Within the Ig superfamily (IgSF), intercellular adhesion molecules (ICAMs) form a subfamily that binds the leukocyte integrin alphaLbeta2. We report a 1.65-A-resolution crystal structure of the ICAM-3 N-terminal domain (D1) in complex with the inserted domain, the ligand-binding domain of alphaLbeta2. This high-resolution structure and comparisons among ICAM subfamily members establish that the binding of ICAM-3 D1 onto the inserted domain represents a common docking mode for ICAM subfamily members. The markedly different off-rates of ICAM-1, -2, and -3 appear to be determined by the hydrophobicity of residues that surround a metal coordination bond in the alphaLbeta2-binding interfaces. Variation in composition of glycans on the periphery of the interfaces influences on-rate.

The crystal structures of the glycosylated N-terminal two domains of ICAM-1 and ICAM-2 provided a framework for understanding the role of glycosylation in the structure and function of intercellular adhesion molecules (ICAMs). The most conserved glycans were less flexible in the structures, interacting with protein residues and contributing to receptor folding and expression. The first N-linked glycan in ICAM-2 contacts an exposed tryptophan residue, defining a conserved glycan-W motif critical for the conformation of the integrin binding domain. The absence of this motif in human ICAM-1 exposes regions used in receptor dimerization and rhinovirus recognition. Experiments with soluble molecules having the N-terminal two domains of human ICAMs identified glycans of the high mannose type N-linked to the second domain of the dendritic cell-specific ICAM-grabbing nonintegrin lectin-ligands ICAM-2 and ICAM-3. About 40% of those receptor molecules bear endoglycosidase H sensitive glycans responsible of the lectin binding activity. High mannose glycans were absent in ICAM-1, which did not bind to the lectin, but they appeared in ICAM-1 mutants with additional N-linked glycosylation and lectin binding activity. N-Linked glycosylation regulate both conformation and immune related functions of ICAM receptors.

Residues important in the interaction between the 23-residue transmembrane (TM) domains of the integrin alpha(IIb)- and beta(3)-subunits were identified by mutating each non-Leu residue to Leu. Leu substitutions of alpha(IIb) at G972, G976, and T981, and of beta(3) at I693 and G708, increased ligand binding. Substitutions with other amino acids at alpha(IIb)G972 and beta(3)G708 could also increase ligand binding. The results are consistent with and extend the helical interface between the integrin alpha- and beta-subunit TM domains previously defined by cysteine scanning and disulfide bond formation. We differentiated between affinity- and valency-based modes of activation by TM domain mutations. The mutant alpha(IIb) W967C forms disulfide-linked alpha(IIb)-subunits within an (alpha(IIb)beta(3))(2) tetramer. This tetramer behaved as an ideal model for the valency mode of regulation, because it exhibited significantly increased binding to multivalent but not monovalent ligands and basally retained the bent conformation. By contrast, the activating Leu mutants showed increased binding to the monovalent, ligand-mimetic PAC-1 Fab and increased exposure of ligand-induced binding site (LIBS) epitopes, suggesting that they partially adopt an extended conformation. Furthermore, the previously described beta(3)G708N mutation in Chinese hamster ovary cells enhanced ligand binding affinity, not valency, and did not alter cell-surface clustering as defined by confocal microscopy. Our studies provide evidence that disrupting the integrin heterodimeric TM helix-helix interface activates ligand binding mainly by increasing the monomeric affinity for ligand, but not the receptor valency, i.e., clustering.

