Search

We recently showed that ligand-mediated cross-linking of FcepsilonRI, the high-affinity receptor for immunoglobulin E, on RBL-2H3 mast cells results in its co-isolation with detergent-resistant membranes (DRM) and its consequent tyrosine phosphorylation by the co-localized tyrosine kinase Lyn that is a critical early event in signaling by this receptor [Field et al. (1997) J. Biol. Chem. 272, 4276-4280]. As part of efforts to determine the structural bases for these interactions, we examined the phospholipid composition of DRM vesicles isolated from RBL-2H3 cells under conditions that preserve FcepsilonRI association. We used positive and negative mode electrospray Fourier transform ion cyclotron resonance mass spectrometry to compare quantitatively the phospholipid composition of isolated DRM to that of total cell lipids and to a plasma membrane preparation. From these analyses, over 90 different phospholipid species were spectrally resolved and unambiguously identified; more than two-thirds of these were determined with a precision of +/-0.5% (absolute) or less. Quantitative characterization of lipid profiles shows that isolated DRM are substantially enriched in sphingomyelin and in glycerophospholipids with a higher degree of saturation as compared to total cellular lipids. Plasma membrane vesicles isolated from RBL-2H3 cells by chemically induced blebbing exhibit a degree of phospholipid saturation that is intermediate between DRM and total cellular lipids, and significant differences in the headgroup distribution between DRM and plasma membranes vesicles are observed. DRM from cells with cross-linked FcepsilonRI exhibit a larger ratio of polyunsaturated to saturated and monounsaturated phospholipids than those from unstimulated cells. Our results support and strengthen results from previous studies suggesting that DRM have a lipid composition that promotes liquid-ordered structure. Furthermore, they demonstrate the potential of mass spectrometry for examining the role of membrane structure in receptor signaling and other cellular processes.

Aggregation of Fc epsilonRI, the high-affinity receptor for IgE, on RBL-2H3 mast cells caused by reversible ligands such as multivalent antigen causes cellular responses that can be halted by subsequent addition of excess monovalent ligand. In contrast, Ca2+ and degranulation responses elicited by effectively irreversible streptavidin cross-linking of biotinylated IgE-Fc epsilonRI are not stopped by addition of excess biotin after stimulation is initiated. These results support previous conclusions based on studies with covalent oligomers of IgE that stable cross-links can continue to deliver stimulatory signals for extended periods of time. Dissociation measured in the presence of monovalent hapten reveals two populations of IgE-Fc epsilonRI cross-linked by multivalent antigen that differ in functional effectiveness. Aggregates with readily dissociable cross-links are normally responsible for triggering essentially all of the degranulation response, whereas aggregates with poorly dissociable cross-links apparently do not trigger this response. Treatment of RBL-2H3 cells with cytochalasin D, an inhibitor of actin polymerization, enhances downstream signaling and enables the less readily dissociable aggregates to stimulate Ca2+ and degranulation responses. Under these conditions, cytochalasin D does not affect hapten-mediated dissociation of multivalent antigen, nor does it prevent hapten from reversing tyrosine phosphorylation of Syk. Cytochalasin D alone causes tyrosine phosphorylation of a protein at approximately 75 kDa, and it reduces hapten-induced reversal of antigen-stimulated tyrosine phosphorylation of several other proteins. Taken together, these results indicate that stimulated actin polymerization normally regulates the coupling of aggregated Fc epsilonRI to downstream signaling pathways, and they provide an explanation for seeming discrepancies between responses to stable and reversible cross-links.

