Biomimetic Droplets for Artificial Engagement of Living Cell Surface

To evaluate the specificity of the cell–droplet association, we also targeted a T-cell receptor not involved in the TCR activation process. To addre...
0 downloads 0 Views 3MB Size
Download PDF
Article pubs.acs.org/Langmuir

Biomimetic Droplets for Artificial Engagement of Living Cell Surface Receptors: The Specific Case of the T-Cell Nadia Bourouina,† Julien Husson,†,∥ Claire Hivroz,‡ and Nelly Henry*,†,§ †

Institut Curie, Centre de Recherche, CNRS, UMR168-Paris, F-75248 France Institut Curie, Centre de Recherche, 3INSERM, U932-Paris, F-75248 France § UPMC Paris, F-75005 France ‡

S Supporting Information *

ABSTRACT: Liquid colloids, in the form of droplets grafted with specific biomolecules, are emerging as potential biomimetic systems. Here we show for the first time the possibility of forming hybrid conjugates between an advanced living cell model, the T-cell of the Jurkat cell line, and a specifically grafted droplet. Using T-cells expressing a fluorescent chimeric protein associated with the TCR/CD3 complex and fluorescent ligand-grafted droplets, we demonstrate formation of an interfacial contact concentrated in linking molecules, the morphology and dynamics of which strongly depend on the targeted receptor. The sequence of events ranges from the initial concentration of molecules following an unbound molecule gradient to active actindriven spreading and fragmentation of the contact, ending with droplet internalization. We observed synchronized colocalization of receptors and ligands driven by cell dynamics and closely mirrored by the droplet interface. Using intracellular calcium probe Fura-2, we also showed that the cell/droplet interaction can trigger the T-cell signaling cascade. By examining molecular dynamics using FRAP measurements, we observed a nearly frozen cell droplet joining interface. Taken together, our results point to liquid colloids as promising new tools both for probing cell surface interactions and receptor dynamics and for manipulating biological cell functions.

■

INTRODUCTION Emulsion droplets offer a unique system of dispersed fluid surfaces. Their physical and physicochemical properties have been widely investigated, and rules for predicting their stability under very different conditions are well-known.1 Emulsions have spread to a large spectrum of technological applications in the food, pharmaceutical, bitumen, and paint industries. Recently, these dispersed materials have also gained ground in microfluid technology as actuatable pico-reservoirs of molecules and single cell containers.2 In addition, droplets have already been shown to form stable conjugates with cells via nonspecific electrostatic interactions.3 However, their outcome as molecularly addressable systems remains limited, as does their potential impact on actuation of biological cells in biomedical applications requiring specific recognition, such as sorting and artificial activation. In addition, the development of intelligent systems that can achieve specific recognition of molecular targets at the surface of cells and tissues for drug targeting and immunotherapy continues to represent a challenge;4 it is worth noting that liquid colloids have long been present in medical applications in the form of parenteral nutrition, demonstrating their innocuity, which underlines the potential for application of these already industrialized and medically authorized materials. Recently, we showed that the droplet fluid surface can be grafted with complementary recognition molecules and form specific droplet−droplet conjugates exhibiting a fluid inter© 2012 American Chemical Society

droplet contact loaded in binding molecules. In these systems, molecular ligands grafted onto the droplet displayed rapid twodimensional diffusion on the droplet surface. These liquid colloids provide the sole model of fluid presentation of a molecule collection on a surface at the scale of a living cell, thereby opening up a whole avenue for devising new tools dedicated to the exploration of biological cell surface interactions.5 Nonetheless, prior to these fundamental and applied developments, the specific interactions that can be formed between a droplet and a living cell must be elucidated. To address this question, we chose to investigate the contact of an antibody-grafted droplet with a T-cell of the Jurkat cell line. The T-cell is a major player in the mammalian immune response in charge of antigen recognition.6 In the organism, this cell exerts its biological function, differentiation and cytokine secretion, after an initial step of activation involving formation of a dynamic contact zone with a specific antigenpresenting cell. This cell−cell contact zone, the so-called immunological synapse (IS), includes several pairs of ligands and receptors, but the central event of activation consists of engagement of the T-cell recognition receptor and its CD3 subunits (TCR-CD3), which trigger the intracellular biochemReceived: January 27, 2012 Revised: March 7, 2012 Published: March 13, 2012 6106

dx.doi.org/10.1021/la300398a | Langmuir 2012, 28, 6106−6113

Langmuir

Article

ical signaling.7 This process is the subject of intensive research to better understand the mechanism of activation of this cell and eventually better control and manipulate its biological response in vaccinal or antitumoral immunotherapeutic strategies.8 In this context, several model systems based on supported membrane lipid bilayers9 or polymer particles engrafted with ligands of the TCR10 have been devised to dissect the link between the formation of surface ligand− receptor bonds and the cell signaling, designating the T- cell as an incomparable model system for evaluating the ability of the liquid colloids to form relevant specific interactions with a living cell. We report herein the specific interaction of a ligand-grafted droplet with a living cell. In order to monitor formation of interfacial contact at the molecular level, we brought into contact T-cells expressing a fluorescent chimeric protein associated with the TCR/CD3 complex (ζCD3-GFP) and fluorescent ligand-grafted droplets. Time-lapse confocal imaging of formation of the cell-droplet hybrid conjugate enabled following the dynamics of the binding molecules on both the cell and the droplet. To evaluate the specificity of the cell− droplet association, we also targeted a T-cell receptor not involved in the TCR activation process. To address the question of the ability of the grafted droplets to act as cell mimes and manipulate cell fate, we examined whether surface molecular binding is able to induce biochemical signaling inside the cell. To clarify this point, we tested T-cell commitment in the signaling pathway upon contact formation, via monitoring of the onset of an intracellular calcium wave, an early sign of Tcell activation triggering.11 To bring to light the driving forces at work in formation of cell−droplet contact, we assessed involvement of the T-cell cytoskeleton in the process using an actin polymerization inhibitor. We then examined molecular dynamics in the contact both on the droplet and on the cell side by analyzing fluorescence recovery after partial photobleaching of the interfacial joining film. The results showed that ligands engrafted at the droplet liquid surface were specifically recognized by the cell, which not only formed a stable hybrid conjugate with the droplet, but also induced intracellular biochemical signaling, depending on the targeted receptor. The dynamics of the interaction was also highly dependent upon the targeted receptor, although a common initial phase resembling the passive physicochemical process observed in the interdroplet interaction was present in all sequences of events that we evidenced. These findings indicate a new avenue for grafted liquid colloids as biomimetic tools for actuating cell surface biological interactions and related cell responses.