The structural integrity of tissue proteins is damaged in processes ranging from remodeling of the extracellular matrix to destruction by microbial pathogens. Leukocytes play a prominent role in tissue surveillance and repair. However, it remains enigmatic what features of structurally decayed proteins prompt recognition by leukocyte cell-surface receptors. Here, we report that adhesion of human neutrophil granulocytes to fibrinogen is greatly increased by plasmin digestion in a mode where alphaXbeta2 dominates the integrin-dependent binding. The bacterial protease subtilisin also enhances binding by alphaXbeta2. The alphaX ligand binding domain has an unusually high affinity for carboxyl groups, with KD at approximately 100 microM. Our findings implicate enhanced accessibility of negatively charged residues in structurally decayed proteins as a pattern recognition motif for alphaXbeta2 integrin. Comparisons among integrins show relevance of these findings to the large number of ligands recognized by alphaMbeta2 and alphaXbeta2 but not alphaLbeta2. The observations suggest that the pericellular proteolysis at the leading edge of neutrophils not only facilitates passage through the extracellular matrix but also manufactures binding sites for alphaXbeta2.

The leukocyte integrin alphaLbeta2 mediates cell adhesion and migration during inflammatory and immune responses. Ligand binding of alphaLbeta2 is regulated by or induces conformational changes in the inserted (I) domain. By using a micropipette, we measured the conformational regulation of two-dimensional (2D) binding affinity and the kinetics of cell-bound intercellular adhesion molecule-1 interacting with alphaLbeta2 or isolated I domain expressed on K562 cells. Locking the I domain into open and intermediate conformations with a disulfide bond increased the affinities by approximately 8000- and approximately 30-fold, respectively, from the locked closed conformation, which has similar affinity as the wild-type I domain. Most surprisingly, the 2D affinity increases were due mostly to the 2D on-rate increases, as the 2D off-rates only decreased by severalfold. The wild-type alphaLbeta2, but not its I domain in isolation, could be up-regulated by Mn2+ or Mg2+ to have high affinities and on-rates. Locking the I domain in any of the three conformations abolished the ability of divalent cations to regulate 2D affinity. These results indicate that a downward displacement of the I domain C-terminal helix, induced by conformational changes of other domains of the alphaLbeta2, is required for affinity and on-rate up-regulation.

Integrins contain two structurally homologous but distantly related domains: an I-like domain that is present in all beta-subunits and an I domain that is present in some alpha-subunits. Atomic resolution and mutagenesis studies of alpha I domains demonstrate a C-terminal, axial displacement of the alpha7-helix that allosterically regulates the shape and affinity of the ligand-binding site. Atomic resolution studies of beta I-like domains have thus far demonstrated no similar alpha7-helix displacement; however, other studies are consistent with the idea that alpha I and beta I-like domains undergo structurally analogous rearrangements. To test the hypothesis that C-terminal, axial displacement of the alpha7-helix, coupled with beta6-alpha7 loop reshaping, activates beta I-like domains, we have mimicked the effect of alpha7-helix displacement on the beta6-alpha7 loop by shortening the alpha7-helix by two independent, four-residue deletions of about one turn of alpha-helix. In the case of integrin alphaLbeta2, each mutant exhibits constitutively high affinity for the physiological ligand intercellular adhesion molecule 1 and full exposure of a beta I-like domain activation-dependent antibody epitope. In the case of analogous mutants in integrin alpha4beta7, each mutant shows the activated phenotype of firm adhesion, rather than rolling adhesion, in shear flow. The results show that integrins that contain or lack alpha I domains share a common pathway of beta I-like domain activation, and they suggest that beta I-like and alpha I domain activation involves structurally analogous alpha7-helix axial displacements.

