Segregation of HLA-C from ICAM-1 at NK Cell Immune Synapses Is Controlled by Its Cell Surface Density

Catarina R. Almeida and Daniel M. Davis
J Immunol November 15, 2006, 177 (10) 6904-6910; DOI: https://doi.org/10.4049/jimmunol.177.10.6904

Abstract

NK cell activity is controlled by the integration of signals from numerous activating and inhibitory receptors at the immunological synapse (IS). However, the importance of segregation and patterning of proteins at the NK cell IS is unknown. In this study, we report that the level of expression of HLA-C on target cells determined its supramolecular organization and segregation from ICAM-1 at the NK cell IS, as well as its capacity to inhibit NK cell cytotoxicity. At YTS NK cell synapses formed with target cells expressing low levels of HLA-C (i.e., 104/cell surface), a multifocal patterning of MHC class I protein predominated, whereas for higher levels of expression (105/cell surface), clusters of HLA-C were more commonly homogeneous, ring-shaped, or containing multiple exclusions. This correlation of protein density with its patterning at the IS was independent of ATP- or actin-driven processes. Importantly, ICAM-1 and HLA-C segregated only at synapses involving target cells expressing high levels of MHC protein. For peripheral blood NK clones, there were specific thresholds in the level of target cell HLA-C needed to inhibit cytotoxicity and to cause segregation of HLA-C from ICAM-1 at the synapse. Thus, the synapse organization of HLA-C, determined by its level of expression, could directly influence NK cell inhibition, e.g., by regulating the proximity of activating and inhibitory receptors. For the first time, this suggests an important function for the assembly of an inhibitory NK cell IS. More broadly, segregation of proteins at intercellular contacts could transmit information about protein expression levels between cells.

Immune cell interactions sometimes trigger the segregation of proteins into micrometer-scale domains at the intercellular contact, thereby creating an immunological synapse (IS)3 (1, 2, 3, 4, 5, 6, 7, 8). Synapses involving many different cell types have been described and even the structure of the IS formed between specific cell types can vary (9, 10, 11). The prototypical synapse, with a central cluster of pMHC/TCR surrounded by a ring of integrin has been seen for both CD4+ T cell-APC and CD8+ T cell-target cell synapses (1, 3, 12). In contrast, multifocal patterns of TCR have been seen for synapses involving thymocytes (13, 14) and for T cell-dendritic cell interactions (15). The IS can have several functions (11, 16), one possibility being that spatial segregation of different proteins across the intercellular contact can regulate intercellular signaling by, for example, regulating the proximity between receptors and kinases or phosphatases.

NK cells respond early to infection and are important in tumor surveillance through both cytokine production and cytotoxicity. Their activity is controlled by the balance of signals from numerous activating and inhibitory cell surface receptors (17, 18). Loss of MHC class I protein expression, for example, caused by some viruses (19), can trigger NK cell responses through a loss of inhibitory receptor ligation (20, 21, 22). At the NK cell IS (23, 24, 25, 26), there is heterogeneity in the patterning of the target cell MHC class I protein HLA-C such that it can cluster as a single patch, within a ring, or contain several discrete regions where the protein is excluded (27). Factors influencing the diversity of patterning at synapses are largely unknown. Here, by imaging conjugates formed between NK cells and target cells expressing different levels of GFP-tagged HLA-Cw6, we directly demonstrate that the supramolecular patterning and segregation of proteins at an intercellular contact reflects protein expression levels and therefore could be used to report such information between cells.

Materials and Methods

Cell lines and transfectants

An EBV-transformed human B cell line, 721.221, that does not express endogenous HLA-A, -B, or -C, was transfected by electroporation with plasmids encoding GFP-tagged HLA-Cw6 (23) and cells expressing different levels of GFP were then selected by flow cytometry. Expression of MHC protein was also confirmed by Western blotting with an anti-GFP mAb (clone JL-8; BD Clontech) used at 14 μg/ml, followed by 16 ng/ml HRP-conjugated goat anti-mouse IgG (ImmunoPure; Pierce).

YTS, an NK cell line, was previously transfected with the inhibitory receptor KIR2DL1, which recognizes HLA-Cw6 (28). Human peripheral blood NK clones were generated as previously described (26), except that NK cells were sorted magnetically, according to manufacturer’s instructions (StemCell Technologies). The purity and phenotype of the human NK clones were determined by staining for CD3, CD56, CD94 (all from BD Biosciences), and KIR2DL1 (EB6; Serotec). NK cell cytotoxicity against different target cells was assessed in standard 5-h [35S]Met release assays. Spontaneous release of [35S]Met was <25% of the maximal release.

Quantifying the number of specific proteins expressed

Populations of microbeads coated in known numbers of Ab-binding sites (Quantum Simply Cellular; Bangs Laboratories), as well as each cell line, were stained with saturating amounts (100 μg/ml) of Cy5-labeled W6/32, an anti-MHC class I mAb, in PBS/3% BSA/5% horse serum for 1 h at 4°C and washed four times in PBS/1% BSA/0.01% sodium azide. After flow cytometric analysis, the number of Ab-binding sites of each cell line was determined by comparing its median fluorescence intensity (MFI) with a standard curve generated with the stained beads.

Similarly, to quantify the amount of KIR2DL1 molecules at the surface of YTS/KIR2DL1 and human NK cells, populations of microbeads as well as each cell line were stained with saturating amounts (20 μg/ml) of EB6 mAb in PBS/1% BSA/0.01% sodium azide for 1 h at 4°C, followed by 15 μg/ml Cy5-conjugated goat anti-mouse IgG (Jackson ImmunoResearch Laboratories). The cells were then washed four times in PBS/1% BSA/0.01% sodium azide before flow cytometry.