■

supplemented with 0.2% v/v of Tween 20. The pH of the buffer was adjusted to 7.2 with NaOH. MES buffer contained 10 mM MES supplemented with 0.2% v/v of Tween 20. The pH was adjusted to 5.3 with NaOH. MS buffer used for all experiments with cells contained NaCl 140 mM, KCl 5 mM, MgCl2 1 mM, CaCl2 1 mM, and HEPES 10 mM. Cells. ζCD3−GFP Jurkat cells (clone 20 obtained from Dr. A. Alcover, Pasteur Institute, Paris, France), transfected as described elsewhere,12 were grown in Glutamax containing RPMI 1640 (Invitrogen Life Technologies, Carlsbad, CA) supplemented with 100 U/mL penicillin, 100 μg/mL streptomycin, and 10% FCS (fetal bovine serum EU-approved origin, Gibco Invitrogen). Liquid Droplet Preparation. The emulsion droplets were prepared by dispersing and shearing soybean oil in an aqueous continuous phase containing 15% v/v F68 and 2% v/v sodium alginate at a final oil fraction equal to 75% as previously described in detail.13 The droplets were then grafted with biotin-(PE0)3-amine and coupled with fluorescent streptavidin.14 Briefly, the emulsion was activated using 120 g/L EDAC in the presence of 20 mg/mL sNHS diluted in PB-Tween 20 buffer for 30 min at room temperature. The emulsion was washed four times in PB-Tween 20 buffer before adding the amino-biotin derivative at a concentration of 10 g/L. The sample was then gently stirred at room temperature for an additional 30 min before extensive washing with PB-Tween 20. Then, the droplets were coated with biotinylated monoclonal antibodies (mAb) directed toward the T-cell receptors TCR/CD3 or CD28 (mouse anti-CD3, UCHT1 and anti-CD28, ref 13-0281, BD Biosciences, France). The mAb were conjugated in the laboratory with biotin using the Molecular Probes coupling kit (F-6347, Invitrogen, France). Coupling with the droplet surface was obtained by 30 min incubation in the presence of 0.013 mg·mL−1 mAb with droplets (4 × 107 mL−1) at room temperature in PB-Tween 20. The droplets were washed twice with PB-Tween 20 and then with MS. Anti-CD3 and anti-CD28 surface densities, measured using a fluorescent goat-anti-mouse (GAM) antibody (Alexa 488-conjugated GAM, Invitrogen, France) and flow cytometry titration,10c were equal to (4.9 ± 1.4) × 103 and (2.6 ± 0,4) × 103 mAb/μm2, respectively. Experimental Chamber. The custom-built chamber consisted of two glass coverslips assembled with vacuum grease to form a 5 mm wide, 18 mm long, 50 μm high channel. Then 4.5 μL of a cell sample volume containing 106 cells/ml in MS buffer was introduced to fill up the channel and allow the cells to sediment for 4 min. Next, the buffer was exchanged with MS containing 0.5% BSA to avoid further cell spreading and to eliminate dead cells. For cell−droplet contact live imaging, the chamber was placed on the microscope stage upside down before introducing the droplets, which became creamy under gravity. Calcium Imaging. A Nikon TiE inverted microscope (Nikon Instruments, France) equipped with a cooled CCD camera (Orca ER, Hamamatsu, France), and a motorized wheel for excitation filters was used to image intracellular calcium using 340 and 380 nm excitation wavelengths and a 510 nm emission wavelength. Ratio images reflecting the intracellular calcium concentration were derived from the images collected at 340 and 380 nm excitation wavelengths, respectively, as in the work by Husson and colleagues.15 Just before the experiments, the cells at a concentration of 106 cells/mL were loaded with a 2 mM Fura-2 AM probe at 37 °C for 20 min in the dark, and then washed and resuspended in MS buffer. FRAP and Fluorescence Imaging. A Nikon TE2000-E inverted microscope equipped with a CSU22 Yokogawa confocal spinning disk and a 60× , Plan Apo 1.4 NA lens was used for fluorescence recovery after photobleaching (FRAP) experiments. Excitation used a MellesGriot 100 mW argon−krypton laser. Emission was filtered using Chroma filter sets (500LP for GFP, BP 630/75 nm for Alexa-555 streptavidin). Photonic Instrument MicroPoint laser technology (488 or 561 nm, 50 mW) was used to create a circular bleached spot of 1 μm radius. After a bleaching laser pulse of 50 ms on a droplet and 200 ms on a cell, time-lapse image acquisition was triggered with illumination light shuttered between acquisitions. Bleaching was triggered after a 400 ms prebleach period recording. The same setup

