Patching and capping of LFA-1 molecules on human lymphocytes

Summary The distribution and dynamics of LFA-1 molecules over the surface of human lymphocytes were analysed using immunogold label-fracture and fracture-flip methods. Patching and capping were induced by incubation at 37°C with antibodies directed against the alpha and beta chains respectively of the heterodimeric LFA-1 molecule, and were followed by immunofluorescence. Treatment with 12-O-tetradecanoylphorbol-13-acetate (TPA) to link LFA-1 molecules to the cytoskeleton increased the percentage of capped cells, implying a faster and more efficient process of capping. At all times of clustering or upon phorbol ester treatment, the concentration of LFA-1 in patches and then in caps was not accompained by a parallel concentration of membrane particles on the freeze-fractured plasma membranes. Our results support the role of the cytoskeleton in regulating the capping phenomenon and in controlling the structural organization of the plasma membranes.     

Making a Little Affinity Go a Long Way: A Topological View of LFA-1 Regulation

Lymphocytes utilize adhesion to navigate in the body and to transiently interact with a variety of potential antigen presenting cells. Interactions of adhesion molecules are governed by the law of mass action and the less understood rules of apposed biological membranes. Biochemical parameters such as adhesion molecule affinity only tell part of the story. Factors such as lateral mobility, membrane alignment and cytoskeletal interactions are equally important in determining the final outcome. Therefore it is important to determine mechanisms by which the properties of cell membranes and the cytoskeleton reinforce or hinder adhesion molecule interactions. Work from my lab has shown that one mechanism by which lymphocyte adhesion molecules cooperate is to align adhering membranes with nanometer precision. Here, I discuss a model for LFA-1 regulation that is dependent on three independent processes: LFA-1 lateral mobility, ligand induced generation of a small amount of high affinity LFA-1 and local membrane alignment. I propose that coordination of these processes allows rapid interconversion between stable adhesion and detachment.     

Surface distribution and partition during freeze-fracture of CD8 antigens on human lymphocytes and on epithelial transfected cells

Freeze-fracture immunocytochemistry was used to analyse the surface distribution, redistribution induced by antibodies, and partition during freeze-fracture, of CD8 molecules on human T lymphocytes and rat epithelial transfected (FRT-U10) cells. Immunogold labelling of CD8 antigens was uniform over the unfractured cell surfaces of both lymphocytes and epithelial transfected cells. After freeze-fracture, the gold particles were associated with the exoplasmic outer leaflets of the plasma membranes in both cell types. In lymphocytes, incubation with antibodies at 37° C up to 20 min induced patching and capping of the antigens on the unfractured cell surface. After fracture, the patched molecules appeared associated with the protoplasmic inner leaflet of the plasma membranes. Parallel antibody-treatment at 37° C of FRT-U10 cells induced clustering of CD8 molecules but failed to cause further aggregation in larger patches or in caps. After freeze-fracture, the immunola-belling was clustered, but associated with the exoplasmic outer leaflet of the plasma membranes as in untreated cells. The different redistribution induced by antibodies and the different behaviour on fracture of the redistributed molecules in the two cell types may be regulated by CD8 interaction with the cytoskeleton.     

Purified intercellular adhesion molecule-1 (ICAM-1) is a ligand for lymphocyte function-associated antigen 1 (LFA-1)

Lymphocyte function-associated antigen 1 (LFA-1) is a leukocyte cell surface glycoprotein that promotes intercellular adhesion in immunological and inflammatory reactions. It is an alpha beta complex that is structurally related to receptors for extracellular matrix components, and thus belongs to the integrin family. ICAM-1 (intercellular adhesion molecule-1) is a distinct cell surface glycoprotein. Its broad distribution, regulated expression in inflammation, and involvement in LFA-1-dependent cell-cell adhesion have suggested that ICAM-1 may be a ligand for LFA-1. We have purified ICAM-1 and incorporated it into artificial supported lipid membranes. LFA-1+ but not LFA-1- cells bound to ICAM-1 in the artificial membranes, and the binding could be specifically inhibited by anti-ICAM-1 treatment of the membranes or by anti-LFA-1 treatment of the cells. The cell binding to ICAM-1 required metabolic energy production, an intact cytoskeleton, and the presence of Mg2+ and was temperature dependent, characteristics of LFA-1- and ICAM-1-dependent cell-cell adhesion.     