For quantifying the total number of GFP molecules in each 721.221 transfectant, beads coated with known amounts of GFP (BD Biosciences) and the transfected cell lines were analyzed by flow cytometry. The number of GFP molecules expressed per cell was calculated by comparing its MFI with a standard curve generated with the beads’ fluorescence.

Imaging the IS

For imaging, conjugates were allowed to form by cells falling together under gravity, and were stained as previously described (26). A total of 10 μg/ml anti-human ICAM-1 mAb (clone HA58; BD Pharmingen), 4 μg/ml Alexa 568-conjugated goat anti-mouse IgG (Molecular Probes), and 3 μg/ml Cy5-conjugated AffiniPure goat anti-mouse IgG (Jackson ImmunoResearch) was used to stain cells as appropriate. Randomly selected conjugates of different cell types were imaged by laser scanning confocal microscopy (TCS SP2; Leica) using excitation wavelengths of 488 and 568 or 633 nm, under a ×63 oil immersion objective.

ATP depletion and inhibition of actin polymerization

Both NK cells and target cells were treated with 100 mM sodium azide for 2 h or with 10 μM cytochalasin D for 1 h at 37°C/5% CO2 or with DMSO, the diluent for cytochalasin D, as a control. ATP depletion was measured by assaying luciferase activity according to the manufacturer’s instructions (ATPLite-M; Packard Instrument). Azide treatment resulted in 70–98% of ATP being depleted (data not shown). The effect of cytochalasin D was confirmed by observing a lack in movement and an altered shape of the treated cells as well as by detecting a loss in cytotoxicity of treated NK cells (data not shown).

Image analysis

When the fluorescence within a part of the intercellular contact was at least twice the average fluorescence intensity at the cell membrane away from the contact, the protein was considered clustered (ImageJ; National Institutes of Health). The percentage of GFP fluorescence and the ratio of fluorescence at the IS compared with another region of the same area were quantified from three-dimensional reconstructions of conjugate fluorescence (Volocity; Improvision). Colocalization or segregation of MHC class I protein and ICAM-1 were determined by visual inspection of en face views of the IS. Pearson coefficients were calculated (ImageJ plug-in Manders coefficients; T. Collins and W. Rasband, ImageJ), after subtraction of background intensity from a region outside the synapse (plug-in background subtraction from region of interest; M. Cammer, ImageJ).

Results

Thresholds in the level of target cell MHC protein needed to inhibit NK cell cytotoxicity

To be able to determine the influence of MHC class I density on the NK cell IS, an EBV-transformed human B cell line, 721.221, that does not express endogenous HLA-A, -B, or -C (29), was transfected to express GFP-tagged HLA-Cw6 (23) and cells expressing different amounts of GFP-tagged MHC class I molecules were then selected by flow cytometry (transfectants 6.1–6.7; Fig. 1⇓A). Transfectants expressed between ∼2 × 105 and 9 × 106 GFP-tagged molecules in total per cell (data not shown). Expression of the GFP-tagged MHC protein in each cell line was confirmed by Western blotting cell lysates with an anti-GFP mAb (Fig. 1⇓B). The brightness of GFP fluorescence in each transfectant correlated with the level of cell surface expression of MHC class I protein, as detected with two anti-human MHC class I mAb (Fig. 1⇓, C and D). Comparing the staining of cells with beads coated in known numbers of Ab-binding sites, we determined that the transfectants expressed between 104 and 2 × 105 MHC class I molecules at the cell surface (Fig. 1⇓E). For comparison, peripheral blood NK cells and T cells expressed between 4 × 104 and 9 × 104 MHC class I proteins at the cell surface, and some immortal cell lines expressed between 2 × 104 and 2 × 105 (Fig. 1⇓, F and G), consistent with previous quantitative studies (30).

FIGURE 1.
FIGURE 1.

Establishment of target cells expressing different levels of HLA-C protein. A, Transfectants of 721.221 were selected to express different amounts of HLA-Cw6-GFP. Open histograms, each cell line expressing different amounts of HLA-Cw6-GFP; filled histogram, untransfected 721.221 cells. B, Expression of HLA-Cw6-GFP was confirmed by Western blot with an anti-GFP mAb. Lane 0, Untransfected cells, whereas lanes 1–7 refer to cell lines (6.1 to 6.7) expressing progressively higher levels of MHC class I protein. Lower row, The same blot after a longer exposure. The GFP fluorescence intensity of the 721.221 transfectants is correlated with the number of MHC class I molecules at the cell surface, as determined by staining with C, W6/32, and D, HC10, mAb that bind MHC class I protein. The extent of cell staining with anti-MHC class I mAb W6/32 (C) or HC10 (D) was plotted against the GFP fluorescence. E, Total number of MHC class I proteins at the cell surface of each transfectant. Means ± SEM from three measurements are shown. F and G, Number of cell surface MHC class I proteins expressed by different immortal cell lines (F) and by different primary human NK and T cell lines (G).

It has been previously established that NK cell cytotoxicity is inversely correlated with MHC class I expression (see for example, Refs. 31 and 32). Here, we found a sharp threshold in the level of target cell expression of MHC class I protein that is required to inhibit the cytotoxicity of polyclonal NK cell lines derived from peripheral blood (Fig. 2⇓A). A sharp threshold in the level of expression of target cell HLA-C required to inhibit NK cells was commonly seen (i.e., in six of eight NK cell lines tested) even where only 6–35% of the cells expressed KIR2DL1 (stained by mAb EB6). It may be important to note that inhibition of polyclonal NK lines was only evident for NK cells that had not received IL-2 for at least 6 days.

FIGURE 2.
FIGURE 2.