MATERIALS AND METHODS

Chemicals and Buffers. Soybean oil, sodium alginate, 1-ethyl-3(3-dimethylaminopropyl) carbodiimide (EDAC), sulfo-N-hydroxysuccinimide (sNHS), polyethylene glycol sorbitan monolaurate (Tween 20), 2-(N-morpholino)ethanesulfonic acid (MES), 4-(2hydroxyethyl)piperazine-1-ethanesulfonic acid (HEPES), bovine serum albumin (BSA), and all chloride salts (NaCl, KCl, MgCl2, CaCl2) were purchased from Sigma-Aldrich (France). Ethylene oxide− propylene oxide triblock copolymer (F68) was from Uniqema (France). Alexa fluor 555-conjugated streptavidin (Alexa 555-SA) and calcium probe Fura-2 AM were purchased from Invitrogen (France), and biotin-(PE0)3-amine from Interchim (France). Buffers were prepared with ultrapure water. Phosphate-Tween 20 buffer (PBTween 20) contained 10 mM monobasic sodium phosphate 6107

dx.doi.org/10.1021/la300398a | Langmuir 2012, 28, 6106−6113

Langmuir

Article

was used for time-lapse image acquisition of cell-droplet contact formation. FRAP data were analyzed following a previously published method16 that consisted of deriving the diffusion coefficient of the bleached molecule from fluorescence intensity profiles obtained from each image of the recovery sequence adjusted to a Gaussian function: ⎡ 2(x − x )2 ⎤ 0 ⎥ F(x) = F0 exp⎢− +K ⎢⎣ ⎥⎦ d2 where K is a constant, x0 is the center ,and d is the width of the profile, related to the diffusion coefficient by the formula: d2 = 8Dt for 2D diffusion. We estimated the diffusion coefficient by plotting d2 as a function of time and using linear regression to estimate the slope of the graph, keeping only the first 5−10 images (depending on the quality of acquisition).

■

RESULTS Anti-CD3 Droplet/T-Cell Contact Formation Dynamics. In order to determine whether a ligand collection grafted onto a liquid interface can form a bidimensional specific contact with targeted receptors at the surface of a living cell, we brought into contact, in a microscope chamber, droplets coupled to an anti-CD3 antibody with T-cells expressing ζCD3-GFP. The droplets flowing into the chamber and randomly colliding with the cells were efficiently arrested by cells previously sedimented in the chamber. Formation of cell-droplet interfacial contact triggered a dynamic sequence of events (Figure 1A) reported by time-lapse imaging of the cell−droplet conjugate in bright field and fluorescence. We recorded droplet ligand (anti CD3SA-alexa 555) and cell receptor (ζCD3-GFP) fluorescence, which reported the corresponding molecular densities versus time (Figure 1B). The initial step of interfacial contact formation consisted of a synchronized molecular concentration of droplet ligands and cell receptors. This first molecular recruitment started immediately after the droplet-cell encounter and remained localized for 55 ± 7 s. Interestingly, this initial phase was quite similar to the molecular contact formed when we brought into contact two droplets bearing complementary ligands and receptors, that is, biotin and streptavidin (Figure 1C and D). For this reason, we called this initial step the passive phase. Then, while droplet−droplet contact remained localized to the initial interdroplet interface and was stable for at least several hours, the cell−droplet conjugate exhibited a spectacular dynamic process beginning with phase II. This step was characterized by spreading of the contact associated with a further increase in the ligand−receptor bidimensional concentration at the interface (Figure 1A, line 3) up to a plateau lasting several tens of seconds (approximately 30−60 s varying from one conjugate to the other). Then, the contact burst into fragments of concentrated ligand−receptor bonds that began to be deployed around the droplet as a hemispheric cup in phase III (Figure 1A, line 4). Phase IV corresponded to spreading out and dilution of phase III foci over the entire droplet, leading to apparent cell membrane closure at the droplet pole opposite the initial contact site (Figure 1A, line 5). At that stage, the droplet was pulled toward the cell body and eventually internalized (Figure 1A, line 6). At the same time, the concentrated clusters of ligands and receptors were relocalized at the initial contact pole of the droplet (see Supporting Information, video 1). The whole process was achieved within 20−30 min depending on the conjugates. We observed strict colocalization of the ligands and receptors throughout the process.

Figure 1. Specific contact formation on the surface of a grafted droplet. (A) Time-lapse images showing formation of anti-CD3 droplet/T-cell contact; the main dynamic steps are represented by lines from top to bottom: initial physical contact (line 1), initial receptor and ligand concentration (line 2), contact dynamic spreading (line 3), contact fragmentation (line 4), droplet engulfment (line 5), and final droplet internalization (line 6). Bright field, GFP, and Alexa 555 fluorescence separated and overlaid are shown from left to right (fluorescence is in false colors). Bar is 5 μm. See also Supporting Information movie S1. (B) ζCD3-GFP (●) and anti CD3-SA-Alexa 555 (gray down triangle) fluorescence intensities in cell-droplet interfacial contact as a function of time from initial contact to contact fragmentation; also shown is fluorescence of the droplet bulk interface (▽). Small square areas (smaller than the contact) of 4 pixels are drawn in the center of the contact and moved if necessary from one image to another to remain in the center of the contact when the cell− droplet moves. Mean fluorescence intensity of this region is plotted versus time. We show here a representative curve of at least four independent experiments; error bars represent SD. (C) Images of formation of interdroplet contact, BF (left) and Alexa 555 fluorescence (right); bar is 5 μm. (D) Fluorescence intensity in the interdroplet contact versus time: biotin and Alexa 555-streptavidin complementary droplets. Measurements taken as in (B).