Antibody to adhesion molecule LFA-1 enhances plasma neutralization of human immunodeficiency virus type 1.

We have shown that a monoclonal antibody to the cell surface adhesion molecule LFA-1 (CD18/CD11a) enhances plasma neutralization of a laboratory isolate (HIVMN) and a primary isolate (HIV28R) of human immunodeficiency virus type 1. Human phytohemagglutinin blasts were infected with HIVMN or HIV28R in the presence of plasma pooled from HIV-positive individuals (AIDS plasma) or immunoglobulin G from AIDS plasma alone or combined with a monoclonal antibody (MAb) to LFA-1. While AIDS plasma alone at a dilution of 1:1,250 neutralized HIVMN and HIV28R infection by 15 and 0%, respectively, in the presence of a saturating concentration of the MAb to LFA-1 the plasma neutralized both viruses by more than 80% at this dilution. Immunoglobulin G purified from AIDS plasma, when used in combination with the MAb to LFA-1, showed the same synergistic effect in HIV neutralization as seen with the AIDS plasma and anti-LFA-1. The MAb against LFA-1 partially neutralized both viral isolates (45 to 55%) on its own. These results demonstrate significant synergy between the plasma and antibody against LFA-1 in the neutralization of HIV. The observations therefore suggest an important role for adhesion molecules in HIV infectivity and transmission. The results have implications for the recently observed host effect on HIV susceptibility to antibody neutralization.     

Molecular cytochemistry of CD3 and CD4 antigens in human lymphocytes as studied by label-fracture and by fracture-label

Label-fracture and fracture-label membrane immunocytochemistry are used to analyze the surface distribution, dynamics and partition on fracture of CD3 and CD4 antigens of human T lymphocytes. Redistribution of the antigens, induced by treatment at 37 degrees C with specific monoclonal antibodies, results in patching and capping of the labeling as observed in label-fractured specimens. Examination of platinum/carbon replicas of freeze-fractured plasma membranes of antibody-treated cells does not reveal recognizable domains of intramembrane particles. However, in cells where the aggregation of intramembrane particles is induced by incubation with glycerol, colloidal gold-labeled CD3 and CD4 molecules are seen confined to particulate domains of the membrane. Therefore, the lack of visible aggregation of intramembrane particles in patched or capped regions of the membrane implies that migration of CD3 and CD4 antigens with concentration in domains of the membrane is achieved contemporaneously with export of other non-capped integral membrane proteins from the same regions, in a process of diffusional equilibrium. Examination of fracture-labeled specimens shows that CD4 molecules partition on fracture with the inner protoplasmic face of the plasma membrane. This partition illustrates the transmembrane attitude of the antigen molecule and is a probable consequence of interaction of the protein with other components of the membrane or with the cytoskeleton.     

Lymphocyte adhesion mediated by lymphocyte function-associated antigen-1. I. Long term augmentation by transient increases in intracellular cAMP.

Lymphocyte adhesion to target cells is mediated, in part, by the interaction of lymphocyte function-associated Ag-1 (LFA-1) with intercellular adhesion molecule-1 (ICAM-1). Cells of the B cell line, JY, express both coreceptors and have been used as a model for intercellular adhesion mediated by these molecules. Elevation of the intracellular cAMP concentration ([cAMP]i), by any of several reagents, for periods as brief as 30 min, led to enhanced intercellular adhesion in a concentration dependent manner 5 to 8 h later. Two protein kinase A inhibitors, KT5720 and H-89, but not the protein kinase C inhibitor calphostin C, blocked the effects of elevated [cAMP]i. These data suggest a role for protein kinase A in this response. The adhesion augmented by increased [cAMP]i was due to LFA-1/ICAM-1 interactions between cells because it was blocked by either anti-LFA-1 or anti-ICAM-1 mAb. Elevated [cAMP]i induced cell surface patching of LFA-1, but not ICAM-1, and this redistribution preceded intercellular adhesion. Finally, redistribution of LFA-1 was not mediated by the cytoskeleton. These results suggest a model in which transient activation of protein kinase A results in increased local concentration of LFA-1 at the cell surface and in augmented long term adhesion mediated by this integrin.     