Inhibition of NK cell cytotoxicity correlates with MHC class I protein expression. Percent lysis of each target cell transfectant by (A) three different peripheral blood human NK cell lines (denoted by ▴, ▪, and •), or (B) three different NK cell clones (denoted ▴, ▪, and •), plotted against the level of expression of GFP-tagged HLA-Cw6. [35S]Met release assays determined target cell lysis over 5 h at a 10:1 E:T ratio. Open symbols (▵, □, and ○) represent data points for untransfected 721.221. C, The percent lysis by untransfected YTS (squares) or YTS/KIR2DL1 (circles) shown against the level of expression of GFP-tagged HLA-Cw6 on each 721.221 transfectant; open symbols representing data points for untransfected 721.221. Means of between 3 and 21 experiments are shown, with SEM. D, The percentage of conjugates where MHC protein clustered at the IS was determined after coincubating YTS/KIR2DL1 cells with the different target cell lines for 45 min by confocal microscopy. E, Quantification of fluorescence at the IS. The percent of fluorescence at the IS relative to the total fluorescence in the cell was calculated from three-dimensional reconstructions of conjugates. The results are not significantly different between the different target cells, as determined by ANOVA. F, The ratio of the fluorescence intensity at the synapse relative to another region of the cell surface, was determined from three-dimensional reconstructions of conjugates. A ratio of one indicates that there was no specific accumulation of protein at the IS. The results are not significantly different between the different target cells, as determined with a nonparametric Kruskal-Wallis test.

Intriguingly, the level of target cell MHC class I proteins required to inhibit different human peripheral blood NK clones varied considerably (examples shown in Fig. 2⇑B). Clearly, diversity in the repertoire of receptors and ligands that control NK cell responses makes a specific analysis of the effect of HLA-C density extremely difficult with primary NK cell clones. Hence, we also used a simple transfectant system. Specifically, we used the human NK cell line YTS transfected to express an inhibitory receptor KIR2DL1 (YTS/KIR2DL1) at a similar level as on the cell surface of peripheral blood NK cells. YTS/KIR2DL1 expressed ∼1–2 × 105 KIR2DL1 molecules at the cell surface, similar to the number expressed on NK clones derived from peripheral blood (data not shown). Ligation of KIR2DL1 by target cell HLA-Cw6 inhibits YTS cytotoxicity in accordance with the “missing self” hypothesis (20). The level of expression of GFP-tagged HLA-Cw6 on each target cell correlated with the level of inhibition of YTS/KIR2DL1 cytotoxicity, in a similar manner to one of the three NK clones shown in Fig. 2⇑B, while not affecting lysis by untransfected YTS cells (Fig. 2⇑C).

Patterning of MHC class I protein reflects its target cell surface density

YTS/KIR2DL1 was coincubated with each target cell line for 45 min at 37°C, fixed, and imaged by confocal microscopy. The frequency of conjugates where MHC class I protein clustered at the contact increased only slightly with the level of target cell surface expression of MHC class I protein (Fig. 2⇑D). More surprisingly, both the percentage of fluorescence at the IS relative to the total fluorescence in the cell and the ratio of GFP-tagged HLA-Cw6 accumulated at the IS compared with elsewhere at the cell surface were constant (Fig. 2⇑, E and F). Thus, the number of HLA-C molecules at the IS is proportional to the overall level of expression on the target cell.

To determine whether patterning of the IS depends on target cell surface density of MHC class I protein, the pattern of GFP-tagged HLA-C at the IS was classified as being either homogeneous, ring-shaped, containing multiple exclusions, multifocal (examples being shown in Fig. 3⇓A), or was unclassified. It may be important to note that a homogeneous distribution of MHC protein does not imply that the cluster of HLA-C fills the whole intercellular contact. After 10 min of coincubation of YTS/KIR2DL1 with target cells expressing different amounts of HLA-C-GFP, a clear transition from multifocal to other patterns of MHC protein (i.e., homogeneous, ring-shaped, or with multiple exclusions) at the IS correlated with increasing density of MHC protein (Fig. 3⇓B). Confirming that these patterns are caused by KIR ligation, a multifocal distribution of MHC class I protein was the most common seen at synapses between untransfected YTS cells and target cells expressing high levels of HLA-C (data not shown). There was no obvious temporal relationship between these categories of MHC distribution at the IS since the same changes in MHC class I patterning, from a multifocal to other distributions for increasing target cell expression of MHC protein, were seen when the cells were coincubated for 45 min (Fig. 3⇓C). Also, live cell imaging showed that patterns could interchange, but that there was no consistent temporal relationship between these categories of MHC distribution (data not shown and Ref. 27).

FIGURE 3.
FIGURE 3.

Density of MHC class I protein determines its patterning at the NK cell IS. A, Representative images of GFP-tagged HLA-C at the IS. Left column, The bright field image of typical cell conjugates, followed by a column with the respective GFP-tagged MHC class I protein distribution. The third panel shows the reconstructed IS for each conjugate and the right column shows the same IS color-coded for the fluorescence intensity of GFP, as shown in the color scale. From top to bottom: multifocal, homogeneous, containing multiple exclusions, and ring-shaped. Scale bar, 10 μm. B, Relative occurrence of each IS pattern for each target cell transfectant (6.1–6.7) after coincubation with YTS/KIR2DL1 for 10 min. Between 20 and 37 conjugates were analyzed for each target cell line. C, Relative occurrence of each IS pattern for each target cell after coincubating the cells for 45 min. Between 46 and 84 conjugates were analyzed for each target cell line.