Most cell−droplet conjugates that were formed displayed this 5-phase sequence of events (17 out of 25, i.e. 68% in a typical experiment), but in some cases the initial and contact spreading phases were very short and the contact burst into clusters in less than 20 s (20%). In a few cases, the dynamic process was interrupted in one or another of the first three phases and did not last until internalization (12%). In control experiments, we brought into contact T-cells and droplets which had not been coupled with anti-CD3 exhibiting only SA on their surface. We observed that cell−droplet collisions induced no efficient contact. In 9% of cases, a droplet was arrested upon cell contact, but no molecular concentration was induced, that is, no fluorescent patch either on the cell or on the droplet side was observed, indicating that the dynamic cell/droplet interaction observed with the anti-CD3 droplet actually resulted from molecular recognition at the liquid interface. 6108

dx.doi.org/10.1021/la300398a | Langmuir 2012, 28, 6106−6113

Langmuir

Article

T-Cell Biochemical Response to Droplet Contact. In order to evaluate the biological relevance of such recognition and to determine whether bidimensional engagement of the Tcell surface receptor by the grafted droplets triggered a specific biochemical pathway, we monitored intracellular calcium upon cell−droplet contact. For this purpose, we brought into contact anti-CD3 droplets and T-cells loaded with a Fura-2 calcium probe in the microscope chamber and then recorded time-lapse ratio images which reported intracellular calcium concentrations. Figure 2 shows a representative response curve. We

initial phase and displayed no contact dynamic spreading, nor contact burst (Figure 3). These findings demonstrate that cell/

Figure 3. Anti-CD3 droplet and latrunculin-B-treated cell interfacial contact. Graph of ζCD3-GFP (●) and anti CD3-SA-Alexa 555 (gray down triangle) fluorescence intensities in the cell−droplet interfacial contact are plotted as a function of time (details as in Figure 1). TCells were treated with 5 × 10−7 M latrunculin B for 20 min before contact with droplets. Bottom panels show fluorescence images acquired at (a) t = 80 s and (b) t = 1620 s following initial contact. ζCD3-GFP and anti CD3-SA-Alexa 555 fluorescence images are successively recorded separately and overlaid for the figure. Stable localized interfacial contact is formed; no dynamic spreading or fragmentation is observed. Figure 2. Calcium wave elicited upon anti-CD3 droplet−T-cell contact. Fura-2 ratio versus time. Ratio images are obtained by dividing image at 340 nm by image at 380 nm excitation wavelength. Cells were loaded with Fura-2 2 mM in MS buffer. Arrow shows cell−droplet initial contact at t = 0. Insets show ratio images of cell−droplet conjugate at (a) initial contact time, (b) maximum peak of calcium, and (c) 8 min after contact when the Ca2+ concentration has returned to baseline.

droplet dynamic interactions are, as expected, supported by actin polymerization. These results also revealed that the dynamic phase was preceded by receptor−ligand recruitment driven only by complementary ligand−receptor bond formation and the ensuing chemical potential gradient of unbound molecules. Engagement of Another T-Cell Receptor. We next sought to determine whether the highly dynamic actindependent behavior observed above depended on the engaged receptor. For this purpose, we coupled droplets with anti-CD28 antibodies, which do not activate the TCR/CD3 signaling pathway. Anti-CD28 droplets were brought into contact with T-cells just as had been previously done; again, we monitored the interaction by collecting time-lapse images of cell−droplet conjugates. The formation of the specific cell−droplet interface was indicated here by the increased fluorescence in the contact on the droplet side only (Figure 4). Logically, no fluorescence increase was visible on the cell side, demonstrating the specificity of surface receptor engagement and confirming that the molecular concentration in the contact was truly due to the formation of specific bonds at the liquid interface. Indeed, ζCD3-GFP receptors, which were not linked by the anti-CD28 droplet, were also not recruited in the contact. As in the case when actin polymerization is inhibited, contact formation displayed only the initial phase. These results indicate that engagement on the cell surface by a grafted droplet that targets receptors not directly coupled to cytoskeleton activity amounts to the formation of an interfacial zone concentrated in linking molecules. Initial Phase Kinetics. The initial phase of the cell/droplet interaction, during which ligand and receptor density increase, went on to a steady state lasting less than 60 s in the case of anti-CD3 droplets and untreated T-cells, just before the active actin-driven phase began. In contrast, the concentrated contact remained stable in anti-CD3 droplet/latrunculin-treated cells, in anti-CD28 droplet/untreated cells and in droplet−droplet

observed the typical calcium transient increase, indicating that the specific biochemical signaling cascade had been triggered. The intracellular calcium concentration increase began after cell−droplet contact with a time delay between 30 and 60 s, that is, during the initial contact phase. The peak of the signal occurred between 60 and 80 s later, which corresponded to the beginning of the third phase when the contact burst started. These results indicate that molecular recognition of a specific ligand collection presented on a liquid interface triggers not only polarized molecular migration of cell receptors toward the cell mime interface but also relevant biochemical signaling. Cytoskeleton Involvement in the Cell/Droplet Interaction. The formation of interfacial contact concentrated in molecules between complementary droplets can be described as an interplay of molecular interactions, chemical gradients and surface forces in a fully physicochemical process. Here the dynamics observed suggested that, beyond the first 30−60 s of the cell/droplet interaction, which strongly resembled a physicochemical mechanism, active cell-driven machinery was implemented at the cell−droplet interface. In order to elucidate this point, we monitored formation of interfacial contacts between anti-CD3 droplets and T-cells which had been previously treated with latrunculin B. This actin polymerization inhibitor, used at a concentration of 5 × 10−7 M for 20 min, prevented actin polymerization without altering the cell viability of Jurkat T-cells (data not shown). Interestingly, the cell− droplet conjugates were formed at the same frequency as with untreated cells; however, all conjugates were blocked in the 6109

