      • Sims T.N.
      • Soos T.J.
      • Dustin M.L.
      • et al.
      Opposing effects of PKCtheta and WASp on symmetry breaking and relocation of the immunological synapse.
      ), a scenario vindicated in a recent study of actin organization and dynamics in spreading Jurkat T cells (
      • Yi J.
      • Wu X.S.
      • Hammer 3rd, J.A.
      • et al.
      Actin retrograde flow and actomyosin II arc contraction drive receptor cluster dynamics at the immunological synapse in Jurkat T cells.
      ).
      Despite the mooted central role of the cytoskeleton (
      • Burkhardt J.K.
      • Carrizosa E.
      • Shaffer M.H.
      The actin cytoskeleton in T cell activation.
      ), and the evident importance of ligand diffusion and mobility as discussed previously, a coherent picture linking molecular diffusion and friction to T cell spreading through actin-generated forces is still missing. Here, we address this issue by first conducting experiments comparing the spreading of Jurkat T cells on OKT3 anti-CD3 ligands that are either mobile or immobilized on chemically identical substrates, and then link the experimental results in a general framework of cell spreading through a quantitative model. We focus on the dynamics of the membrane and overall cell morphology during spreading and, after 15 min, on the final adhesion state in terms of membrane adhesion and actin structure. We elucidate the role of the actin-myosin force generating system through selective use of drugs and show that on blocking myosin-II, the cell behavior on mobile and immobilized ligands are similar. Based on these experiments, we propose a physical model that provides a deeper understanding of how ligand-generated friction is linked to spreading.

      Materials and Methods

      Substrates

      Anti-CD3 (Ortho-clone OKT3, Janssen-Cilag, Issy-les-Moulineaux, France) was monobiotinylated according to (
      • Fleire S.J.
      • Batista F.D.
      Studying cell-to-cell interactions: an easy method of tethering ligands on artificial membranes.
      ) and was conjugated to Atto647 according to manufacturers kit instructions (Life Technologies, Saint Aubin, France). The modified antibody, henceforth called α-CD3, was coupled to glass substrates though Neutravidin (Life Technologies), which was either bound to biotin-conjugated bovine serum albumin (BSA-biotin, Sigma, Saint-Quentin Fallavier, France) adsorbed directly on glass (henceforth called Pos) or to SLBs containing 2% 1,2-dipalmitoyl-sn-glycero-3-phosphoethanolamine-N-(cap biotinyl) (All lipids from Avanti Polar Lipids, Coger, Paris, France) dispersed in a matrix of 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC, Avanti Polar Lipids). The SLBs were either deposited from lipid monolayers formed at air/water interface by the Langmuir-Blodgett technique (
      • Fenz S.F.
      • Merkel R.
      • Sengupta K.
      Diffusion and intermembrane distance: case study of avidin and E-cadherin mediated adhesion.
      ), or by the vesicle spreading method (
      • Lin W.-C.
      • Yu C.-H.
      • Groves J.T.
      • et al.
      Supported membrane formation, characterization, functionalization, and patterning for application in biological science and technology.
      ), using a flow chamber with chamber height of 500 μm (FCS2, Bioptechs, Butler, PA). In both cases, the fluidity of the lipids in the bilayer was measured at around 10 μm2/s by continuous photobleaching (CPB) of tracer lipids (C12 Bodipy, Life Technologies) (
      • Fenz S.F.
      • Merkel R.
      • Sengupta K.
      Diffusion and intermembrane distance: case study of avidin and E-cadherin mediated adhesion.
      ) (see Fig. S3 in the Supporting Material). The diffusion of fluorescent anti-CD3 was also measured: CPB shows that, in the vesicle fusion case, α-CD3 is mobile (Fig. S5 A) with a diffusion constant D = 5 μm2/s, henceforth called Mob; conversely, diffusion is not measurable using CPB in the Langmuir-Blodgett case, henceforth called Fix (Fig. S5 B). This last observation is consistent with earlier reports (
      • Ma Z.
      • Janmey P.A.
      • Finkel T.H.
      • et al.
      Improved method of preparation of supported planar lipid bilayers as artificial membranes for antigen presentation.
      ). Single-particle tracking provides on Fix an upper bound for D ≃ 0.001 μm2/s (Fig. S5, C and D). Fix and Mob substrates have exactly the same chemical composition and differ only in the diffusion of α-CD3.
      In certain experiments, the SLBs contained additional lipids—5% 1,2-dioleoyl-sn-glycero-3-[(N-(5-amino-1-carboxypentyl)iminodiacetic acid)succinyl] (nickel salt), which were then bound to recombinant human ICAM1-Fc chimera with histidine tag at the C-terminus (Sinobiological, Beijing, China). Such double functionalized substrates containing both ICAM-Fc and α-CD3 are henceforth called Fix+ICAM or Mob+ICAM in analogy with the previous description. The negative control (without α-CD3) for Pos is henceforth called Neg and for Fix and Mob are called Neg-Fix and Neg-Mob, respectively.
      The relative α-CD3 density on the various substrates was compared by measuring the intensity of fluorescence of Atto647 labeled α-CD3 (see Fig. S1 C). The intensity distribution on the different substrate types was comparable and the differences in the median intensity between types was negligible compared to intrasubstrate spread, showing that all the experiments were performed at a standard α-CD3 surface density. In addition, to facilitate comparison with other works, the absolute α-CD3 density was estimated to be ≃ 20–30 molecules/μm2 (Fig. S2 and accompanying text). Table S1 summarizes the features of the eight different substrates used. Further details on substrate preparation and characterization can be found in the Supporting Material.

      Cell culture, fixation, drug treatment, and labeling

      Jurkat T cells (Clone E6-1, ATCC) were cultured in complete RPMI 1640 medium (Life Technologies) containing red phenol and L-glutamine supplemented with 1% glutamax (Life Technologies) and 10% fetal calf serum (Life Technologies). 1 mL of the medium containing around 0.6 millions cells was removed from the culture flask for experimentation. For live cell imaging at 37°C, cells were centrifuged and resuspended in phosphate buffered saline (PBS) containing 0.1% BSA. To obtain the static data, the cells were allowed to spread for 15 min and were fixed by incubation in 2% paraformaldehyde (Sigma) for 15 min. For drug treatment, the suspended cells were incubated at 37°C and 5% CO2 for 30 min in 50 μM final concentration of the drug (Blebbistatin or CK666, Sigma). Cells were then rapidly washed in PBS as above and observed in PBS + 0.1% BSA + 50 μM of the drug. Cells were fixed after 15 min as above. Particular care was taken to prevent cell detachment at all steps. To visualize the actin cytoskeleton, fixed cells were incubated in 20μg/mL rhodamine-labeled phalloidin (Sigma) in PBS for 45 min at room temperature, and then washed with PBS.