Patterning of MHC class I protein does not depend on ATP-dependent processes or actin polymerization

Previously, we found that the frequency of MHC class I accumulation at the NK cell IS does not require ATP-dependent processes (23). To test whether variation in the supramolecular organization of HLA-C with its density required ATP-dependent processes, we depleted ATP in both NK cells and target cells with azide. After treatment with azide, the transition from multifocal to other patterns of MHC protein distribution (i.e., homogeneous, ring-shaped, or with multiple exclusions) with increasing levels of MHC class I protein still occurs, demonstrating that HLA-C spontaneously organizes into specific patterns at intercellular contacts. Small differences seen after azide treatment (Fig. 4⇓A) may be due to factors such as a change in cell morphology or a change in cell surface expression of proteins as well as a direct effect of ATP depletion. Therefore, we next specifically tested the influence of the actin cytoskeleton on patterning of HLA-C at the IS by treating both NK cells and target cells with 10 μM cytochalasin D, which inhibits actin polymerization (33). Here, we studied the influence of cytochalasin D on the structure of the IS formed after 45 min of coincubating cells, by which time the frequency of HLA-C clustered at the IS is known to be unaffected (34). Patterning of HLA-C at the NK cell IS across the range of transfectants with different MHC densities was unaffected by cytochalasin D (Fig. 4⇓, B and C). The percentage of conjugates with tight accumulation of F-actin at the activating IS formed between 721.221 and YTS/KIR2DL1 cells, decreased from 70 to 30% upon cytochalasin D or azide treatment, indicating that conjugates were affected by drug treatment (data not shown). Thus, correlation of the patterning of MHC protein at the NK cell synapse with the level of target cell expression of MHC class I protein is not a function of ATP-dependent processes or actin polymerization.

FIGURE 4.
FIGURE 4.

MHC class I protein patterning at the NK cell IS is not determined by ATP or actin-dependent processes. Both the target cells and YTS/KIR2DL1 were treated with (A) azide, (B) DMSO (vehicle control for the use of cytochalasin D), or (C) cytochalasin D before mixing each target with the NK cells and incubating for 45 min. Each graph shows the relative occurrence of each IS pattern, for target cells expressing different levels of MHC class I protein. Between 23 and 35 conjugates were analyzed, in each condition, for each target cell line.

Segregation of MHC class I protein from ICAM-1 depends on its cell surface density

Micrometer-scale segregation of proteins across an intercellular contact is the hallmark characteristic of an IS (1, 2, 3) and could regulate, for example, the balance between kinase and phosphatase activities on specific receptors (35). Segregation of proteins at the intercellular contact, perhaps initially on a spatial scale less than the resolution of light microscopy, could be critical for the initial triggering of T cell responses by the TCR (36, 37). At inhibitory NK cell synapses, it is well-established that the integrins ICAM-1/LFA-1 can segregate from MHC class I protein/KIR (23). Thus, here we set out to test whether the extent of this segregation varies with the target cell surface density of HLA-C.

Conjugates of YTS/KIR2DL1 with target cells expressing different levels of MHC class I protein were fixed after 45 min of coincubation and stained for ICAM-1. Segregation of ICAM-1 from clusters of MHC class I protein at the IS was determined by visual inspection of synapse reconstructions, examples being shown in Fig. 5⇓A. The number of synapses with at least some ICAM-1 segregated from MHC class I protein increased dramatically with increased expression of MHC class I protein (Fig. 5⇓B). It may be useful to note that the extent of HLA-C/ICAM-1 segregation did not vary across the three different patterns of HLA-C dominant at synapses involving target cells expressing high levels of MHC protein (data not shown). In addition, mAb EB6 marking KIR2DL1 did colocalize with MHC class I protein at the IS with target cells expressing high levels of MHC protein (data not shown), demonstrating that our detection of segregation from ICAM-1 in these synapses was not due to, for example, accessibility of mAb to only some areas of the IS. Pearson correlation coefficients, to correlate the intensity of ICAM-1 and MHC class I protein at the IS, confirmed that ICAM-1 and MHC protein were largely colocalized for target cells expressing low levels of MHC protein and segregated for high levels of MHC protein expression (Fig. 5⇓C). Demonstrating that segregation occurs specifically at synapses, ICAM-1 was not segregated from MHC class I protein in unconjugated target cells expressing high levels of HLA-C (Fig. 5⇓C). Intriguingly, although assayed on different time scales, the fraction of immune synapses where HLA-C segregated from ICAM-1, correlated with the extent of inhibition by each transfectant (Fig. 5⇓D).

FIGURE 5.
FIGURE 5.

Segregation of ICAM-1 from MHC class I protein depends on MHC density. A, Representative cases of colocalization (two top panels) and segregation (two bottom panels) of ICAM-1 from MHC class I protein. Bright field, HLA-Cw6-GFP and ICAM-1 images are shown for each conjugate. Below each conjugate, en face views of the IS are shown. Green, HLA-Cw6-GFP; red, ICAM-1; yellow, colocalization. The R number refers to the Pearson coefficient of each IS. Scale bar, 10 μm. B, Target cells expressing different amounts of MHC class I protein were incubated with YTS/KIR2DL1 for 45 min, were fixed and stained for ICAM-1. Segregation or colocalization of ICAM-1 and MHC class I proteins was determined by visually inspecting reconstructed synapses. Between 20 and 30 conjugates were analyzed for each target cell line. C, A correlation coefficient (Pearson coefficient) was determined for each IS. Individual conjugates (○) as well as the mean (bar) are shown. For comparison, the correlation coefficient for ICAM-1 and HLA-C within surface areas of unconjugated cells (transfectant 6.6) is also shown (•). D, Graph shows the percentage of transfectants 6.1–6.7 lysed by YTS/KIR2DL1 cells against the percentage of conjugates in which MHC class I protein segregated from ICAM-1 at the IS.