The ligand binding function of integrins can be modulated by various monoclonal antibodies by both direct and indirect mechanisms. We have characterized an anti-beta(1) antibody, SG/19, that had been reported to inhibit the function of the beta(1) integrin on the cell surface. SG/19 recognized the wild type beta(1) subunit that exists in a conformational equilibrium between the high and low affinity states but bound poorly to a mutant beta(1) integrin that had been locked in a high affinity state. Epitope mapping of SG/19 revealed that Thr(82) in the beta(1) subunit, located at the outer face of the boundary between the I-like and hybrid domains, was the key binding determinant for this antibody. Direct visualization of the alpha (5)beta(1) headpiece fragment in complex with SG/19 Fab with electron microscopy confirmed the location of the binding surface and showed that the ligand binding site is not occluded by the bound Fab. Surface plasmon resonance showed that alpha (5)beta(1) integrin bound by SG/19 maintained a low affinity toward its physiological ligand fibronectin (Fn) whereas binding by function-blocking anti-alpha(5) antibodies resulted in a complete loss of fibronectin binding. Thus a class of the anti-beta antibodies represented by SG/19 attenuate the ligand binding function by restricting the conformational shift to the high affinity state involving the swing-out of the hybrid domain without directly interfering with ligand docking.

We explore the binding sites for mAbs to the alpha I domain of the integrin alphaLbeta2 that can competitively inhibit, allosterically inhibit, or activate binding to the ligand ICAM-1. Ten mAbs, some of them clinically important, were mapped to species-specific residues. The results are interpreted with independent structures of the alphaL I domain determined in seven different crystal lattices and in solution, and which are present in three conformational states that differ in affinity for ligand. Six mAbs bind to adjacent regions of the beta1-alpha1 and alpha3-alpha4 loops, which show only small (mean, 0.8 angstroms; maximum, 1.8 angstroms) displacements among the eight I domain structures. Proximity to the ligand binding site and to noncontacting portions of the ICAM-1 molecule explains competitive inhibition by these mAbs. Three mAbs bind to a segment of seven residues in the beta5-alpha6 loop and alpha6 helix, in similar proximity to the ligand binding site, but on the side opposite from the beta1-alpha1/alpha3-alpha4 epitopes, and far from noncontacting portions of ICAM-1. These residues show large displacements among the eight structures in response to lattice contacts (mean, 3.6 angstroms; maximum, 9.4 angstroms), and movement of a buried Phe in the beta5-alpha6 loop is partially correlated with affinity change at the ligand binding site. Together with a lack of proximity to noncontacting portions of ICAM-1, these observations explain variation among this group of mAbs, which can either act as competitive or allosteric antagonists. One agonistic mAb binds distant from the ligand binding site of the I domain, to residues that show little movement (mean, 0.5 angstroms; maximum, 1.0 angstroms). Agonism by this mAb is thus likely to result from altering the orientation of the I domain with respect to other domains within an intact integrin alphaLbeta2 heterodimer.
Copyright 2004 The American Association of Immunologists, Inc.

The extracellular portions of cell surface receptor proteins are often comprised of independently folding protein domains. As they are translated into the endoplasmic reticulum (ER), some of these domains require protein chaperones to assist in their folding. Members of the low-density lipoprotein receptor (LDLR) family require the chaperone called Boca in Drosophila or its ortholog, Mesoderm development, in the mouse. All LDLRs have at least one six-bladed beta-propeller domain, which is immediately followed by an epidermal growth factor (EGF) repeat. We show here that Boca is specifically required for the maturation of these beta-propeller/EGF modules through the secretory pathway, but is not required for other LDLR domains. Protein interaction data suggest that as LDLRs are translated into the ER, Boca binds to the beta-propeller. Subsequently, once the EGF repeat is translated, the beta-propeller/EGF module achieves a more mature state that has lower affinity for Boca. We also show that Boca-dependent beta-propeller/EGF modules are found not only throughout the LDLR family but also in the precursor to the mammalian EGF ligand.

We test with molecular dynamics the hypothesis that interdomain forces in integrins, simulated with a spring attached to the C-terminal alpha 7-helix of an integrin I domain, can allosterically stabilize alternative I domain conformations. Depending on the force applied and timecourse, in alpha(L) and alpha(M) I domains the beta 6-alpha 7 loop moves successively between three ratchet positions; i.e. from closed to intermediate, and then to open. More distal, linked alterations in MIDAS loops and metal coordination closely resemble those seen when the MIDAS becomes ligated. Simulations show that the intermediate state is populated over a wider range of forces for alpha(L) than alpha(M) I domains. Simulations with mutant I domains suggest that specific ratchet residues regulate conformational equilibria. Simulations with alpha(1) and alpha(2) I domains reveal a lack of the intermediate conformation, owing to Phe to Glu substitution at the second ratchet residue. The findings have important implications for biological regulation of integrin adhesiveness.