      Microscopy, imaging, and analysis

      Total internal reflection microscopy (TIRFM) and reflection interference contrast microscopy (RICM) were performed using an inverted microscope (AxioObserver, Zeiss, Göttingen, Germany), equipped with an EM-CCD camera (iXon, Andor, Belfast, North-Ireland). Acquisition was performed using Andor iQ software or Micro-Manager (
      • Edelstein A.D.
      • Amodaj N.
      • Stuurman N.
      • et al.
      Computer control of microscopes using MicroManager.
      ). TIRF and RICM images were taken with a 100 × 1.45 NA oil or a custom 100 × 1.46 NA oil antiflex objective (Zeiss). For TIRF exposure time was 1 s and fluorescence filter set adapted to Alexa488 or Rhodamine was used. The atto647 fluorophore was imaged in epifluorescence illumination. To enlarge the field of view, RICM images and time sequences were taken also with a 63x 1.25 NA oil antiflex objective (Zeiss) (
      • Limozin L.
      • Sengupta K.
      Quantitative reflection interference contrast microscopy (RICM) in soft matter and cell adhesion.
      ). Exposure time was 100 to 300 ms and rate 3 to 5 images/s for movies. Confocal images were taken with a Leica confocal microscope equipped with a 63×, NA = 1.4 oil objective. Images were analyzed using Fiji software (
      • Schindelin J.
      • Arganda-Carreras I.
      • Cardona A.
      • et al.
      Fiji: an open-source platform for biological-image analysis.
      ) and Igor Pro (Wavemetrics, Lake Oswego, OR). In RICM, the region of the cell that resides in the vicinity of the substrate (up to a height of a fraction of a micron), and in particular its contour, exhibits a high spatial intensity variance that permits an easy segmentation from the background, which exhibits low variance (
      • Limozin L.
      • Sengupta K.
      Quantitative reflection interference contrast microscopy (RICM) in soft matter and cell adhesion.
      ,
      • Sengupta K.
      • Aranda-Espinoza H.
      • Hammer D.
      • et al.
      Spreading of neutrophils: from activation to migration.
      ). The procedure is illustrated in Fig. S7, AE.
      The gathering of Atto647 labeled α-CD3 under the cell was quantified in terms of an enrichment index and the centralization was quantified in terms of a central-SMAC (cSMAC) number (details in Fig. S7, FH, and corresponding legend).
      The velocity of the cell edge was measured from the time sequence of RICM images as illustrated in Fig. S8. For quantitative comparison between substrates, edge velocities were subsequently pooled over time and cell perimeter and binned in the form of histograms of fractions of events.

      Results

      Ligand mobility determines cell morphology and supramolecular organization

      The final spread area of T cells interacting with different kinds of substrates with immobilized (Pos and Fix) or mobile (Mob) α-CD3, after 15 min of spreading, was quantified from RICM images. The Pos substrates have been extensively used in literature and provided a basis for comparison with other work. Fix and Mob are chemically identical and thus provide an ideal platform to isolate the effect of ligand mobility.
      On Pos, α-CD3 was bound to glass via physisorbed BSA-biotin and neutravidin, and was immobile. Cells spread well and reached a median area of 300 μm2 (mean ± SD: 350 ± 180, 300 cells on 5 independent substrates, Fig. 1). In the literature, final spread area of 400 μm2 was reported (
      • Lam Hui K.
      • Wang C.
      • Upadhyaya A.
      • et al.
      Membrane dynamics correlate with formation of signaling clusters during cell spreading.
      ) on anti-CD3 (Hit3) with about 10 times more surface coverage, and using techniques similar to ours. From older work on similar substrates (
      • Bunnell S.C.
      • Hong D.I.
      • Samelson L.E.
      • et al.
      T cell receptor ligation induces the formation of dynamically regulated signaling assemblies.
      ), an area of ∼200 μm2 can be inferred. We conclude that for Pos, the cell area can vary depending on details of sample preparation. This can be understood in light of the fact that even in the absence of α-CD3 (Neg), a small but detectable adhesion with median area of ∼130 μm2 was observed (Fig. 1), arising from nonspecific interactions of physical origin with glass. Such residual adhesion has also been reported for B cells spreading on similar substrates (
      • Ketchum C.
      • Miller H.
      • Upadhyaya A.
      • et al.
      Ligand mobility regulates B cell receptor clustering and signaling activation.
      ).
      Figure thumbnail gr1
      Figure 1Ligand mobility determines the T cell spread area and actin distribution. Row 1: Scheme of the functionalization of the various substrates with monobiotin fluorescent anti-CD3 (α-CD3) coupled via Neutravidin (Nav). Neg: no α-CD3 on glass. Pos: immobilized α-CD3 on glass. Fix: immobilized α-CD3 on SLB. Mob: mobile α-CD3 on SLB. Row 2: RIC Micrographs of Jurkat T cells after 15 min engagement on substrates Neg, Pos, Fix, Mob. NB: no adhesion was measurable on supported bilayers in the absence of α-CD3 (Fix-Neg and Mob-Neg). Row 3: Corresponding fluorescence micrographs of labeled α-CD3. NA: not applicable. Row 4: TIRF micrographs after labeling actin with Rhodamine Phalloidin. Scale bars: 5μm. Row 5: Scatter dot plot of cell adhesion area, anti-CD3 enrichment index and cSMAC number at 15 min after seeding. Enrichment index represents the recruitment of ligands in the adhesion zone. cSMAC number shows the centralization of α-CD3 in the middle of the adhesion zone. Black horizontal lines represent the median of each distribution. To see this figure in color, go online.
      The F-actin distribution on Pos, as imaged with TIRFM and in agreement with the literature (
      • Bunnell S.C.
      • Hong D.I.
      • Samelson L.E.
      • et al.
      T cell receptor ligation induces the formation of dynamically regulated signaling assemblies.
      ), exhibits a strong enrichment at the periphery of the contact zone, concomitant with a depletion at the center (Fig. 1). Comparison of RICM and TIRF images suggests that the peripheral zone exhibits structures that are reminiscent of lamellipodia and filopodia and confocal images show that the cells have a fried egg shape (Fig. S9). On Neg, the distribution of F-actin is fairly uniform, and confocal images confirm that it corresponds to the homogeneous cortical actin of a rounded, nonactivated cell (Fig. S9).
      On Mob, α-CD3 was bound to a SLB via a biotinylated lipid and neutravidin, and was free to diffuse in the plane of the bilayer. Here, in absence of α-CD3 (Mob-Neg), no adhesion was detected. On Mob, the measured median area was ∼90 μm2 (mean ± SD: 100 ± 40, 202 cells on seven independent substrates, Fig. 1). This value is in fact a slight overestimation because some of the very weakly adhered cells are probably washed away during the fixation process. This value compares very well with the only other report of T cells adhering to SLBs carrying α-CD3 (and no ICAM)—we infer a spread area of ∼80 μm2 (from images of fluorescent actin because no area or RICM data are provided) (
      • Hsu C.-J.
      • Hsieh W.-T.
      • Baumgart T.
      • et al.
      Ligand mobility modulates immunological synapse formation and T cell activation.
      ).
      As can be seen from Fig. 1, the F-actin distribution on Mob is central, in the form of an inhomogeneous disc, with strongly fluorescent structures evoking filopodia visible both with RICM and TIRF (Fig. 1 and Fig. S9). Confocal images show that on Mob, as in Neg, the cells keep a rounded shape (Fig. S9).
      On Fix, α-CD3 was bound to a SLB via a biotinylated lipid and neutravidin, but was immobilized. Being chemically identical to Mob, Fix provided an ideal substrate to estimate the influence of mobility without interference from any other factors. Furthermore, in the absence of α-CD3 (Fix-Neg), no adhesion was detected. On Fix, the cells spread less compared to functionalized glass (Pos) and after 15 min of spreading attained a median area of 184 μm2 (mean ± SD: 200 ± 80, 178 cells on 3 independent substrates, Fig. 1). This type of substrates have not been previously reported in the literature, because the Pos substrate is usually used as the standard for immobilized ligands. The significant difference in cell adhesion area between Pos and Fix emphasizes the absence of contribution from nonspecific adhesion in Fix. The distribution of F-actin, as well as the overall cell shape on Fix is similar to that on Pos (Fig. 1 and Fig. S9). These results together show that on chemically identical substrates, cell spreading is strongly attenuated, and actin distribution is altered when ligands are mobile.
      As expected, on Mob substrates the ligands were enriched under the cells with a median enrichment index of 1.8 on Mob (Fig. 1, row 5). In comparison, median enrichment indices on Fix and Pos were 1.2 and 1.4, respectively (these values differ from 1 because of the feeble but detectable autofluorescence of the cells. It was checked on Fix substrate that using TIRF instead of epifluorescence for Atto647 imaging reduces the median value of the enrichment index from 1.2 to 1.05). The median of cSMAC number, characterizing the centralization of the ligands, increases from 1.4 (Pos) to 2.7 (Mob) (Fig. 1, row 5), revealing an α-CD3 centralization reminiscent of the cSMAC structure observed in the classical synapse (
      • Grakoui A.
      • Bromley S.K.
      • Dustin M.L.
      • et al.
      The immunological synapse: a molecular machine controlling T cell activation.
      ). The ligand enrichment and the central organization in the form of cSMAC in Mob are attributed to the mobility of the ligands. All the area and enrichment data are summarized in Table S2.
      The level of cell activation was also compared between the different substrates by measuring the surface density of phosphorylated Zap by using a fluorescent antibody, as shown in Fig. S10. It is low on Neg substrates, but high and similar on all α-CD3-coated substrates.