Previous studies in our laboratory established that the symmetrical bivalent ligand, N,N'-bis-[[epsilon-(2,4-dinitrophenyl)amino]caproyl]-L-tyrosyl]-L-cystin e ((DCT)2-cys), stably cross-links anti-2,4-dinitrophenyl-immunoglobulin E (IgE) bound to high affinity receptors Fc epsilonRI on the surface of RBL-2H3 cells, forming mostly cyclic dimers containing two IgE-Fc epsilonRI and two (DCT)2-cys (Posner et al. (1995) J. Immunol. 155, 3601-3609). These cyclic dimers do not trigger Ca2+ or degranulation responses under a variety of conditions. However, we find that the linearly cross-linked IgE-Fc epsilonRI formed at higher concentrations of (DCT)2-cys do trigger degranulation in the presence of cytochalasin D, an inhibitor of actin polymerization. We further investigated stimulation by (DCT)2-cys of the earliest known events in the functional response, i.e., tyrosine phosphorylation of the beta and gamma subunits of Fc epsilonRI. At the higher (DCT)2-cys concentrations corresponding to linear dimers and maximal degranulation, tyrosine phosphorylation of both beta and gamma are observed. At lower (DCT)2-cys concentrations where cross-linking is maximal and cyclic dimers are overwhelmingly dominant, only gamma tyrosine phosphorylation is observed. Cytochalasin D does not affect these phosphorylation patterns, but instead appears to enhance coupling to downstream signaling events. Phosphorylation of Syk occurs at the higher (DCT)2-cys concentrations in parallel with beta phosphorylation but does not occur in its absence at the lower (DCT)2-cys concentrations. These results suggest that cyclic dimers of IgE-Fc epsilonRI are sterically restricted such that they stimulate tyrosine phosphorylation of gamma but not beta, and this is not sufficient for Syk binding and/or activation.

Erythrosin-labeled immunoglobulin E (IgE) and time-resolved phosphorescence anisotropy were used to monitor the rotational dynamics of transfected wild-type (alpha beta gamma 2) and four mutant Fc epsilon RI receptors in the monomeric and dimeric state on P815 cells. Erythrosin-IgE bound to Fc epsilon RI on cells transfected with either beta or gamma subunits with truncated COOH-terminal cytoplasmic segments exhibit faster rotational motion than when bound to Fc epsilon RI on cells transfected with wild-type subunits. Deletion of the NH2-terminal cytoplasmic segment of the beta subunit or the COOH-terminal cytoplasmic segment of the alpha subunit does not cause any significant change in the anisotropy decay. Dimers of IgE-receptor complexes formed with anti-IgE monoclonal antibody B1E3 exhibit substantially slower anisotropy decays for all the receptor constructs used, including a receptor construct that only contains the ectodomain of the alpha subunit anchored to Chinese hamster ovary (CHO) cell membranes through a lipid tail. This loss of rotational motion of dimeric IgE-Fc epsilon RI complexes may be due to nonspecific entanglement or to specific interactions involving IgE or the extracellular portion of alpha. The results suggest that the beta and gamma subunits of the tetrameric alpha beta gamma 2 receptor participate in interactions with other membrane components even in the absence of receptor aggregation. The loss of such interactions may be related to the functional impairments previously determined for these mutants.

We have used sonicated liposomes of phosphatidylcholine (PC), sphingomyelin (SM), or a mixture of cholesterol (chol) and PC to investigate the role of cellular lipid composition in Fc epsilon RI-mediated stimulation of RBL-2H3 cells. Overnight treatment with either PC or SM liposomes causes a substantial enhancement of antigen-stimulated degranulation and phospholipase A2 activity, whereas treatment with a PC/chol mixture results in partial inhibition of the antigen-stimulated response. The most consistent change in the cellular lipid composition that results from the PC and SM liposome treatments is an approximate 40% decrease in the chol/phospholipid (PL) ratio. The lipid treatments do not alter degranulation stimulated by AlF4- or by Ca2+ ionophore in the presence or absence of PMA, suggesting that lipid alteration affects a receptor-specific signaling process. The lipid treatments do not appear to alter antigen-stimulated tyrosine phosphorylation or Ca2+ mobilization. Possible involvement of protein kinase C (PKC) activation in the signal-enhancing effect of the PL treatments was investigated by using calphostin C and phorbol-12-myristol-13-acetate (PMA) to inhibit PKC activity and degranulation in RBL-2H3 cells. Both SM and PC treatment restore the antigen-mediated degranulation response that is inhibited by long-term treatment (> or = 16 h) with 100 nM PMA or short-term treatment (10 min) with 5 microM calphostin C. The results indicate that a decreased chol/PL ratio facilitates or enhances the receptor-mediated activation of a PKC-like pathway that plays an important role in Fc epsilon RI-stimulated degranulation.(ABSTRACT TRUNCATED AT 250 WORDS)