dx.doi.org/10.1021/la300398a | Langmuir 2012, 28, 6106−6113

Langmuir

Article

Table 1. Initial Contact Formation Half-Times conjugate droplet anti-CD3/T-cell droplet anti-CD3/latrunculin T-cell anti-CD28/T-cell droplet SA/droplet biotin

half-time (s) ± SDa 49 173 153 58

± ± ± ±

16 26 19 9

a

Means and standard deviations obtained from three different curves giving the fluorescence increase in the initial contact. t1/2 is time during which intensity I1/2 is equal to (Imax-I0)/2.

cells, anti-CD3 droplets and T-cells in the presence of latrunculin B and anti-CD28 droplets and T-cells. We bleached a 1 μm radius region focused on half of the cell−droplet contact. Immediately after this partial bleaching, we monitored fluorescence recovery on both the bleached and unbleached zones of contact using small regions for analysis (2 pixels, i.e., 0.2 μm side length). Figure 6 shows images recorded before

Figure 4. Anti-CD28 droplet and T-cell interfacial contact. Graph of ζCD3-GFP (●) and anti CD3-SA-Alexa 555 (gray down triangle) fluorescence intensities in cell−droplet interfacial contact are plotted as a function of time. Details as in Figure 1. Bottom panels show fluorescence images acquired at t = 1350 s following initial contact, i.e., at saturation. ζCD3-GFP, anti-CD28-SA-Alexa 555 and their overlay are shown from left to right. Only anti-CD28-SA-Alexa 555 is observed in the contact; no concentration of ζCD3-GFP, indicating that only linking molecules accumulate in the interface. These images show stable contact.

conjugates. To compare initial contact formation kinetics in the different cases, we normalized the curves giving fluorescence intensity in the contact versus time to maximal amplitude of the fluorescence increase (Figure 5). We observed that the

Figure 5. Contact formation kinetics. Normalized fluorescence intensity of the Alexa 555-SA coupled ligand on the droplet side of cell−droplet interfacial contact is plotted versus time for anti-CD3 droplet/T-cell (+), anti-CD3 droplet/latrunculin B-treated T-cell (gray tilted square), anti-CD28 droplet/T-cell (◆), and droplet− droplet contact (◇). Normalized values are obtained by dividing contact fluorescence intensity at time t (I) subtracted from fluorescence intensity of the same region before contact (I0) by fluorescence intensity at the plateau (Imax) also subtracted I0, i.e., (I − I0)/ (Imax − I0). For the sake of clarity, we represent error bars on only one curve (◆). Error bars on the other curves were similar and below 10% of the signal.

Figure 6. Contact FRAP experiments. Latrunculin-B-treated T-cell− droplet interfacial contact was partially bleached using a 1 μm radius spot at: (A) 488 nm to extinguish ζCD3-GFP fluorescence on the cell membrane (top row) or 561 nm to extinguish anti CD3-SA-Alexa 555 fluorescence on the droplet side (bottom row). Bright field image shows conjugates and images show cell or droplet fluorescence before bleaching (left image), immediately after bleaching (middle image) and after 15 s bleaching (right image). Different conjugates were used to bleach GFP and Alexa-555. (B) The same experiment is shown on a droplet−droplet conjugate. In this case, fluorescence that accumulated in the contact spreads out into the joining interface within the first 5 s after bleaching. Middle column of the figure shows kimographs giving spatiotemporal evolution along the dotted line drawn on the figures. Left column shows graphs of fluorescence recovery in a 2-pixel side square region located within the bleached part of the contact (●), within the unbleached zone of the contact (gray circle), on the bulk interface (◇), and in an empty zone background (○).

characteristic time (half-time) of contact formation increased by a factor of 2 for cell−droplet contact compared to droplet− droplet contact, except for anti-CD3 droplet/untreated T-cell initial contact (Table 1). In the latter case, half time was even slightly shorter than that of interdroplet contact. These results show that contact formation kinetics depends on targeted receptor properties. Dynamics of Ligands and Receptors in Intercell− Droplet Contact. To evaluate dynamics of the cell−droplet joining film initially formed within the first 30−60 s after contact, we performed a series of FRAP experiments on cell− droplet conjugates obtained with anti-CD3 droplets and T-

and after the bleaching flash either on the cell or droplet side as well as corresponding kymograph and fluorescence recovery curves for anti-CD3 droplet/latrunculin-treated cell contact. On the time scale of the experiment, fluorescence intensity in the unbleached region of interfacial contact remained constant both on the cell and on the droplet side, indicating that ligand and 6110