      Effect of myosin-II and Arp2/3 inhibition

      The role of cell-generated forces on cell spreading and actin organization was explored by inhibition of myosin-II by blebbistatin and/or inhibition of Arp2/3 by CK666. Blebbistatin is known to reduce binding of myosin-II to actin, therefore reducing cell contractility. Area of spreading was not appreciably modified by the drug on Neg or Fix; the area was slightly decreased on Pos, as observed previously (
      • Lam Hui K.
      • Wang C.
      • Upadhyaya A.
      • et al.
      Membrane dynamics correlate with formation of signaling clusters during cell spreading.
      ). Remarkably, on Mob, the area was strongly enhanced in presence of the drug, showing that reduction in contractility facilitates spreading on mobile ligands (Fig. 2). Enrichment of α-CD3 was still observed, but the enrichment index was reduced to 1.5 (compared to 1.8 in the absence of the drug). Additionally, α-CD3 distribution under the cell was centralized, as quantified by cSMAC number of 2.2, suggesting that active mechanisms were still present and gathered α-CD3. Interestingly, on Mob, the amount of ligands collected under the cell, as measured by an integrated enrichment, does not change upon blebbistatin addition (see Fig. S11), which could suggest that myosin-II may not regulate anti-CD3 binding. On Pos and Fix, the actin distribution at the cell scale was similar with or without drug, exhibiting a peripheral enrichment. However, strikingly, on Mob the actin distribution is peripheral in the presence of the drug, whereas it was centrally enriched in the absence of the drug. Therefore, we conclude that inhibition of contractile forces restores peripheral actin organization, which was abrogated on mobile ligands in the absence of the drug (see Fig. 2).
      Figure thumbnail gr2
      Figure 2Combined Effects of myosin-II inhibition and ligand mobility on cell spreading area and actin organization. Jurkat T cells were pretreated with 50 μM blebbistatin and engaged on various substrates for 15 min. Row 1: RICMicrographs of blebbistatin-treated cells on substrates Neg, Pos, Fix, Mob, corresponding fluorescence micrograph of labeled α-CD3 on Mob. Scale Bar: 5 μm. Row 2: TIR Fluorescence micrographs of Rhodamine Phalloidin labeled actin on each substrate. Corresponding RICM is shown as inset. Row 3: Scatter dot plot of spreading area, ligand enrichment, and cSMAC number, at 15 min with and without blebbistatin. Black horizontal lines represent the median of each distribution. Fluorescence intensity (A.U.) of actin radial distribution averaged from at least 5 individual cells imaged in TIRF (mean ± SD) for Fix and Mob, with or without blebbistatin treatment. To see this figure in color, go online.
      The small molecule CK666 is known to stabilize the inactive state of the Arp2/3 complex (
      • Nolen B.J.
      • Tomasevic N.
      • Pollard T.D.
      • et al.
      Characterization of two classes of small molecule inhibitors of Arp2/3 complex.
      ), which is an actin nucleator essential for the formation of lamellipodia. CK666 was used to inhibit the formation of lamellipodia during spreading. On Pos, the spreading area was strongly reduced by the addition of CK666 (Fig. S11, left), suggesting that lamellipodia are one of the major driving mechanisms for spreading. Treatment with a mixture of CK666 and blebbistatin restored the spreading area to the large values observed in the absence of the drugs, again emphasizing the role of myosin-generated contractile forces. However, RICM observations also revealed that the cell-substrate contact zone exhibits numerous white areas, indicating defective adhesion (Fig. S11, right).