Synapses with peripheral blood KIR2DL1+LIR1− NK cell clones did not show such a clear increase in frequency from multifocal to other types of patterns with increasing expression of target cell HLA-C, either after 10 min (Fig. 6⇓A) or 45 min of coincubation (data not shown). This may reflect the difference in size between NK cell clones and YTS/KIR2DL1 or could reflect that NK cell clones, unlike YTS, express CD94/NKG2A which will bind target cell HLA-E and perhaps influence the organization of HLA-C. However, similar to YTS/KIR2DL1, the percentage of conjugates where MHC class I protein segregated from ICAM-1 at the synapse did correlate with the level of HLA-C expression on the target cells (Fig. 6⇓B).

FIGURE 6.
FIGURE 6.

Segregation of ICAM-1 from MHC class I protein at synapses with primary human NK cell clones depends on MHC density. A, Frequency of each IS pattern for target cell transfectants 6.2–6.7 after coincubation with KIR2DL1+LIR1− NK cell clones for 10 min. A total of 14–55 conjugates were analyzed for each target cell transfectant. B, Segregation or colocalization of ICAM-1 and MHC class I proteins was determined by visually inspecting reconstructed synapses formed between target cells expressing different amounts of MHC class I protein and peripheral blood human NK cell clones. Between 20 and 35 conjugates were analyzed for each target cell line. Each graph shows the result with one individual NK clone. C, Graph shows the percentage of transfectants 6.1–6.7 lysed by KIR2DL1+LIR1− NK cell clones against the percentage of conjugates in which MHC class I protein segregation from ICAM-1 at the IS. Data from several NK cell clones are shown, each data point representing coincubation of one specific NK clone with one of the transfectants.

The level of HLA-C expression at which HLA-C segregated from ICAM-1 varied between different human NK clones (Fig. 6⇑B). Most importantly, across different human NK clones, as with YTS/KIR2DL1 cells, inhibition of NK cell cytotoxicity correlated with the fraction of conjugates in which HLA-C clearly segregated from ICAM-1 at the IS (Fig. 6⇑C). ICAM-1 is an important ligand for activating NK- and especially YTS-mediated lysis (38, 39, 40). Thus, in summary, the level of expression of target cell HLA-C determines the extent to which it segregates from ICAM-1 at the IS, which could in turn directly influence NK cell receptor organization and inhibition of NK cytotoxicity.

Discussion

We generated a panel of target cell transfectants to express different levels of GFP-tagged HLA-Cw6, and confirmed that NK cell inhibition correlates with MHC class I expression, consistent with the “missing self” hypothesis. Levels of MHC protein on our transfectants covered the range of endogenous expression on different cell types. However, it is important to note that HLA-C is generally only ∼10% of the average level of HLA-A or -B expressed endogenously by human cells (41). Thus, a relatively high density of HLA-C was required for inhibition of NK cell cytotoxicity. This is perhaps because the target cells used here, 721.221 cells, are transformed cell lines particularly susceptible to NK cell lysis.

We found that the density of cell surface MHC class I protein determines its patterning at the NK cell IS. Specifically, the distribution of HLA-C at the YTS NK cell IS changed from being multifocal, with target cells expressing low levels of MHC protein, to clusters that were homogeneous, ring-shaped, or containing multiple exclusions, with targets expressing high levels of MHC protein. Multifocal patterning may require local fold increases in protein density that cannot be reached with the target cells expressing high MHC protein levels, so that other patterns, more diffuse through the synapse, emerge.

Importantly, we found that increased levels of MHC class I protein expression also correlated with enhanced segregation between MHC class I protein and ICAM-1, both with YTS/KIR2DL1 and with primary NK cell clones. Thus, an important next goal is to determine whether patterning and segregation of proteins at the IS directly influences downstream NK cell signaling. For example, it remains to be clarified whether and how the supramolecular organization of HLA-C, and its extent of segregation from ICAM-1, correlates with microclusters of inhibitory KIR signaling, recently visualized by Förster resonance energy transfer (42). Studies of other types of immune synapse, however, do suggest that patterning of proteins at synapses can directly influence downstream signaling: thymocyte multifocal patterns maintain Lck at high levels for a longer period (13, 14) than mature T cell synapses (43) and geometrically constrained patterns in T cell synapses correlated with different signaling activity (44, 45). Also, it has recently been shown that the extent of segregation of CD28-CD80 from TCR-pMHC at the T cell IS correlates with stronger costimulation (46).

If signaling is directly influenced by the extent of segregation of proteins at an IS, then our data suggest that this could provide the foundation for a simple mechanism by which NK cells detect the loss of MHC protein caused by viral infection or tumor transformation. It is well-established that viral proteins commonly interfere with mechanisms for immune surveillance. For example, CMV proteins can prevent expression of stress-inducible MHC class I-like protein (19). However, it could be difficult for a virus to interfere with the segregation or patterning of proteins at an IS, especially processes that are thermodynamically rather than ATP driven.

An additional importance of the current study is that it offers an explanation for the many clear discrepancies across previous studies describing the organization of immune synapses. For example, in the NK cell IS, studies have shown that KIR/MHC clusters homogeneously within specific foci (47), whereas earlier studies had shown KIR/MHC to cluster within a range of patterns including rings and clusters with several areas of exclusion (23, 27). Also, KIR2DL1 has been observed to cluster uniformly throughout the contact between YTS/KIR2DL1 and the Drosophila SC2 cell line transfected to express the inhibitory ligand HLA-Cw4 (48) in contrast to earlier evidence of specific patterning of KIR/HLA-C at synapses (23, 27). Data presented here demonstrate how these discrepancies in the literature can simply reflect differences in levels of protein expression in the cells used. For example, specific patterns of HLA-C would only occur for sufficiently high levels of HLA-C expression and also likely require the presence of ICAM-1/LFA-1 interactions (or other receptor/ligand interactions).