dx.doi.org/10.1021/la300398a | Langmuir 2012, 28, 6106−6113

Langmuir

Article

that we studied implies that contact formation velocity is not limited by molecular diffusion, but rather by molecular binding. Indeed, molecular diffusion at the droplet surface is about 10 times faster than at the cell surface, while characteristic times of initial contact formation were similar for cell−droplet and interdroplet conjugates. Depending on the nature of the receptor targeted on the cell surface, this initial phase was followed by a dynamic sequence of events. In the particular example of the T-cell examined here, we observed that engagement of the central receptor of activation, TCR-CD3, by an anti-CD3 droplet induced spectacular spreading and fragmentation of the contact ending in droplet traction and internalization by the cell. Using an inhibitor, we showed that this active behavior was supported by actin polymerization. In addition, we demonstrated that formation of the dynamic T-cell−anti-CD3 droplet conjugate induced a transient intracellular calcium peak, a hallmark of the T-cell activation and triggering of the intracellular biochemical cascade. In contrast, formation of cell/droplet conjugates resulting from engagement of a different receptor (not involved in T-cell triggering) using anti-CD28 droplets was limited to the initial phase, like conjugates formed in the presence of actin polymerization inhibitors and interdroplet conjugates. These findings demonstrate that recognition of specific ligands anchored at the liquid surface of an emulsion droplet triggers biologically active relevant processes selectively installed by the cell in response to specific surface interactions. Previously, several other model systems had been used to investigate the T-cell response to defined molecular arrays; they also achieved intracellular signaling and showed dynamic aspects of the response upon TCR engagement. However, dynamic fragmentation of the interfacial zone into multiple foci of bonds that migrated from initial contact to the entire droplet surface had never been previously described. Studies based on supported bilayer membranes19 are intrinsically restricted to exploration of flat contact, which might lock the process in the initial phase. Traction and internalization steps were described in experiments using polymer particles, but not the initial contact fragmentation in dynamic clusters of bonds, likely due to the solid nature of the system.15 Using fluorescent labeling of binding molecules on both the droplet and cell side, we were able to follow their recruitment during contact formation. In addition FRAP measurements revealed molecular dynamics within the initial contact once formed. We observed that cell receptors (ζCD3-GFP) and droplet ligand anchors (SA-Alexa 555) exhibited in the initial phase colocalized and synchronized molecular traffic toward interfacial cell−droplet contact. During the contact formation step, molecular kinetics appeared to be driven by cell dynamics, implying that molecular diffusion on the fluid droplet surface showed no opposition to the movement of the cell receptors. Thus, molecular dynamics on the droplet surface directly reflected targeted receptor diffusion on the cell surface. These cell receptor dynamics are of primary importance in numerous cell biology issues and the droplet system provides an exceptional tool for probing cell receptor dynamics in all cases in which cell receptors are not labeled, as in primary cells or cases in which labeling disturbs the biological function. Exciting information might thus be obtained concerning cell receptor segregation and patterning using grafting of multiple ligands to the droplet surface, with no need for cell labeling.

receptor bonds were immobile in the concentrated contact. We observed slight recovery up to the fluorescence level of the bulk interface, suggesting that unbound molecules could still migrate to the contact. Similar results were obtained on the contact formed by the anti-CD28 droplet and T-cell contact, as well as on the contact formed by anti-CD3 droplets and untreated Tcells (Supporting Information, Figure S1). This behavior contrasted with that observed on interdroplet joining film, where we detected diffusion of fluorescence from the unbleached to the bleached region (Figure 6, panel B), indicating that ligand−receptor bonds were mobile in the contact. Meanwhile, FRAP experiments performed on droplet and cell contours exhibited recovery fractions equal to 0.87 ± 0.07 and 0.80 ± 0.16, which was a sign that ligands and receptors diffused freely on the droplet and cell surface. The diffusion coefficient of Alexa 555-SA coupled with anti-CD3 antibody on the droplet surface was equal to 0.97 ± 0.35 μm2/s (Supporting Information Figure S2), showing that the presence of the antibodies reduced molecule mobility at the interface; indeed, D had previously been found to be equal to 3 μm2/s in the absence of antibody. The dynamic behavior of the T-cell strongly blurred fluorescence recovery profiles, that could not be correctly fitted to the Gaussian model, preventing accurate determination of the diffusion coefficient for ζCD3-GFP in the cell membrane. However, the characteristic time of recovery enabled a rough estimation of the diffusion coefficient on the order of 0.05 to 0.1 μm2/s, consistent with previous measurement of membrane protein mobility in T-cells.17 This suggested that receptor-free diffusion in the cell membrane is 10 times slower than ligand-free diffusion on the droplet surface, supporting the hypothesis of cell/droplet interaction dynamics driven by cell properties. These results indicated that the specific contact formed between cells and grafted droplets leads to formation of an interfacial zone concentrated in bonds which appear to be frozen at a time scale of a few tens of seconds, i.e., a time scale at which mobility is measured both on cell and droplet bulk interfaces and droplet−droplet interfacial contact.

■

DISCUSSION We explored here the potential of emulsion droplets grafted with specific biomolecules as biomimetic tools for probing living cell surfaces and manipulating surface receptor-dependent functions. For this first study, we focused our attention on the interaction of the grafted droplet with the T-cell, a major player in the immune system and currently the object of intense research7,18 Our results demonstrate that presentation of ligand collection at the liquid surface of a droplet supports formation of stable hybrid conjugates with the living cell through exquisitely specific receptor−ligand recognition and binding between cell and droplet surfaces. This interaction is initiated by formation of an interfacial cell-droplet joining contact of increased receptor and ligand densities very similar to the contact zone that forms between complementary droplets.5 In the latter case, the initial phase is driven by physicochemical molecular gradients resulting from formation of the first complementary molecular bonds and a consequent decrease in unbound molecule chemical potential in the contact zone. The similarity of fluorescence patterns observed in cell−droplet conjugates suggests that the same physicochemical driving force supports the initial molecular concentration in the cell−droplet interface. The kinetics of the initial phase in the different cases 6111