      Effect of ICAM ligands on spreading

      Doubly functionalized substrates with mobile ligands (α-CD3 and ICAM1) have been widely studied (
      • Kaizuka Y.
      • Douglass A.D.
      • Vale R.D.
      • et al.
      Mechanisms for segregating T cell receptor and adhesion molecules during immunological synapse formation in Jurkat T cells.
      ). We therefore compare Fix and Mob substrates featuring α-CD3 alone with Fix+ICAM and Mob+ICAM—the corresponding cases with ICAM present (Fig. S12). In agreement with the literature, the cells spread well on Mob+ICAM (median area ∼300 μm2, compared to ∼100 μm2 for Mob) and exhibit peripheral actin. As expected, α-CD3 is enriched under the cell with a cSMAC number of 3.2 (compared to 3.0 on α-CD3 alone). On Fix+ICAM, the median spreading area reaches 400 μm2 (compared to 184 μm2 on Fix). All the area and enrichment data are summarized in Table S2. Overall, inclusion of ICAM restores the area on mobile ligands to values we report for immobilized α-CD3. The spreading on SLBs functionalized with ICAM alone was very limited (median 30 μm2), in agreement with the expectations of limited engagement of integrins in the absence of activation through TCR.

      Dynamics at the leading edge

      The dynamics of the leading edge of spreading T cells on the various substrates was explored by quantitative analysis of RICM image sequences recorded in the course of spreading (
      • Limozin L.
      • Sengupta K.
      Quantitative reflection interference contrast microscopy (RICM) in soft matter and cell adhesion.
      ,
      • Sengupta K.
      • Limozin L.
      Adhesion of soft membranes controlled by tension and interfacial polymers.
      ), with or without blebbistatin treatment. In our analysis, the cell edge is defined as the boundary of the contact zone between the cell membrane and the substrate. In addition to the regions of tight contact, this also contains the regions where the cell membrane is extended over the substrate and, even if not fully bound, is potentially able to interact with it.
      Consistent with the area measured after 15 min of spreading, on Pos and Fix, as well as on Mob with blebbistatin, cells exhibit global spreading, whereas on Mob, globally the cells spread very little beyond an initial deformation and alignment of the proximal surface with the substrate—the cell edge however undergoes continuous extension and retraction (see Movie). This is evidenced in the representation of the radius as a function of time and space, as in (
      • Gauthier N.C.
      • Fardin M.A.
      • Sheetz M.P.
      • et al.
      Temporary increase in plasma membrane tension coordinates the activation of exocytosis and contraction during cell spreading.
      ) (curvilinear coordinates around the contour, Figs. S14–S16). It is seen that on Pos, Fix, and Mob with blebbistatin, a growth phase, during which the radius grows regularly on the average, is followed by a steady phase where the radial growth is saturated but the edge nevertheless continues to be dynamic (see Fig. S14, Fig. S16, and corresponding Movies). By convention, we defined the steady phase as starting when 80% of the final area is reached. On Pos, the growth occurs on the scale of minutes as reported previously (
      • Cretel E.
      • Touchard D.
      • Pierres A.
      • et al.
      A new method for rapid detection of T lymphocyte decision to proliferate after encountering activating surfaces.
      ,
      • Lam Hui K.
      • Wang C.
      • Upadhyaya A.
      • et al.
      Membrane dynamics correlate with formation of signaling clusters during cell spreading.
      ,
      • Brodovitch A.
      • Bongrand P.
      • Pierres A.
      T lymphocytes sense antigens within seconds and make a decision within one minute.
      ). The edge velocity, calculated with a time resolution of 2 s is presented in the form of a velocity map (
      • Lam Hui K.
      • Kwak S.I.
      • Upadhyaya A.
      Adhesion-dependent modulation of actin dynamics in Jurkat T cells.
      ,
      • Sims T.N.
      • Soos T.J.
      • Dustin M.L.
      • et al.
      Opposing effects of PKCtheta and WASp on symmetry breaking and relocation of the immunological synapse.
      ) (Figs. S14–S16), where the occurrence of extensions/retractions of the edge can be seen as red (v > 0)/blue (v < 0) regions. The same data are also presented in the form of velocity histograms (Fig. 3).
      Figure thumbnail gr3
      Figure 3Effect of ligand mobility and myosin inhibition on the velocity of the leading edge. Histograms count the fraction of events in a particular velocity bin. Lines joining points are drawn as a guide to the eye. Top row compares velocities on each substrate, in the absence of blebbistatin, either during the growth phase of spreading (A) or during the steady phase (B). Bottom row shows velocities for different conditions (growth or steady phase, with or without blebbistatin), either on Pos (C) or Mob (D). Thick lines correspond to empirical fits described in the . To see this figure in color, go online.
      Strikingly, as already noted for the velocity maps, both positive and negative velocities are observed during the growth phase (Fig. 3 A) as well as during the steady phase (Fig. 3 B). During the growth phase, velocity distribution is quasi-symmetrical around zero on mobile ligands, although it is asymmetric on fixed ligands, with a most probable velocity at around 0.05 μm/s (Fig. 3 A). Addition of blebbistatin does not significantly change the negative velocities on any substrate and slightly reduces the positive velocities on fixed ligands (Fig. 3, C and D), as already observed in (
      • Lam Hui K.
      • Wang C.
      • Upadhyaya A.
      • et al.
      Membrane dynamics correlate with formation of signaling clusters during cell spreading.
      ). However, on mobile ligands, blebbistatin significantly increases the small positive velocities. Overall, treatment with blebbistatin has a stronger impact on the advancing than on the retracting events. On Mob+ICAM, the velocity distribution is strikingly different from Mob and closely resembles that on Fix or Pos (Fig. S12).