A prediction that arises from mathematical models of synapse assembly is that T cells that express lower numbers of TCRs would preferentially assemble a multifocal distribution of TCR at synapses (49, 50). Consistent with this, thymocytes, that do express lower numbers of TCRs, form disorganized or multifocal synapse patterns (13, 14). In addition, mathematical models of IS assembly have also predicted that below a certain threshold of TCR expression ICAM would not segregate from MHC protein (49). Segregation of ICAM-1 from MHC protein at the synapse may be driven by differences in sizes of the extracellular domains of receptor/ligand pairs (51, 52). Exclusion of ICAM-1 from the MHC class I-rich regions would thus only occur when MHC class I density at the interface is enough to sufficiently increase the energetic penalty on interspersing ICAM-1/LFA-1 with MHC class I/TCR pairs (49). Because our data for NK cells are consistent with these theoretical predictions made in specific models of the T cell IS, variation in synapse patterning with protein density may be applicable broadly.

Great breadth and beauty in the mechanisms that transmit information between cells have been discovered in a wide variety of biological systems (53). Yet it is still unclear how immune cells, for example, “count” ligands on other cells to make decisions. One way is that intracellular signaling networks create thresholds in the sum of the number of receptor/ligand-binding events. A complementary mechanism suggested by our data is that the spontaneous supramolecular distribution of proteins at immune synapses could be used to report protein expression levels between cells.

Acknowledgments

We thank A. Rae for assistance with cell sorting, R. Leung for technical assistance, F. Gordon for statistical advice, R. Mehr, and members of our laboratory for useful discussions.

Disclosures

The authors have no financial conflict of interest.

Footnotes

  • The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

  • ↵1 This work was supported by the Fundação para a Ciência e a Tecnologia, the Medical Research Council, the Biotechnology and Biological Science Research Council, and a Lister Institute Research Prize.

  • ↵2 Address correspondence and reprint requests to Dr. Daniel M. Davis, Division of Cell and Molecular Biology, Sir Alexander Fleming Building, Imperial College London, South Kensington Campus, London SW7 2AZ, U.K. E-mail address: d.davis{at}imperial.ac.uk

  • ↵3 Abbreviations used in this paper: IS, immunological synapse; MFI, median fluorescence intensity.

  • Received February 22, 2006.
  • Accepted August 25, 2006.