dx.doi.org/10.1021/la300398a | Langmuir 2012, 28, 6106−6113

Langmuir

■

Interestingly, the cell−droplet interface appeared to be frozen when examined by FRAP in the three cases that we observed (conjugates resulting from the T-cell/anti-CD3 droplet, latrunculin-B-treated T-cell/anti-CD3 droplet and Tcell/anti-CD28 droplet). No molecular diffusion was observed on a time scale of a few tens of seconds, in contrast to what is observed in concentrated interdroplet contact, where a diffusion coefficient equal to 0.05 μm 2/s was previously found.17 Meanwhile, we measured no alteration in molecular diffusion on the droplet or cell contour. Zech and co-workers recently pointed out early alteration of lipid composition and membrane condensation at sites of TCR activation that might explain the apparent contact freeze observed here.20 However, the absence of molecular diffusion of the bonds in the contact was also noted upon engagement of anti-CD28. The molecular contact freeze could also result from transmembrane receptor anchorage to the cytoskeleton upon external binding, which would limit receptor mobility before the active phase starts. These aspects were not explored here and warrant further investigation for clarification. Our findings present grafted droplets as valuable cell mimes that could be advantageously used to dissect the mechanisms of cell responses triggered by collective surface molecular interactions. Moreover, they offer new tools for artificially activating cells in a realistic geometry enabling bidimensional engagement of surface receptors on a 3D surface.

■

CONCLUSION

■

ASSOCIATED CONTENT

Article

AUTHOR INFORMATION

Corresponding Author

*Mailing address: Institut Curie, Centre de Recherche, CNRS, UMR168, UPMC-Paris, F-75248 France. Telephone: +33 156 247 495. Fax: +33 140 51 06 36. E-mail: [email protected]. Present Address ∥

Laboratoire d’Hydrodynamique (LadHyX), Ecole Polytechnique-CNRS, 91128 Palaiseau Cedex, France. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was in part supported by grants from the Agence Nationale pour la Recherche PCV (07-188055) and the Association pour la Recherche contre le Cancer. N.B. was partly supported by a fellowship from the Fondation pour la Recherche Médicale. The authors thank the Nikon Imaging Center at the Institut Curie-CNRS and François Waharte for his help with FRAP image acquisition and analysis.

■

REFERENCES

(1) (a) Tadros, T. F. Emulsion Science and Technology; Wiley-VCH Verlag GmbH & Co. KGaA: Weinheim, Germany, 2009. (b) LealCalderon, F.; Schmitt, V.; Bibette, J. Emulsion science: basic principles; 2nd version; Springer: New York, 2007. (c) Bibette, J.; Morse, D. C.; Witten, T. A.; Weitz, D. A. Stability criteria for emulsions. Phys. Rev. Lett. 1992, 69, 2439−2442. (2) (a) Zagnoni, M.; Cooper, J. M. Droplet microfluidics for highthroughput analysis of cells and particles. Methods Cell Biol. 2011, 102, 25−48. (b) Brouzes, E.; Medkova, M.; Savenelli, N.; Marran, D.; Twardowski, M.; Hutchison, J. B.; Rothberg, J. M.; Link, D. R.; Perrimon, N.; Samuels, M. L. Droplet microfluidic technology for single-cell high-throughput screening. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 14195−14200. (3) Ravaine, V.; Bibette, J.; Henry, N. Wetting of liquid droplets on living cells. J. Colloid Interface Sci. 2002, 255, 270−273. (4) (a) Bonacucina, G.; Cespi, M.; Misici-Falzi, M.; Palmieri, G. F. Colloidal soft matter as drug delivery system. J. Pharm. Sci. 2009, 98, 1−42. (b) Vogel, F. R.; Caillet, C.; Kusters, I. C.; Haensler, J. Emulsion-based adjuvants for influenza vaccines. Expert Rev. Vaccines 2009, 8, 483−492. (c) Ott, G.; Singh, M.; Kazzaz, J.; Briones, M.; Soenawan, E.; Ugozzoli, M.; O’Hagan, D. T. A cationic sub-micron emulsion (MF59/DOTAP) is an effective delivery system for DNA vaccines. J. Controlled Release 2002, 79, 1−5. (5) Bourouina, N.; Husson, J.; Waharte, F.; Pansu, R. B.; Henry, N. Formation of specific receptor−ligand bonds between liquid interfaces. Soft Matter 2011, 7, 9130−9139. (6) Lanzavecchia, A.; Sallusto, F. Dynamics of T lymphocyte responses: intermediates, effectors, and memory cells. Science (New York, N.Y.) 2000, 290, 92−97. (7) Smith-Garvin, J. E.; Koretzky, G. A.; Jordan, M. S. T Cell Activation. Annu. Rev. Immunol. 2009, 27, 591−619. (8) Tartour, E.; Sandoval, F.; Bonnefoy, J. Y.; Fridman, W. H. Cancer immunotherapy: recent breakthroughs and perspectives. Med. Sci. (Paris) 2011, 27, 833−841. (9) (a) Mossman, K.; Groves, J. Micropatterned supported membranes as tools for quantitative studies of the immunological synapse. Chem. Soc. Rev. 2007, 36, 46−54. (b) DeMond, A. L.; Groves, J. T. Interrogating the T-cell synapse with patterned surfaces and photoactivated proteins. Curr. Opin. Immunol. 2007, 19, 722−727. (10) (a) Trickett, A.; Kwan, Y. L. T cell stimulation and expansion using anti-CD3/CD28 beads. J. Immunol. Methods 2003, 275, 251− 255. (b) Wulfing, C.; Davis, M. M. A receptor/cytoskeletal movement triggered by costimulation during T cell activation. Science (New York, N.Y.) 1998, 282, 2266−2269. (c) Carpentier, B.; Pierobon, P.; Hivroz,

Grafted droplets appear to be novel biomimetic artificial systems in a landscape currently occupied mainly by liposomes, supported bilayers and polymer particles. Yet liquid droplets combine properties previously not encountered in only one system. Unlike supported membranes, which impose flat geometry or polymer particles, the solid nature of which blocks ligand surface diffusion, they constitute 3D surfaces with a totally fluid interface that more closely probes cell properties. They can be easily prepared in quantity, and have already gone through important steps of industrial processes and marketing authorizations for medical products; indeed, emulsions are already being used in clinics for patient parenteral nutrition. Taken together, the results presented here herald the strong potential of specifically grafted droplets for living cell fundamental science and applications.