      Discussion

      The dynamics as well as static data presented here can be interpreted in terms of a model that links substrate friction to the dynamics of the cell edge. Consider a one-dimensional model of the cell edge, which moves with radial velocity ve parallel to the substrate (Fig. 4 A). The main driving force for spreading arises from lamellipodia-like protrusions mediated by actin polymerization and branching. This is in accordance with the observation that Arp2/3 inhibition prevents spreading on all types of substrates (Fig. S11, left), and was also mooted in earlier work (
      • Babich A.
      • Li S.
      • Burkhardt J.K.
      • et al.
      F-actin polymerization and retrograde flow drive sustained PLCγ1 signaling during T cell activation.
      ). In analogy with neuronal growth cones (
      • Craig E.M.
      • Van Goor D.
      • Mogilner A.
      • et al.
      Membrane tension, myosin force, and actin turnover maintain actin treadmill in the nerve growth cone.
      ) and filopodia (
      • Bornschlögl T.
      • Romero S.
      • Bassereau P.
      • et al.
      Filopodial retraction force is generated by cortical actin dynamics and controlled by reversible tethering at the tip.
      ), the local edge velocity is assumed to be imposed by the dynamics of the underlying actin, which is itself the result of polymerization at velocity vp (this may include potential actin disassembly which we shall not detail further (
      • Craig E.M.
      • Van Goor D.
      • Mogilner A.
      • et al.
      Membrane tension, myosin force, and actin turnover maintain actin treadmill in the nerve growth cone.
      )) and a retrograde actin flow at velocity va: ve = vpva. Physically, this means that if there is no retrograde flow, the cell edge would advance at the same rate as the growth of the tip of the actin.
      Figure thumbnail gr4
      Figure 4Frictional coupling model for cell edge velocity. (A) The edge velocity ve is imposed by the growing tip of actin, which is itself governed by polymerization at velocity vp and the actin retrograde flow at velocity va. The retrograde flow of actin arises due to actomyosin contraction and membrane tension. TC Receptors are put in motion by a frictional coupling with actin (Friction a/r), whereas the membrane (Friction m/r) and the substrate resists their motion (Friction s/r). Friction with the substrate occurs through anchoring of the ligands that are in turn bound to the receptors (α-CD3 ligands are not explicitly represented). (B) Model prediction of cell edge velocity for variable α-CD3 density n and diffusion on the substrate Ds/r. The only fitted parameter is the friction between the actin and the TCR: ζa/TCRFix1pN.s/μm. Other parameters were taken from the literature (see text and for the details of the model). The thick black dots indicate the experimental conditions realized for substrates Fix and Mob. (C) Model prediction of cell final spread area, assuming a linear dependence between tension and spreading area T = gA with a coefficient g = 0.5 pN.sm3. To see this figure in color, go online.
      The velocity of the retrograde actin flow is set by the balance between a tension term T (which includes the actomyosin contraction and membrane tension—in line with recent reports of cellular traction on α-CD3 bonds (
      • Bashour K.T.
      • Gondarenko A.
      • Kam L.C.
      • et al.
      CD28 and CD3 have complementary roles in T-cell traction forces.
      ,
      • Brodovitch A.
      • Bongrand P.
      • Pierres A.
      T lymphocytes sense antigens within seconds and make a decision within one minute.
      )), and the friction Fr/a between the actin and all the receptors. Such a frictional coupling was identified previously as a major player in receptor and cytoskeleton motion in T cells (
      • Mossman K.D.
      • Campi G.
      • Dustin M.L.
      • et al.
      Altered TCR signaling from geometrically repatterned immunological synapses.
      ,
      • Yu C.H.
      • Wu H.-J.
      • Groves J.T.
      • et al.
      Altered actin centripetal retrograde flow in physically restricted immunological synapses.
      ). As for any frictional dissipation, Fr/a is the product of the friction coefficient for receptor to actin coupling, and the relative velocity between the actin and the receptors. All the receptors are of course subjected to an equal and opposite force that drags them backward away from the cell edge. In addition, they are subject to friction with the membrane, assumed to be immobile. Receptors that are bound to a ligand are subjected to an additional friction with the substrate, which tends to resist receptor motion. The force balance on the actin and on the receptors leads to equations for the edge velocity in terms of the actin polymerization rate, the membrane tension, the number of bound and unbound ligands and the various frictional coefficients (see the Supporting Material for detailed hypotheses and equations).
      These considerations are sufficient to qualitatively understand the dynamics data. Simple geometrical considerations imply that if the rate of actin polymerization is constant, the velocity of the cell edge decreases when the actin retrograde flow is high and increases when it is low. At a constant low tension, the retrograde flow is regulated entirely by the various frictional couplings. If the anchorage to the substrate is low, the receptor is advected with almost the same velocity as the actin without much influence on the actin velocity; if the anchorage is high and there are many receptors, the resistance to dragging slows down the actin itself. The former case corresponds to ligands being mobile and results in slower edge velocity, whereas the latter corresponds to immobilized receptors and results in larger edge velocity—in qualitative agreement with our observations.
      To achieve quantitative comparison, the values for the parameters in the model are either taken from the literature (actin polymerization rate (
      • Lam Hui K.
      • Kwak S.I.
      • Upadhyaya A.
      Adhesion-dependent modulation of actin dynamics in Jurkat T cells.
      ,
      • Bornschlögl T.
      • Romero S.
      • Bassereau P.
      • et al.
      Filopodial retraction force is generated by cortical actin dynamics and controlled by reversible tethering at the tip.
      ), typical tension in leukocytes (
      • Lomakina E.B.
      • Waugh R.E.
      Micromechanical tests of adhesion dynamics between neutrophils and immobilized ICAM-1.
      ), receptor diffusion (
      • Dushek O.
      • Mueller S.
      • Valitutti S.
      • et al.
      Effects of intracellular calcium and actin cytoskeleton on TCR mobility measured by fluorescence recovery.
      )) or are measured for our experimental conditions. The only remaining free parameter is the frictional coupling between the receptors and the actin, which could be estimated to be 1 pN.s/μm. In a model of stochastic sliding friction (
      • Sens P.
      Rigidity sensing by stochastic sliding friction.
      ) this would correspond to having a stiffness of the molecular complex of 0.01 pN/nm and an off-rate of 10 s−1.
      The calculated dependence of the edge velocity on the density and diffusion of the ligands is shown in Fig. 4 B. At low ligand diffusion, as expected intuitively, increasing ligand density increases ve. However, at high ligand diffusion, the edge velocity is hardly affected by ligand density. At all ligand densities, ve decreases with increasing ligand diffusion. To fit the edge velocity histograms, ve is assumed to be an oscillatory function (oscillation can arise from actin polymerization or from the retrograde flow). Such a treatment fits the experimental histograms very well (Fig. 3, Fig. S16, and Table S3).
      The local dynamics of the leading edge cannot, only by itself, account for the final spread area, which is probably limited by the global tension that increases as the cell spreads. We assumed a linear relation between the tension and the area to make connection with the static data. The results show that reducing the actin retrograde flow, using either blebbistatin or fixed ligands, leads to an increase of the spreading time and of the maximal area. The calculated dependence of final area on the density and diffusion of the ligands is given in Fig. 4 C. Again, as expected, at low ligand diffusion, area increases with ligand density. However, on diffusing ligands, the area is virtually independent of ligand density. The crossover from Fix-like to Mob-like occurs around the point where the ligand diffusion on the substrate falls below the diffusion of receptors on the cell membrane. At a constant ligand density, above this crossover threshold, the area strongly decreases with increasing ligand diffusion.
      For simplicity, we have not explicitly introduced the membrane or bulk dissipation (the latter was introduced in the literature in (
      • Cuvelier D.
      • Théry M.
      • Mahadevan L.
      • et al.
      The universal dynamics of cell spreading.
      ) to explain global features of spreading data). The adhesion energy provided by ligand-receptor binding was neglected according to our result with Arp2/3 inhibition. However, when both Arp2/3 and myosin-II are inhibited, the terms vp as well as T are reduced simultaneously. The residual driving force for the spreading observed in Fig. S11 (right panel) may stem from Arp2/3 independent polymerization. Additionally, the adhesion energy of the ligand/receptor bonds may begin to play a role, in analogy with liposomes studies (
      • Fenz S.F.
      • Merkel R.
      • Sengupta K.
      Diffusion and intermembrane distance: case study of avidin and E-cadherin mediated adhesion.
      ,
      • Sengupta K.
      • Limozin L.
      Adhesion of soft membranes controlled by tension and interfacial polymers.
      ,
      • Smith A.-S.
      • Sengupta K.
      • Sackmann E.
      • et al.
      Force-induced growth of adhesion domains is controlled by receptor mobility.
      ).
      The model, which was so far used to discuss edge dynamics and global spreading, can also account for receptor accumulation and centralization in analogy with a purely passive system of integrin-mediated adhesion of a liposome (
      • Smith A.-S.
      • Sengupta K.
      • Sackmann E.
      • et al.
      Force-induced growth of adhesion domains is controlled by receptor mobility.
      ), where on application of a pulling force, the individual adhesion domains either resist the force (immobile) or are compacted and pulled inward (mobile). Here, the retrograde actin flow similarly results in the inward motion of mobile receptors leading to their accumulation and centralization.
      The one-dimensional model developed here is by its nature not suitable for explaining the lifting of the peripheral membrane away from the substrate, frequently observed on Mob but not on Fix or Pos (see arrow on RICM sequence of Fig. S15). To explain this phenomenon, we hypothesize that T may have a component perpendicular to the substrate. On Fix and Pos, the immobile receptors effectively pin the membrane in place and oppose the lifting force. However, on Mob, the receptors slide backward in response to the pulling (
      • Smith A.-S.
      • Sengupta K.
      • Sackmann E.
      • et al.
      Force-induced growth of adhesion domains is controlled by receptor mobility.
      ), and in absence of a stable anchor, the cell edge is lifted off the plane of the substrate.
      The classical immune synapse was described on SLBs with mobile ICAM1 and pMHC (
      • Fooksman D.R.
      • Vardhana S.
      • Dustin M.L.
      • et al.
      Functional anatomy of T cell activation and synapse formation.
      ,
      • Grakoui A.
      • Bromley S.K.
      • Dustin M.L.
      • et al.
      The immunological synapse: a molecular machine controlling T cell activation.
      ). Integrin LFA-1 binding to ICAM1 provides adhesion between the T cell and the substrate and increases the sensitivity to antigen by a factor 100 (
      • Springer T.A.
      • Dustin M.L.
      Integrin inside-out signaling and the immunological synapse.
      ). In systems with α-CD3 and ICAM1 on SLB, spreading and cluster motion comparable to the classical case has been reported (
      • Kaizuka Y.
      • Douglass A.D.
      • Vale R.D.
      • et al.
      Mechanisms for segregating T cell receptor and adhesion molecules during immunological synapse formation in Jurkat T cells.
      ). In accordance with these previous works, we saw that on additional grafting of ICAM1 on the α-CD3-coated substrates, cells could spread well even on mobile ligands. Spread area is further increased on immobile, doubly functionalized substrate, but in a limited manner (see Table S2). In addition, on dually functionalized mobile bilayers, Kumari et al. (
      • Kumari S.
      • Vardhana S.
      • Dustin M.L.
      • et al.
      T lymphocyte myosin IIA is required for maturation of the immunological synapse.
      ) observed a limited increase of area induced by blebbistatin. All these results show that the presence of integrins may mask the dramatic effect of ligand mobility seen with α-CD3 alone.
      Interestingly, Fig. 4 C shows that the difference in area on Mob in the presence of integrins ligands cannot be explained simply by an increase in ligand density. Indeed, our model accounts quantitatively for these observations by supposing an increase in the friction between actin and LFA-1 upon binding its ligand ICAM. We estimate this increases to 0.14 pN.s/μm (see the Supporting Material).
      The present case, in the absence of integrin ligands, represents an interesting situation where nonintegrin mechanotransduction may occur (
      • Bashour K.T.
      • Gondarenko A.
      • Kam L.C.
      • et al.
      CD28 and CD3 have complementary roles in T-cell traction forces.
      ,
      • Yu Y.
      • Smoligovets A.A.
      • Groves J.T.
      Modulation of T cell signaling by the actin cytoskeleton.
      ). Beyond the mechanistic considerations presented previously, the indirect role of force may occur through a change in protein organization, which then may have a direct impact on activation. Whether impeded translocation of clusters increases or decreases signaling is a debated question. Our experiments on phosphorylation of Zap-70 reveal no differences on the basis of ligand mobility but cannot entirely rule out differences in other markers of activation. Our experiments and model suggest that the impact of ligand mobility on spreading, can be understood on the basis of ligand-mediated friction in analogy with earlier models of integrin based migration models.