3D single-particle tracking and optical trap measurements on adhesion proteins.

A three-dimensional single-particle tracking system was combined with an optical trap to investigate the behavior of transmembrane adhesion proteins. We exploited this setup to investigate which part of the cell adhesion protein LFA-1 forms a connection to the cytoskeleton after binding to its ligand ICAM-1. LFA-1 is an integrin consisting of an alpha and a beta chain. Thus far, only the cytoplasmic tail of the beta chain is known to form a connection to the cytoskeleton. We investigated cells that express a mutant form of LFA-1 that lacks the complete beta cytoplasmic tail and therefore is not thought to bind to the cytoskeleton. Interestingly, single-particle tracking measurements using beads coated with the ligand ICAM-1 indicate that this mutant form of LFA-1 does not move freely within the cell membrane, suggesting that LFA-1 is still connected to the cytoskeleton network. This finding is strongly supported by the observation that LFA-1 exhibits a more diffusive motion when the cytoskeleton network is disrupted and confirmed by the optical trap measurements used to force the proteins to move through the membrane. Collectively, our findings suggest that the interaction of LFA-1 with the cytoskeleton cannot solely be attributed to the cytoplasmic part of the beta chain.     

Epstein-Barr virus latent infection membrane protein alters the human B-lymphocyte phenotype: deletion of the amino terminus abolishes activity.

A latent infection membrane protein (LMP) encoded by the Epstein-Barr virus (EBV) genome in latently infected, growth-transformed lymphocytes alters the phenotype of a human EBV-negative B-lymphoma cell line (Louckes) when introduced by gene transfer. These LMP-expressing cells exhibit increased homotypic adhesion due to increased expression of the adhesion molecules LFA-1 and ICAM-1. Increased homotypic adhesion could foster B-cell growth by facilitating autocrine growth factor effects. LFA-3 expression is also induced. The induction of LFA-3 and ICAM-1 results in increased heterotypic adhesion to T lymphocytes. This could result in more effective T-cell immune surveillance. Since LMP is expressed in EBV-transformed lymphocytes and has been demonstrated to transform rodent fibroblasts in vitro, a wide range of possible effects on B-lymphoma cell growth were assayed. In the Louckes B-lymphoma cell line, EBV LMP causes increased cell size, acid production, plasma membrane ruffling, and villous projections. Although cell proliferation rate was not greatly affected, the steady-state intracellular free calcium level, transforming growth factor beta responsiveness, and expression of the lymphocyte activation markers (CD23 and transferrin receptor) were increased. Thus, LMP appears to be a mediator of EBV effects on B-cell transformation. In transfected lymphoma cells, LMP localizes to patches at the cell periphery and associates with the cytoskeleton as it does in EBV-transformed B lymphocytes or in rodent fibroblasts. A partially deleted form of LMP (D1LMP) does not aggregate in patches or associate with the cytoskeleton and had little effect on B-cell growth. Thus, cytoskeletal association may be integral to LMP activity.     

Synaptic clusters of MHC class II molecules induced on DCs by adhesion molecule-mediated initial T-cell scanning

Initial adhesive contacts between T lymphocytes and dendritic cells (DCs) facilitate recognition of peptide-MHC complexes by the TCR. In this report, we studied the dynamic behavior of adhesion and Ag receptors on DCs during initial contacts with T-cells. Adhesion molecules LFA-1- and ICAM-1,3-GFP as well as MHC class II-GFP molecules were very rapidly concentrated at the DC contact area. Binding of ICAM-3, and ICAM-1 to a lesser extent, to LFA-1 expressed by mature but not immature DC, induced MHC-II clustering into the immune synapse. Also, ICAM-3 binding to DC induced the activation of the Vav1-Rac1 axis, a regulatory pathway involved in actin cytoskeleton reorganization, which was essential for MHC-II clustering on DCs. Our results support a model in which ICAM-mediated MHC-II clustering on DC constitutes a priming mechanism to enhance antigen presentation to T-cells.     