S Supporting Information *

Figure S1: FRAP experiments on anti-CD3 droplet/T-cell and anti-CD28/T-cell conjugates. Fluorescence recovery curves and images. Figure S2: Determination of the diffusion coefficient of anti-CD3-Alexa 555-SA molecules on the surface of a droplet. Intensity profile and square width time dependence. Video 1: Movie of the formation of an anti-CD3 droplet/T-cell conjugate. ζCD3-GFP and anti-CD3-SA-Alexa 555 fluorescence signals were combined. Movie is played at 7 fps. Frames were acquired every 5 s. Contact was initiated at image 4. This material is available free of charge via the Internet at http:// pubs.acs.org. 6112

dx.doi.org/10.1021/la300398a | Langmuir 2012, 28, 6106−6113

Langmuir

Article

C.; Henry, N. T-cell artificial focal triggering tools: linking surface interactions with cell response. PLoS One 2009, 4, e4784. (11) Oh-hora, M.; Rao, A. Calcium signaling in lymphocytes. Curr. Opin. Immunol. 2008, 20, 250−258. (12) Blanchard, N.; Di Bartolo, V.; Hivroz, C. In the immune synapse, ZAP-70 controls T cell polarization and recruitment of signaling proteins but not formation of the synaptic pattern. Immunity 2002, 17, 389−399. (13) Mason, T. G.; Bibette, J. Shear Rupturing of Droplets in Complex Fluids. Langmuir 1997, 13, 4600−4613. (14) Fattaccioli, J.; Baudry, J.; Emerard, J.; Bertrand, E.; Goubault, C.; Henry, N.; Bibette, J. Size and fluorescence measurements of individual droplets by flow cytometry. Soft Matter 2009, 5, 7. (15) Husson, J.; Chemin, K.; Bohineust, A.; Hivroz, C.; Henry, N. Force generation upon T cell receptor engagement. PLoS One 2011, 6, e19680. (16) (a) Seiffert, S.; Oppermann, W. Systematic evaluation of FRAP experiments performed in a confocal laser scanning microscope. J. Microsc. 2005, 220, 20−30. (b) Waharte, F.; Steenkeste, K.; Briandet, R.; Fontaine-Aupart, M. P. Diffusion measurements inside biofilms by image-based fluorescence recovery after photobleaching (FRAP) analysis with a commercial confocal laser scanning microscope. Appl. Environ. Microbiol. 2010, 76, 5860−5869. (17) Zhu, D. M.; Dustin, M. L.; Cairo, C. W.; Golan, D. E. Analysis of two-dimensional dissociation constant of laterally mobile cell adhesion molecules. Biophys. J. 2007, 92, 1022−1034. (18) (a) Lewinsohn, D. A.; Gold, M. C.; Lewinsohn, D. M. Views of immunology: effector T cells. Immunol. Rev. 2011, 240, 25−39. (b) Dustin, M. L.; Chakraborty, A. K.; Shaw, A. S. Understanding the structure and function of the immunological synapse. Cold Spring Harbor Perspect. Biol. 2010, 2, a002311. (c) Kennedy, R.; Celis, E. Multiple roles for CD4+ T cells in anti-tumor immune responses. Immunol. Rev. 2008, 222, 129−144. (19) (a) DeMond, A. L.; Mossman, K. D.; Starr, T.; Dustin, M. L.; Groves, J. T. T cell receptor microcluster transport through molecular mazes reveals mechanism of translocation. Biophys. J. 2008, 94, 3286− 3292. (b) Doh, J.; Irvine, D. J. Immunological synapse arrays: patterned protein surfaces that modulate immunological synapse structure formation in T cells. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 5700−5705. (c) Manz, B. N.; Jackson, B. L.; Petit, R. S.; Dustin, M. L.; Groves, J. T-cell triggering thresholds are modulated by the number of antigen within individual T-cell receptor clusters. Proc. Natl. Acad. Sci. U.S.A. 2011, 108, 9089−9094. (d) Grakoui, A.; Bromley, S. K.; Sumen, C.; Davis, M. M.; Shaw, A. S.; Allen, P. M.; Dustin, M. L. The immunological synapse: a molecular machine controlling T cell activation. Science (New York, N.Y.) 1999, 285, 221−227. (e) Salaita, K.; Nair, P. M.; Petit, R. S.; Neve, R. M.; Das, D.; Gray, J. W.; Groves, J. T. Restriction of receptor movement alters cellular response: physical force sensing by EphA2. Science 2010, 327, 1380−1385. (20) Zech, T.; Ejsing, C. S.; Gaus, K.; de Wet, B.; Shevchenko, A.; Simons, K.; Harder, T. Accumulation of raft lipids in T-cell plasma membrane domains engaged in TCR signalling. EMBO J. 2009, 28, 466−476.

6113

dx.doi.org/10.1021/la300398a | Langmuir 2012, 28, 6106−6113