      Conclusion

      Our results show that the ability of ligand anchors to either withstand pulling forces or to respond by sliding along the cell membrane is central to the way a cell adheres, spreads, and organizes its cytoskeleton. Although actin polymerization provides the main driving force, the spreading dynamics is determined by a combination of actomyosin contraction and membrane tension on one hand, and frictional coupling of the receptors to the actin retrograde flow, the membrane as well as the substrate (via the ligands to which they are bound) on the other hand. A comparison of our experiments lacking a ligand for integrins with experiments involving either T cells interacting with antigen plus ICAM on SLBs (
      • Kumari S.
      • Vardhana S.
      • Dustin M.L.
      • et al.
      T lymphocyte myosin IIA is required for maturation of the immunological synapse.
      ) or fibroblasts on mobile RGDs (
      • Yu C.H.
      • Rafiq N.B.
      • Sheetz M.P.
      • et al.
      Integrin-matrix clusters form podosome-like adhesions in the absence of traction forces.
      ), point to the fact that integrins can partially override the consequences of sliding anchors on overall cell spreading. In the absence of integrins ligands, and on mobile ligands, cells can be made to spread by blocking myosin—probably by releasing the cortical tension. These results together point to two different possible mechanisms that can abrogate the lack of stable anchoring to proceed with spreading—generating enough friction to stabilize anchoring or relaxation of actomyosin contraction and membrane tension, facilitating actin-induced protrusions that then lead to cell spreading without pulling against an anchor. Our simple experimental system allowed us to pinpoint the major players that influence T cell spreading on diffusing ligands, and thus to build a quantitative model linking molecular diffusion to cell spreading through actin-generated forces.
      We thank A. M. Lellouch, S. Cadra, M. Biarnes for assistance with cell culture and labeling, D. Tribouillois for development in ImageJ, P. Bongrand, A. Pierres, A. M. Lellouch, P. H. Puech, and M. Rao for stimulating discussions, P. H. Puech for careful reading of the manuscript. L. L. thanks Labex Inform for providing stimulating discussion forums.
      K. S. thanks European Research Council for funding via grant No. 307104 FP/2007-2013/ERC. This work was in part supported by the intramural research program of National Institute of Allergy and Infectious Diseases (NIAID), National Institutes of Health (NIH).