Two activation-dependent Abs to the integrin alphaL-subunit were used to study conformational rearrangement of alphaLbeta2 on the cell surface. Activation lowered the concentration of Ca2+ required for maximal expression of each epitope. Each Ab requires the Ca2+-binding loop in the integrin genu and nearby species-specific residues in the thigh domain. Key thigh residues are shielded from Ab in the bent integrin conformation by the alpha-subunit calf-1 domain and the nearby bent beta leg, suggesting that extension at the genu is required for epitope exposure. Activating stimuli and alpha/beta I-like small molecule antagonists demonstrate that exposure of epitopes in the integrin alpha- and beta-subunit legs is coordinate during integrin activation. A coordinating residue donated by the calf-1 domain is as important as Ca2+ for mAb binding. Together with inspection of the alphaV structure, this result suggests that the genu/calf-1 interface is maintained in integrin activation, and that extension occurs by a rearrangement at the thigh/genu interface.

Abciximab, a derivative of the murine mAb 7E3, protects against ischemic complications of percutaneous coronary interventions by inhibiting ligand binding to the alphaIIbbeta3 receptor. In this study we identified regions on integrin beta3 that control 7E3 binding. Murine/human amino acid substitutions were created in two regions of the betaA domain that previous studies found to influence 7E3 binding: the C177-C184 loop and K125-N133. The T182N substitution and a K125Q mutation reduced 7E3 binding to human beta3 in complex with alphaIIb. The introduction of both the human C177-C184 region and human W129 into murine beta3 was necessary and sufficient to permit 7E3 binding to the human alphaIIb/murine beta3 complex. Although we cannot exclude allosteric effects, we propose that 7E3 binds between C177-C184 and W129, which are within 15 A of each other in the crystal structure and close to the beta3 metal ion-dependent adhesion site. We previously demonstrated that 7E3 binds more rapidly to activated than unactivated platelets. Because it has been proposed that alphaIIbbeta3 changes from a bent to an extended conformation upon activation, we hypothesized that 7E3 binds less well to the bent than the extended conformation. In support of this hypothesis we found that 7E3 bound less well to an alphaIIbbeta3 construct locked in a bent conformation, and unlocking the conformation restored 7E3 binding. Thus, our data are consistent with alphaIIbbeta3 existing in variably bent conformations in equilibrium with each other on unactivated platelets, and activation resulting in alphaIIbbeta3 adopting a more extended conformation.

The function of some multidomain proteins is regulated by interdomain communication. We use second-site suppressor cysteine mutations to test a hypothesis on how the inserted (I)-like domain in the integrin beta-subunit regulates ligand binding by the neighboring I domain in the integrin alpha-subunit [Huth, J. R., Olejniczak, E. T., Mendoza, R., Liang, H., Harris, E. A., et al. (2000) Proc. Natl. Acad. Sci. USA 97, 5231-5236; and Alonso, J. L., Essafi, M., Xiong, J. P., Stehle, T. & Arnaout, M. A. (2002) Curr. Biol. 12, R340-R342]. The hypothesis is that an interaction between the beta I-like metal ion-dependent adhesion site (MIDAS) and an intrinsic ligand in the linker following the alpha I domain, Glu-310, exerts a pull that activates the alpha I domain. Individual mutation of alpha(L) linker residue Glu-310 or beta(2) MIDAS residues Ala-210 or Tyr-115 to cysteine abolishes I domain activation, whereas the double mutation of alpha(L)-E310C with either beta(2)-A210C or beta(2)-Y115C forms a disulfide bond that constitutively activates ligand binding. The disulfide-bonded mutant is resistant to small molecule antagonists that bind to the beta I-like domain near its interface with the alpha I domain and inhibit communication between these domains but remains susceptible to small molecule antagonists that bind underneath the I domain alpha 7-helix and certain allosteric antagonistic antibodies. Thus, the alpha 7-helix and its linker are better modeled as a pull spring than a bell rope. The results suggest that alpha(L) residue Glu-310, which is universally conserved in all I domain-containing integrins, functions as an intrinsic ligand for the beta I-like domain, and that when integrins are activated, the beta I-like MIDAS binds to Glu-310, pulls the spring, and thereby activates the alpha I domain.