References

  1. ↵
    Monks, C. R., B. A. Freiberg, H. Kupfer, N. Sciaky, A. Kupfer. 1998. Three-dimensional segregation of supramolecular activation clusters in T cells. Nature 395: 82-86.
    OpenUrlCrossRefPubMed
  2. ↵
    Dustin, M. L., M. W. Olszowy, A. D. Holdorf, J. Li, S. Bromley, N. Desai, P. Widder, F. Rosenberger, P. A. van der Merwe, P. M. Allen, A. S. Shaw. 1998. A novel adaptor protein orchestrates receptor patterning and cytoskeletal polarity in T-cell contacts. Cell 94: 667-677.
    OpenUrlCrossRefPubMed
  3. ↵
    Grakoui, A., S. K. Bromley, C. Sumen, M. M. Davis, A. S. Shaw, P. M. Allen, M. L. Dustin. 1999. The immunological synapse: a molecular machine controlling T cell activation. Science 285: 221-227.
    OpenUrlAbstract/FREE Full Text
  4. ↵
    Maldonado, R. A., D. J. Irvine, R. Schreiber, L. H. Glimcher. 2004. A role for the immunological synapse in lineage commitment of CD4 lymphocytes. Nature 431: 527-532.
    OpenUrlCrossRefPubMed
  5. ↵
    Krogsgaard, M., J. B. Huppa, M. A. Purbhoo, M. M. Davis. 2003. Linking molecular and cellular events in T-cell activation and synapse formation. Semin. Immunol. 15: 307-315.
    OpenUrlCrossRefPubMed
  6. ↵
    Kupfer, A., H. Kupfer. 2003. Imaging immune cell interactions and functions: SMACs and the immunological synapse. Semin. Immunol. 15: 295-300.
    OpenUrlCrossRefPubMed
  7. ↵
    Davis, D. M.. 2006. Intrigue at the immune synapse. Sci. Am. 294: 48-55.
    OpenUrlPubMed
  8. ↵
    Fleire, S. J., J. P. Goldman, Y. R. Carrasco, M. Weber, D. Bray, F. D. Batista. 2006. B cell ligand discrimination through a spreading and contraction response. Science 312: 738-741.
    OpenUrlAbstract/FREE Full Text
  9. ↵
    Friedl, P., A. T. den Boer, M. Gunzer. 2005. Tuning immune responses: diversity and adaptation of the immunological synapse. Nat. Rev. Immunol. 5: 532-545.
    OpenUrlCrossRefPubMed
  10. ↵
    Trautmann, A., S. Valitutti. 2003. The diversity of immunological synapses. Curr. Opin. Immunol. 15: 249-254.
    OpenUrlCrossRefPubMed
  11. ↵
    Davis, D. M., M. L. Dustin. 2004. What is the importance of the immunological synapse?. Trends Immunol. 25: 323-327.
    OpenUrlCrossRefPubMed
  12. ↵
    Stinchcombe, J. C., G. Bossi, S. Booth, G. M. Griffiths. 2001. The immunological synapse of CTL contains a secretory domain and membrane bridges. Immunity 15: 751-761.
    OpenUrlCrossRefPubMed
  13. ↵
    Richie, L. I., P. J. Ebert, L. C. Wu, M. F. Krummel, J. J. Owen, M. M. Davis. 2002. Imaging synapse formation during thymocyte selection: inability of CD3ζ to form a stable central accumulation during negative selection. Immunity 16: 595-606.
    OpenUrlCrossRefPubMed
  14. ↵
    Hailman, E., W. R. Burack, A. S. Shaw, M. L. Dustin, P. M. Allen. 2002. Immature CD4+CD8+ thymocytes form a multifocal immunological synapse with sustained tyrosine phosphorylation. Immunity 16: 839-848.
    OpenUrlCrossRefPubMed
  15. ↵
    Brossard, C., V. Feuillet, A. Schmitt, C. Randriamampita, M. Romao, G. Raposo, A. Trautmann. 2005. Multifocal structure of the T cell-dendritic cell synapse. Eur. J. Immunol. 35: 1741-1753.
    OpenUrlCrossRefPubMed
  16. ↵
    Lin, J., M. J. Miller, A. S. Shaw. 2005. The c-SMAC: sorting it all out (or in). J. Cell Biol. 170: 177-182.
    OpenUrlAbstract/FREE Full Text
  17. ↵
    Lanier, L. L.. 2005. NK cell recognition. Annu. Rev. Immunol. 23: 225-274.
    OpenUrlCrossRefPubMed
  18. ↵
    Vivier, E., J. A. Nunes, F. Vely. 2004. Natural killer cell signaling pathways. Science 306: 1517-1519.
    OpenUrlAbstract/FREE Full Text
  19. ↵
    Lodoen, M. B., L. L. Lanier. 2005. Viral modulation of NK cell immunity. Nat. Rev. Microbiol. 3: 59-69.
    OpenUrlCrossRefPubMed
  20. ↵
    Ljunggren, H. G., K. Karre. 1990. In search of the “missing self”: MHC molecules and NK cell recognition. Immunol. Today 11: 237-244.
    OpenUrlCrossRefPubMed
  21. ↵
    Karre, K., H. G. Ljunggren, G. Piontek, R. Kiessling. 2005. Selective rejection of H-2-deficient lymphoma variants suggests alternative immune defence strategy. 1986. J. Immunol. 174: 6566-6569.
    OpenUrlFREE Full Text
  22. ↵
    Lanier, L. L.. 2005. Missing self, NK cells, and the white album. J. Immunol. 174: 6565
    OpenUrlFREE Full Text
  23. ↵
    Davis, D. M., I. Chiu, M. Fassett, G. B. Cohen, O. Mandelboim, J. L. Strominger. 1999. The human natural killer cell immune synapse. Proc. Natl. Acad. Sci. USA 96: 15062-15067.
    OpenUrlAbstract/FREE Full Text
  24. ↵
    Vyas, Y. M., K. M. Mehta, M. Morgan, H. Maniar, L. Butros, S. Jung, J. K. Burkhardt, B. Dupont. 2001. Spatial organization of signal transduction molecules in the NK cell immune synapses during MHC class I-regulated noncytolytic and cytolytic interactions. J. Immunol. 167: 4358-4367.
    OpenUrlAbstract/FREE Full Text
  25. ↵
    Vyas, Y. M., H. Maniar, B. Dupont. 2002. Cutting edge: differential segregation of the SRC homology 2-containing protein tyrosine phosphatase-1 within the early NK cell immune synapse distinguishes noncytolytic from cytolytic interactions. J. Immunol. 168: 3150-3154.
    OpenUrlAbstract/FREE Full Text
  26. ↵
    McCann, F. E., B. Vanherberghen, K. Eleme, L. M. Carlin, R. J. Newsam, D. Goulding, D. M. Davis. 2003. The size of the synaptic cleft and distinct distributions of filamentous actin, ezrin, CD43, and CD45 at activating and inhibitory human NK cell immune synapses. J. Immunol. 170: 2862-2870.
    OpenUrlAbstract/FREE Full Text
  27. ↵
    Carlin, L. M., K. Eleme, F. E. McCann, D. M. Davis. 2001. Intercellular transfer and supramolecular organization of human leukocyte antigen C at inhibitory natural killer cell immune synapses. J. Exp. Med. 194: 1507-1517.
    OpenUrlAbstract/FREE Full Text
  28. ↵
    Cohen, G. B., R. T. Gandhi, D. M. Davis, O. Mandelboim, B. K. Chen, J. L. Strominger, D. Baltimore. 1999. The selective downregulation of class I major histocompatibility complex proteins by HIV-1 protects HIV-infected cells from NK cells. Immunity 10: 661-671.