      Supporting Material

      • Loading ...

      Supporting Citations

      References (
      • Robert P.
      • Nicolas A.
      • Limozin L.
      • et al.
      Minimal encounter time and separation determine ligand-receptor binding in cell adhesion.
      ,
      • Dedecker P.
      • Duwé S.
      • Zhang J.
      • et al.
      Localizer: fast, accurate, open-source, and modular software package for superresolution microscopy.
      ) appear in the Supporting Material.

      References

        • Fooksman D.R.
        • Vardhana S.
        • Dustin M.L.
        • et al.
        Functional anatomy of T cell activation and synapse formation.
        Annu. Rev. Immunol. 2010; 28: 79-105
        • Crites T.J.
        • Padhan K.
        • Varma R.
        • et al.
        TCR Microclusters pre-exist and contain molecules necessary for TCR signal transduction.
        J. Immunol. 2014; 193: 56-67
        • Cretel E.
        • Touchard D.
        • Pierres A.
        • et al.
        A new method for rapid detection of T lymphocyte decision to proliferate after encountering activating surfaces.
        J. Immunol. Methods. 2011; 364: 33-39
        • Dustin M.L.
        Hunter to gatherer and back: immunological synapses and kinapses as variations on the theme of amoeboid locomotion.
        Annu. Rev. Cell Dev. Biol. 2008; 24: 577-596
        • Chan P.-Y.
        • Lawrence M.B.
        • Springer T.A.
        • et al.
        Influence of receptor lateral mobility on adhesion strengthening between membranes containing LFA-3 and CD2.
        J. Cell Biol. 1991; 115: 245-255
        • Dustin M.L.
        • Ferguson L.M.
        • Golan D.E.
        • et al.
        Visualization of CD2 interaction with LFA-3 and determination of the two-dimensional dissociation constant for adhesion receptors in a contact area.
        J. Cell Biol. 1996; 132: 465-474
        • Fenz S.F.
        • Merkel R.
        • Sengupta K.
        Diffusion and intermembrane distance: case study of avidin and E-cadherin mediated adhesion.
        Langmuir. 2009; 25: 1074-1085
        • Grakoui A.
        • Bromley S.K.
        • Dustin M.L.
        • et al.
        The immunological synapse: a molecular machine controlling T cell activation.
        Science. 1999; 285: 221-227
        • Varma R.
        • Campi G.
        • Dustin M.L.
        • et al.
        T cell receptor-proximal signals are sustained in peripheral microclusters and terminated in the central supramolecular activation cluster.
        Immunity. 2006; 25: 117-127
        • Kaizuka Y.
        • Douglass A.D.
        • Vale R.D.
        • et al.
        Mechanisms for segregating T cell receptor and adhesion molecules during immunological synapse formation in Jurkat T cells.
        Proc. Natl. Acad. Sci. USA. 2007; 104: 20296-20301
        • Hartman N.C.
        • Nye J.A.
        • Groves J.T.
        Cluster size regulates protein sorting in the immunological synapse.
        Proc. Natl. Acad. Sci. USA. 2009; 106: 12729-12734
        • Mossman K.D.
        • Campi G.
        • Dustin M.L.
        • et al.
        Altered TCR signaling from geometrically repatterned immunological synapses.
        Science. 2005; 310: 1191-1193
        • Yu C.H.
        • Wu H.-J.
        • Groves J.T.
        • et al.
        Altered actin centripetal retrograde flow in physically restricted immunological synapses.
        PLoS ONE. 2010; 5: e11878
        • Bunnell S.C.
        • Hong D.I.
        • Samelson L.E.
        • et al.
        T cell receptor ligation induces the formation of dynamically regulated signaling assemblies.
        J. Cell Biol. 2002; 158: 1263-1275
        • Bunnell S.C.
        • Kapoor V.
        • Samelson L.E.
        • et al.
        Dynamic actin polymerization drives T cell receptor-induced spreading: a role for the signal transduction adaptor LAT.
        Immunity. 2001; 14: 315-329
        • Bunnell S.C.
        Multiple microclusters: diverse compartments within the immune synapse.
        in: Immunological Synapse. Vol. 340. Springer-Verlag, Berlin Heidelberg, 2010: 123-154
        • Hsu C.-J.
        • Hsieh W.-T.
        • Baumgart T.
        • et al.
        Ligand mobility modulates immunological synapse formation and T cell activation.
        PLoS ONE. 2012; 7: e32398
        • Segura J.-M.
        • Guillaume P.
        • Luescher I.F.
        • et al.
        Increased mobility of major histocompatibility complex I-peptide complexes decreases the sensitivity of antigen recognition.
        J. Biol. Chem. 2008; 283: 24254-24263
        • Ma Z.
        • Sharp K.A.
        • Finkel T.H.
        • et al.
        Surface-anchored monomeric agonist pMHCs alone trigger TCR with high sensitivity.
        PLoS Biol. 2008; 6: e43
        • Luxembourg A.T.
        • Brunmark A.
        • Cai Z.
        • et al.
        Requirements for stimulating naive CD8+ T cells via signal 1 alone.
        J. Immunol. 1998; 161: 5226-5235
        • Judokusumo E.
        • Tabdanov E.
        • Kam L.C.
        • et al.
        Mechanosensing in T lymphocyte activation.
        Biophys. J. 2012; 102: L5-L7
        • Bashour K.T.
        • Gondarenko A.
        • Kam L.C.
        • et al.
        CD28 and CD3 have complementary roles in T-cell traction forces.
        Proc. Natl. Acad. Sci. USA. 2014; 111: 2241-2246
        • Lam Hui K.
        • Kwak S.I.
        • Upadhyaya A.
        Adhesion-dependent modulation of actin dynamics in Jurkat T cells.
        Cytoskeleton (Hoboken). 2014; 71: 119-135
        • Cuvelier D.
        • Théry M.
        • Mahadevan L.
        • et al.
        The universal dynamics of cell spreading.
        Curr. Biol. 2007; 17: 694-699
        • Jurado C.
        • Haserick J.R.
        • Lee J.
        Slipping or gripping? Fluorescent speckle microscopy in fish keratocytes reveals two different mechanisms for generating a retrograde flow of actin.
        Mol. Biol. Cell. 2005; 16: 507-518
        • Craig E.M.
        • Van Goor D.
        • Mogilner A.
        • et al.
        Membrane tension, myosin force, and actin turnover maintain actin treadmill in the nerve growth cone.
        Biophys. J. 2012; 102: 1503-1513
        • Giannone G.
        • Dubin-Thaler B.J.
        • Sheetz M.P.
        • et al.