Although integrin alpha subunit I domains exist in multiple conformations, it is controversial whether integrin beta subunit I-like domains undergo structurally analogous movements of the alpha7-helix that are linked to affinity for ligand. Disulfide bonds were introduced into the beta(3) integrin I-like domain to lock its beta6-alpha7 loop and alpha7-helix in two distinct conformations. Soluble ligand binding, ligand mimetic mAb binding and cell adhesion studies showed that disulfide-bonded receptor alpha(IIb)beta(3)(T329C/A347C) was locked in a low affinity state, and dithiothreitol treatment restored the capability of being activated to high affinity binding; by contrast, disulfide-bonded alpha(IIb)beta(3)(V332C/M335C) was locked in a high affinity state. The results suggest that activation of the beta subunit I-like domain is analogous to that of the alpha subunit I domain, i.e. that axial movement in the C-terminal direction of the alpha7-helix is linked to rearrangement of the I-like domain metal ion-dependent adhesion site into a high affinity conformation.

Dynamic regulation of integrin adhesiveness is required for immune cell-cell interactions and leukocyte migration. Here, we investigate the relationship between cell adhesion and integrin microclustering as measured by fluorescence resonance energy transfer, and macroclustering as measured by high resolution fluorescence microscopy. Stimuli that activate adhesion through leukocyte function-associated molecule-1 (LFA-1) failed to alter clustering of LFA-1 in the absence of ligand. Binding of monomeric intercellular adhesion molecule-1 (ICAM-1) induced profound changes in the conformation of LFA-1 but did not alter clustering, whereas binding of ICAM-1 oligomers induced significant microclustering. Increased diffusivity in the membrane by cytoskeleton-disrupting agents was sufficient to drive adhesion in the absence of affinity modulation and was associated with a greater accumulation of LFA-1 to the zone of adhesion, but redistribution did not precede cell adhesion. Disruption of conformational communication within the extracellular domain of LFA-1 blocked adhesion stimulated by affinity-modulating agents, but not adhesion stimulated by cytoskeleton-disrupting agents. Thus, LFA-1 clustering does not precede ligand binding, and instead functions in adhesion strengthening after binding to multivalent ligands.

We examined the effect of conformational change at the beta(7) I-like/hybrid domain interface on regulating the transition between rolling and firm adhesion by integrin alpha(4)beta(7). An N-glycosylation site was introduced into the I-like/hybrid domain interface to act as a wedge and to stabilize the open conformation of this interface and hence the open conformation of the alpha(4) beta(7) headpiece. Wild-type alpha(4)beta(7) mediates rolling adhesion in Ca(2+) and Ca(2+)/Mg(2+) but firm adhesion in Mg(2+) and Mn(2+). Stabilizing the open headpiece resulted in firm adhesion in all divalent cations. The interaction between metal binding sites in the I-like domain and the interface with the hybrid domain was examined in double mutants. Changes at these two sites can either counterbalance one another or be additive, emphasizing mutuality and the importance of multiple interfaces in integrin regulation. A double mutant with counterbalancing deactivating ligand-induced metal ion binding site (LIMBS) and activating wedge mutations could still be activated by Mn(2+), confirming the importance of the adjacent to metal ion-dependent adhesion site (ADMIDAS) in integrin activation by Mn(2+). Overall, the results demonstrate the importance of headpiece allostery in the conversion of rolling to firm adhesion.