    OpenUrlCrossRefPubMed
  29. ↵
    Shimizu, Y., R. DeMars. 1989. Production of human cells expressing individual transferred HLA-A,-B,-C genes using an HLA-A,-B,-C null human cell line. J. Immunol. 142: 3320-3328.
    OpenUrlAbstract
  30. ↵
    Trucco, M., S. de Petris, G. Garotta, R. Ceppellini. 1980. Quantitative analysis of cell surface HLA structures by means of monoclonal antibodies. Hum. Immunol. 1: 233-243.
    OpenUrlCrossRefPubMed
  31. ↵
    Storkus, W. J., D. N. Howell, R. D. Salter, J. R. Dawson, P. Cresswell. 1987. NK susceptibility varies inversely with target cell class I HLA antigen expression. J. Immunol. 138: 1657-1659.
    OpenUrlFREE Full Text
  32. ↵
    Ohlen, C., M. T. Bejarano, A. Gronberg, S. Torsteinsdottir, L. Franksson, H. G. Ljunggren, E. Klein, G. Klein, K. Karre. 1989. Studies of sublines selected for loss of HLA expression from an EBV-transformed lymphoblastoid cell line: changes in sensitivity to cytotoxic T cells activated by allostimulation and natural killer cells activated by IFN or IL-2. J. Immunol. 142: 3336-3341.
    OpenUrlAbstract
  33. ↵
    Cooper, J. A.. 1987. Effects of cytochalasin and phalloidin on actin. J. Cell Biol. 105: 1473-1478.
    OpenUrlFREE Full Text
  34. ↵
    Standeven, L. J., L. M. Carlin, P. Borszcz, D. M. Davis, D. N. Burshtyn. 2004. The actin cytoskeleton controls the efficiency of killer Ig-like receptor accumulation at inhibitory NK cell immune synapses. J. Immunol. 173: 5617-5625.
    OpenUrlAbstract/FREE Full Text
  35. ↵
    Freiberg, B. A., H. Kupfer, W. Maslanik, J. Delli, J. Kappler, D. M. Zaller, A. Kupfer. 2002. Staging and resetting T cell activation in SMACs. Nat. Immunol. 3: 911-917.
    OpenUrlCrossRefPubMed
  36. ↵
    Choudhuri, K., D. Wiseman, M. H. Brown, K. Gould, P. A. van der Merwe. 2005. T-cell receptor triggering is critically dependent on the dimensions of its peptide-MHC ligand. Nature 436: 578-582.
    OpenUrlCrossRefPubMed
  37. ↵
    Davis, S. J., P. A. van der Merwe. 1996. The structure and ligand interactions of CD2: implications for T-cell function. Immunol. Today 17: 177-187.
    OpenUrlCrossRefPubMed
  38. ↵
    Barber, D. F., M. Faure, E. O. Long. 2004. LFA-1 contributes an early signal for NK cell cytotoxicity. J. Immunol. 173: 3653-3659.
    OpenUrlAbstract/FREE Full Text
  39. ↵
    Azuma, M., M. Cayabyab, D. Buck, J. H. Phillips, L. L. Lanier. 1992. Involvement of CD28 in MHC-unrestricted cytotoxicity mediated by a human natural killer leukemia cell line. J. Immunol. 149: 1115-1123.
    OpenUrlAbstract
  40. ↵
    Chen, X., D. S. Allan, K. Krzewski, B. Ge, H. Kopcow, J. L. Strominger. 2006. CD28-stimulated ERK2 phosphorylation is required for polarization of the microtubule organizing center and granules in YTS NK cells. Proc. Natl. Acad. Sci. USA 103: 10346-10351.
    OpenUrlAbstract/FREE Full Text
  41. ↵
    Le Bouteiller, P.. 1994. HLA class I chromosomal region, genes, and products: facts and questions. Crit. Rev. Immunol. 14: 89-129.
    OpenUrlCrossRefPubMed
  42. ↵
    Treanor, B., P. M. Lanigan, S. Kumar, C. Dunsby, I. Munro, E. Auksorius, F. J. Culley, M. A. Purbhoo, D. Phillips, M. A. Neil, et al 2006. Microclusters of inhibitory killer immunoglobulin-like receptor signaling at natural killer cell immunological synapses. J. Cell Biol. 174: 153-161.
    OpenUrlAbstract/FREE Full Text
  43. ↵
    Lee, K. H., A. D. Holdorf, M. L. Dustin, A. C. Chan, P. M. Allen, A. S. Shaw. 2002. T cell receptor signaling precedes immunological synapse formation. Science 295: 1539-1542.
    OpenUrlAbstract/FREE Full Text
  44. ↵
    Mossman, K. D., G. Campi, J. T. Groves, M. L. Dustin. 2005. Altered TCR signaling from geometrically repatterned immunological synapses. Science 310: 1191-1193.
    OpenUrlAbstract/FREE Full Text
  45. ↵
    Doh, J., D. J. Irvine. 2006. Immunological synapse arrays: patterned protein surfaces that modulate immunological synapse structure formation in T cells. Proc. Natl. Acad. Sci. USA 103: 5700-5705.
    OpenUrlAbstract/FREE Full Text
  46. ↵
    Tseng, S. Y., M. Liu, M. L. Dustin. 2005. CD80 cytoplasmic domain controls localization of CD28, CTLA-4, and protein kinase Cθ in the immunological synapse. J. Immunol. 175: 7829-7836.
    OpenUrlAbstract/FREE Full Text
  47. ↵
    Borszcz, P. D., M. Peterson, L. Standeven, S. Kirwan, M. Sandusky, A. Shaw, E. O. Long, D. N. Burshtyn. 2003. KIR enrichment at the effector-target cell interface is more sensitive than signaling to the strength of ligand binding. Eur. J. Immunol. 33: 1084-1093.
    OpenUrlCrossRefPubMed
  48. ↵
    Faure, M., D. F. Barber, S. M. Takahashi, T. Jin, E. O. Long. 2003. Spontaneous clustering and tyrosine phosphorylation of NK cell inhibitory receptor induced by ligand binding. J. Immunol. 170: 6107-6114.
    OpenUrlAbstract/FREE Full Text
  49. ↵
    Lee, S. J., Y. Hori, A. K. Chakraborty. 2003. Low T cell receptor expression and thermal fluctuations contribute to formation of dynamic multifocal synapses in thymocytes. Proc. Natl. Acad. Sci. USA 100: 4383-4388.
    OpenUrlAbstract/FREE Full Text
  50. ↵
    Weikl, T. R., R. Lipowsky. 2004. Pattern formation during T-cell adhesion. Biophys. J. 87: 3665-3678.
    OpenUrlCrossRefPubMed
  51. ↵
    Qi, S. Y., J. T. Groves, A. K. Chakraborty. 2001. Synaptic pattern formation during cellular recognition. Proc. Natl. Acad. Sci. USA 98: 6548-6553.
    OpenUrlAbstract/FREE Full Text
  52. ↵
    Burroughs, N. J., C. Wulfing. 2002. Differential segregation in a cell-cell contact interface: the dynamics of the immunological synapse. Biophys. J. 83: 1784-1796.
    OpenUrlCrossRefPubMed
  53. ↵
    Niehoff, D.. 2005. The Language of Life: How Cells Communicate in Health and Disease Joseph Henry Press, Washington, DC.
View Abstract
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section