        Lamellipodial actin mechanically links myosin activity with adhesion-site formation.
        Cell. 2007; 128: 561-575
        • Schwartz M.A.
        Integrins and extracellular matrix in mechanotransduction.
        Cold Spring Harb. Perspect. Biol. 2010; 2: a005066
        • Yu C.H.
        • Rafiq N.B.
        • Sheetz M.P.
        • et al.
        Integrin-matrix clusters form podosome-like adhesions in the absence of traction forces.
        Cell Reports. 2013; 5: 1456-1468
        • Sims T.N.
        • Soos T.J.
        • Dustin M.L.
        • et al.
        Opposing effects of PKCtheta and WASp on symmetry breaking and relocation of the immunological synapse.
        Cell. 2007; 129: 773-785
        • Yi J.
        • Wu X.S.
        • Hammer 3rd, J.A.
        • et al.
        Actin retrograde flow and actomyosin II arc contraction drive receptor cluster dynamics at the immunological synapse in Jurkat T cells.
        Mol. Biol. Cell. 2012; 23: 834-852
        • Burkhardt J.K.
        • Carrizosa E.
        • Shaffer M.H.
        The actin cytoskeleton in T cell activation.
        Annu. Rev. Immunol. 2008; 26: 233-259
        • Fleire S.J.
        • Batista F.D.
        Studying cell-to-cell interactions: an easy method of tethering ligands on artificial membranes.
        Methods Mol. Biol. 2009; 462: 145-154
        • Lin W.-C.
        • Yu C.-H.
        • Groves J.T.
        • et al.
        Supported membrane formation, characterization, functionalization, and patterning for application in biological science and technology.
        in: Arkin A.P. Mahal L. Romesberg F. Shah K. Shamu C. Thomas C. Current Protocols in Chemical Biology. John Wiley & Sons, Hoboken, NJ2010: 235-269
        • Ma Z.
        • Janmey P.A.
        • Finkel T.H.
        • et al.
        Improved method of preparation of supported planar lipid bilayers as artificial membranes for antigen presentation.
        Microsc. Res. Tech. 2011; 74: 1174-1185
        • Edelstein A.D.
        • Amodaj N.
        • Stuurman N.
        • et al.
        Computer control of microscopes using MicroManager.
        in: Ausubel F.M. Current Protocols in Molecular Biology. John Wiley & Sons, NY2010: 1-17
        • Limozin L.
        • Sengupta K.
        Quantitative reflection interference contrast microscopy (RICM) in soft matter and cell adhesion.
        ChemPhysChem. 2009; 10: 2752-2768
        • Schindelin J.
        • Arganda-Carreras I.
        • Cardona A.
        • et al.
        Fiji: an open-source platform for biological-image analysis.
        Nat. Methods. 2012; 9: 676-682
        • Sengupta K.
        • Aranda-Espinoza H.
        • Hammer D.
        • et al.
        Spreading of neutrophils: from activation to migration.
        Biophys. J. 2006; 91: 4638-4648
        • Lam Hui K.
        • Wang C.
        • Upadhyaya A.
        • et al.
        Membrane dynamics correlate with formation of signaling clusters during cell spreading.
        Biophys. J. 2012; 102: 1524-1533
        • Ketchum C.
        • Miller H.
        • Upadhyaya A.
        • et al.
        Ligand mobility regulates B cell receptor clustering and signaling activation.
        Biophys. J. 2014; 106: 26-36
        • Nolen B.J.
        • Tomasevic N.
        • Pollard T.D.
        • et al.
        Characterization of two classes of small molecule inhibitors of Arp2/3 complex.
        Nature. 2009; 460: 1031-1034
        • Sengupta K.
        • Limozin L.
        Adhesion of soft membranes controlled by tension and interfacial polymers.
        Phys. Rev. Lett. 2010; 104: 088101
        • Gauthier N.C.
        • Fardin M.A.
        • Sheetz M.P.
        • et al.
        Temporary increase in plasma membrane tension coordinates the activation of exocytosis and contraction during cell spreading.
        Proc. Natl. Acad. Sci. USA. 2011; 108: 14467-14472
        • Brodovitch A.
        • Bongrand P.
        • Pierres A.
        T lymphocytes sense antigens within seconds and make a decision within one minute.
        J. Immunol. 2013; 191: 2064-2071
        • Babich A.
        • Li S.
        • Burkhardt J.K.
        • et al.
        F-actin polymerization and retrograde flow drive sustained PLCγ1 signaling during T cell activation.
        J. Cell Biol. 2012; 197: 775-787
        • Bornschlögl T.
        • Romero S.
        • Bassereau P.
        • et al.
        Filopodial retraction force is generated by cortical actin dynamics and controlled by reversible tethering at the tip.
        Proc. Natl. Acad. Sci. USA. 2013; 110: 18928-18933
        • Lomakina E.B.
        • Waugh R.E.
        Micromechanical tests of adhesion dynamics between neutrophils and immobilized ICAM-1.
        Biophys. J. 2004; 86: 1223-1233
        • Dushek O.
        • Mueller S.
        • Valitutti S.
        • et al.
        Effects of intracellular calcium and actin cytoskeleton on TCR mobility measured by fluorescence recovery.
        PLoS ONE. 2008; 3: e3913
        • Sens P.
        Rigidity sensing by stochastic sliding friction.
        EPL. 2013; 104 (Europhysics Letters): 38003
        • Smith A.-S.
        • Sengupta K.
        • Sackmann E.
        • et al.
        Force-induced growth of adhesion domains is controlled by receptor mobility.
        Proc. Natl. Acad. Sci. USA. 2008; 105: 6906-6911
        • Springer T.A.
        • Dustin M.L.
        Integrin inside-out signaling and the immunological synapse.
        Curr. Opin. Cell Biol. 2012; 24: 107-115
        • Kumari S.
        • Vardhana S.
        • Dustin M.L.
        • et al.
        T lymphocyte myosin IIA is required for maturation of the immunological synapse.
        Front. Immunol. 2012; 3: 1-13
        • Yu Y.
        • Smoligovets A.A.
        • Groves J.T.
        Modulation of T cell signaling by the actin cytoskeleton.
        J. Cell Sci. 2013; 126: 1049-1058
        • Robert P.
        • Nicolas A.
        • Limozin L.
        • et al.
        Minimal encounter time and separation determine ligand-receptor binding in cell adhesion.
        Biophys. J. 2011; 100: 2642-2651
        • Dedecker P.
        • Duwé S.
        • Zhang J.
        • et al.
        Localizer: fast, accurate, open-source, and modular software package for superresolution microscopy.
        J. Biomed. Opt. 2012; 17: 126008
      Advertisement