The major histocompatibility complex-restricted antigen receptor on T cells. IX. Role of accessory molecules in recognition of antigen plus isolated IA

Antibody inhibition studies were done to determine which molecules on the surface of the T cell hybridomas other than their receptors for antigen plus IAd were involved in interaction with antigen-presenting B cells, with artificial IAd membranes on glass beads, or with anti-receptor antibodies coupled to Sepharose beads. We found that T cell LFA-1 was only involved when B cells were used to present antigen plus IAd, whereas T cell L3T4 was involved in the response of T cells to antigen plus IAd either on cells or in artificial membranes, but not if anti-receptor antibodies were used to stimulate the T cells. From these results we concluded that LFA-1 may be involved in the recognition of a ligand on cells that was not present in artificial membranes, but that L3T4 might interact with a nonpolymorphic portion of class II molecules present in both intact antigen-presenting cells and the antigen-presenting artificial membranes.     

The actin cytoskeleton regulates LFA-1 ligand binding through avidity rather than affinity changes.

To elucidate the role of the cytoskeleton regulating avidity or affinity changes in the leukocyte adhesion receptor lymphocyte function-associated antigen-1 (LFA-1) (alpha(L)beta(2)), we generated mutant cytoplasmic LFA-1 receptors and expressed these into the erythroleukemic cell line K562. We determined whether intercellular adhesion molecule-1 (ICAM-1)-mediated adhesion of LFA-1, lacking parts of its cytoplasmic tails, is regulated through receptor diffusion/clustering and/or by altered ligand binding affinity. All cytoplasmic deletion mutants that lack the complete beta(2) cytoplasmic tail and/or the conserved KVGFFKR sequence in the alpha(L) cytoplasmic tail were constitutively active and expressed high levels of the activation epitopes NKI-L16 and M24. Surprisingly, whereas these mutants showed a clustered cell surface distribution of LFA-1, the ligand-binding affinity as measured by titration of soluble ligand ICAM-1 remained unaltered. The notion that redistribution of LFA-1 does not alter ligand-binding affinity is further supported by the finding that disruption of the cytoskeleton by cytochalasin D did not alter the binding affinity nor adhesion to ICAM-1 of these mutants. Most cytoplasmic deletion mutants that spontaneously bound ICAM-1 were not capable to spread on ICAM-1, demonstrating that on these mutants LFA-1 is not coupled to the actin cytoskeleton. From these data we conclude that LFA-1-mediated cell adhesion to ICAM-1 is predominantly regulated by receptor clustering and that affinity alterations do not necessarily coincide with strong ICAM-1 binding.     

Dual role of the actin cytoskeleton in regulating cell adhesion mediated by the integrin lymphocyte function-associated molecule-1.

Intracellular signals are required to activate the leukocyte-specific adhesion receptor lymphocyte function-associated molecule-1 (LFA-1; CD11a/CD18) to bind its ligand, intracellular adhesion molecule-1 (ICAM-1). In this study, we investigated the role of the cytoskeleton in LFA-1 activation and demonstrate that filamentous actin (F-actin) can both enhance and inhibit LFA-1-mediated adhesion, depending on the distribution of LFA-1 on the cell surface. We observed that LFA-1 is already clustered on the cell surface of interleukin-2/phytohemagglutinin-activated lymphocytes. These cells bind strongly ICAM-1 and disruption of the actin cytoskeleton inhibits adhesion. In contrast to interleukin-2/phytohemagglutinin-activated peripheral blood lymphocytes, resting lymphocytes, which display a homogenous cell surface distribution of LFA-1, respond poorly to intracellular signals to bind ICAM-1, unless the actin cytoskeleton is disrupted. On resting peripheral blood lymphocytes, uncoupling of LFA-1 from the actin cytoskeleton induces clustering of LFA-1 and this, along with induction of a high-affinity form of LFA-1, via "inside-out" signaling, results in enhanced binding to ICAM-1, which is dependent on intact intermediate filaments, microtubules, and metabolic energy. We hypothesize that linkage of LFA-1 to cytoskeletal elements prevents movement of LFA-1 over the cell surface, thus inhibiting clustering and strong ligand binding. Release from these cytoskeletal elements allows lateral movement and activation of LFA-1, resulting in ligand binding and "outside-in" signaling, that subsequently stimulates actin polymerization and stabilizes cell adhesion.     