The small GTPase Rap1 induces integrin-mediated adhesion and changes in the actin cytoskeleton. The mechanisms that mediate these effects of Rap1 are poorly understood. We have identified RIAM as a Rap1-GTP-interacting adaptor molecule. RIAM defines a family of adaptor molecules that contain a RA-like (Ras association) domain, a PH (pleckstrin homology) domain, and various proline-rich motifs. RIAM also interacts with Profilin and Ena/VASP proteins, molecules that regulate actin dynamics. Overexpression of RIAM induced cell spreading and lamellipodia formation, changes that require actin polymerization. In contrast, RIAM knockdown cells had reduced content of polymerized actin. RIAM overexpression also induced integrin activation and cell adhesion. RIAM knockdown displaced Rap1-GTP from the plasma membrane and abrogated Rap1-induced adhesion. Thus, RIAM links Rap1 to integrin activation and plays a role in regulating actin dynamics.

In vivo, beta(2) integrins and particularly alpha(L)beta(2) (LFA-1) robustly support firm adhesion of leukocytes, but can also cooperate with other molecules in supporting rolling adhesion. Strikingly, a small molecule alpha/beta I-like allosteric antagonist, XVA143, inhibits LFA-1-dependent firm adhesion, while at the same time it enhances adhesion in shear flow and rolling both in vitro and in vivo. XVA143 appears to induce the extended conformation of integrins as shown by increased activation epitope exposure. Fab to the beta(2) I-like domain converts firm adhesion to rolling adhesion, but does not enhance adhesion. Residue alpha(L)-Glu-310 in the linker following the I domain is critical for communication to the beta(2) I-like domain, rolling, integrin extension, and activation by Mn(2+) of firm adhesion. The results demonstrate the importance of integrin extension in rolling, and suggest that rolling and firm adhesion are mediated by extended conformations of alpha(L)beta(2) that differ in the affinity of the alpha(L) I domain for ICAM-1.

Conformational communication across the plasma membrane between the extracellular and intracellular domains of integrins is beginning to be defined by structural work on both domains. However, the role of the alpha and beta subunit transmembrane domains and the nature of signal transmission through these domains have been elusive. Disulfide bond scanning of the exofacial portions of the integrin alpha(IIbeta) and beta(3) transmembrane domains reveals a specific heterodimerization interface in the resting receptor. This interface is lost rather than rearranged upon activation of the receptor by cytoplasmic mutations of the alpha subunit that mimic physiologic inside-out activation, demonstrating a link between activation of the extracellular domain and lateral separation of transmembrane helices. Introduction of disulfide bridges to prevent or reverse separation abolishes the activating effect of cytoplasmic mutations, confirming transmembrane domain separation but not hinging or piston-like motions as the mechanism of transmembrane signaling by integrins.

Integrins are important adhesion receptors in all Metazoa that transmit conformational change bidirectionally across the membrane. Integrin alpha and beta subunits form a head and two long legs in the ectodomain and span the membrane. Here, we define with crystal structures the atomic basis for allosteric regulation of the conformation and affinity for ligand of the integrin ectodomain, and how fibrinogen-mimetic therapeutics bind to platelet integrin alpha(IIb)beta3. Allostery in the beta3 I domain alters three metal binding sites, associated loops and alpha1- and alpha7-helices. Piston-like displacement of the alpha7-helix causes a 62 degrees reorientation between the beta3 I and hybrid domains. Transmission through the rigidly connected plexin/semaphorin/integrin (PSI) domain in the upper beta3 leg causes a 70 A separation between the knees of the alpha and beta legs. Allostery in the head thus disrupts interaction between the legs in a previously described low-affinity bent integrin conformation, and leg extension positions the high-affinity head far above the cell surface.