Diffusion, patching, and capping of stearoylated dextrans on 3T3 cell plasma membranes

Fluorescence-labeled trinitrophenylated stearoylated dextrans have been used as controllable analogues of cell membrane proteins on model membranes and on a variety of natural cell membranes. This paper reports their behavior on 3T3 mouse fibroblast plasma membranes. Spatial distribution on the membrane was studied by fluorescence microscopy, and molecular mobility was measured by fluorescence photobleaching recovery. At concentrations from 10(2) to 3 X 10(3) molecules/micron2 essentially homogeneous fluorescence was observed after treatment with these stearoyldextrans in culture. Diffusion coefficients and fractional recovery of fluorescence after photobleaching were cvoncentration independent. For 3 X 10(3) molecules/micron2 we found at 23 degrees C D = (3.0 +/- 1.8) X 10(-10) cm2/s with 65 +/- 17% recovery and at 37 degrees C D = (7.0 +/- 5.0) X 10(-10) cm2/s without a change of the fractional recovery. Cross-linking with antibodies stopped diffusion on a macroscopic scale and sometimes induced patching, mottling (defined as the development of gaps in the fluorescence layer), and capping (defined as the confinement of the fluorescence to less than 50% of the cell). Capping required approximately 3 h at 37 degrees C and was inhibited by metabolic poisons and cytochalasin B. These drugs did not affect stearoyldextran diffusion or fractional recovery. Colchicine, which did not dramatically affect capping, slowed diffusion two- to threefold but did not affect fractional recovery. The antibody inhibition of the diffusion of stearoyldextrans precedent to capping did not affect the diffusion of a lipid probe or fluorescein isothiocyanate labeled membrane proteins. When the trinitrophenylated stearoyldextran was cleared from most of the surface by capping and the surface subsequently relabeled with stearoyldextran, the diffusion coefficient and fractional recovery of the second label were identical with those of the first label prior to capping. Thus, capping does not clear an immobilizing factor from the membrane.     

Organization of the integrin LFA-1 in nanoclusters regulates its activity

The beta2-integrin LFA-1 facilitates extravasation of monocytes (MOs) into the underlying tissues, where MOs can differentiate into dendritic cells (DCs). Although DCs express LFA-1, unlike MOs, they cannot bind to ICAM-1. We hypothesized that an altered integrin organization on the DC plasma membrane might cause this effect and investigated the relationship between membrane organization and function of LFA-1 on MOs and DCs. High-resolution mapping of LFA-1 surface distribution revealed that on MOs LFA-1 function is associated with a distribution in well-defined nanoclusters (100-150-nm diameter). Interestingly, a fraction of these nanoclusters contains primed LFA-1 molecules expressing the specific activation-dependent L16-epitope. Live imaging of MO-T-cell conjugates showed that only these primed nanoclusters are dynamically recruited to the cellular interface forming micrometer-sized assemblies engaged in ligand binding and linked to talin. We conclude that besides affinity regulation, LFA-1 function is controlled by at least three different avidity patterns: random distributed inactive molecules, well-defined ligand-independent proactive nanoclusters, and ligand-triggered micrometer-sized macroclusters.     

Annexin V relocates to the periphery of activated platelets following thrombin activation: an ultrastructural immunohistochemical approach

We have previously shown biochemically that the physiological agonist thrombin can cause translocation of endogenous annexin V to a fraction containing all platelet membranes. This paper reports ultrastructural immunohistochemical data revealing that annexin V molecules localize with plasma membranes of blood platelets following thrombin activation. When ultrathin sections of resting platelets were examined by immunogold staining, annexin V was found to be cytosolic, having a generalized distribution throughout the platelet. After thrombin activation, annexin V became peripheral in location and plasmalemma association increased. Morphometric analysis of gold particles shows that annexin V relocates specifically to the plasma membrane and its underlying cytoskeleton following treatment with thrombin. In control platelets 6.1% +/- 0.78 of annexin V is present at the plasma membrane and 15.0% +/- 0.82 in the region corresponding to the membrane cytoskeleton (10-80 nm); after stimulation with 0.5 unit/ml thrombin for 2 min this increased to 16.7% +/- 0.22 and 40.4% +/- 0.53, respectively.     

Binding of plasma membrane glycoproteins to the cytoskeleton during patching and capping is consistent with an entropy-enhancement model

Concentrations of concanavalin A that induced patching and capping of cell surface receptors on Dictyostelium discoideum also induce binding of the receptors to the cortical cytoskeleton, which was isolated by density-gradient centrifugation. The receptors were solubilized by deoxycholate, purified by affinity chromatography, and used to determine whether the receptors bound directly to the cytoskeletal protein, actin. As the concentration of actin was increased, many of the receptors became bound to purified filamentous rabbit muscle actin, even in the absence of concanavalin A. As in the ligation-induced binding of receptors to the cortical cytoskeleton in cells, concanavalin A induced much stronger binding of the purified receptors to filamentous actin. The results were consistent with a previously stated hypothesis that induction of receptor binding to the cytoskeleton during their patching and capping is driven by clustering the receptors, which reduces their translational entropy and by doing so enhances their avidity for the cytoskeleton.     

Calmodulin-binding proteins of bovine thyroid plasma membranes

Calmodulin binding proteins in bovine thyroid plasma membranes were investigated using the ¹²⁵I-labeled calmodulin gel overlay technique. The purified thyroid plasma membranes contained two calmodulin binding proteins with molecular weights of approx. 220 000 and 150 000 respectively. The binding of ¹²⁵I-labeled calmodulin to the calmodulin binding proteins was inhibited by excess unlabeled calmodulin, 100 μM trifluoperazine or 1 mM EGTA, indicating that the binding was calmodulin-specific and calcium-dependent. The calmodulin binding proteins appear to be components of the cytoskeleton since they remained in the pellet after treatment of the thyroid plasma membranes with 1% Triton X-100. Similar calmodulin binding proteins were present in rat liver plasma membranes, but not in human red blood cell plasma membranes. These two calmodulin binding proteins may interact with other components of the cytoskeleton and regulate endocytosis, exocytosis and hormone secretion in thyroid cells.     

Antibody-induced changes on rabbit sperm surface inhibit gamete interaction

Interaction of specific ligands with cell surface molecules may induce reorganization of surface components. A monoclonal antibody (B-12) against sperm surface antigens of 40kDa size induced molecules on the plasma membrane overlying the acrosome of rabbit sperm to cluster in small aggregates at 0°C (patching). At an elevated temperature of 37°C these clusters of antigen antibody complexes collected into a large aggregate on one pole of the cell forming a cap (capping). This cap-like structure showed a reduction in size over a period of time and eventually disappeared from the sperm surface. Inhibition of capping by sodium azide indicated that it is an energy-dependent process. Patching of antigens did not require energy. Involvement of sperm head cytoskeleton in the process of capping was evident from potentiation of cap formation by cytoskeleton disrupting agents like cytochalasin B and D. Patching of antigen antibody complexes was not affected by either of the agents. The loss of antigen antibody complexes from sperm surface was mainly due to shedding of the complexes in the surrounding media. Sperm with patches of antigen antibody complexes did not adhere to oocytes. Sperm from the group where a majority of the sperm were denuded of the antigen antibody complexes also did not bind to oocytes. © 1994 Wiley-Liss, Inc.