Store-operated Ca(2+) entry (SOCE) is a ubiquitous signaling process in eukaryotic cells in which the endoplasmic reticulum (ER)-localized Ca(2+) sensor, STIM1, activates the plasma membrane-localized Ca(2+) release-activated Ca(2+) (CRAC) channel, Orai1, in response to emptying of ER Ca(2+) stores. In efforts to understand this activation mechanism, we recently identified an acidic coiled-coil region in the C-terminus of Orai1 that contributes to physical association between these two proteins, as measured by fluorescence resonance energy transfer, and is necessary for Ca(2+) influx, as measured by an intracellular Ca(2+) indicator. Here, we present evidence that a positively charged sequence of STIM1 in its CRAC channel activating domain, human residues 384-386, is necessary for activation of SOCE, most likely because this sequence interacts directly with the acidic coiled coil of Orai1 to gate Ca(2+) influx. We find that mutation to remove positive charges in these residues in STIM1 prevents its stimulated association with wild-type Orai1. However, association does occur between this mutant STIM1 and Orai1 that is mutated to remove negative charges in its C-terminal coiled coil, indicating that other structural features are sufficient for this interaction. Despite this physical association, we find that thapsigargin fails to activate SOCE following coexpression of mutant STIM1 with either wild type or mutant Orai1, implicating STIM1 residues 384-386 in transmission of the Ca(2+) gating signal to Orai1 following store depletion.

Dynamic conformations of two distinct immunoglobulin (Ig) isotypes, murine IgE and human IgG1, were examined with fluorescence resonance energy transfer measurements. The IgE mutant epsilon/C gamma 3* and the IgG1 mutant gamma/C gamma 3* each bind [5-(dimethylamino)naphthalen-1-yl]sulfonyl (DNS) in two identical antigen binding sites at the amino (N)-terminal ends of the Ig in the Fab segments. Eosin-DNS bound in these Fab sites served as the acceptor probe in these studies. Both Ig have a carboxy (C)-terminal domain (C gamma 3*) which contains genetically introduced cysteine residues. Modification of these cysteine sulfhydryls with fluorescein maleimide provided donor probes near the C-terminal ends of the Ig in the Fc segment. Energy transfer between the C-terminal and N-terminal ends was compared for these two Ig in solution and when they were found to their respective high-affinity receptors on plasma membranes: IgE-Fc epsilon RI on RBL cell membranes and IgG1-Fc gamma RI on U937 cell membranes. Previous energy-transfer measurements with these probes yielded an average end-to-end distance of 71 A for IgE in solution and 69 A for IgE bound to Fc epsilon RI, indicating that in both situations IgE is bent such that the axes of the Fab segments and the axis of the Fc segment do not form a planar Y-shape [Zheng, Shopes, Holowka, & Baird (1991) Biochemistry 30, 9125]. In the current study we found the average end-to-end distance for IgG1 in solution is 75 A and greater than or equal to 85 A for IgG1 bound to Fc gamma RI, suggesting an average bend conformation for IgG1 as well. The contributions of segmental flexibility to the average distances were assessed directly by measuring the efficiency of energy transfer as a function of variations in donor quantum yield caused by a collisional quencher and using these data to extract a Gaussian distribution of end-to-end distances. The distribution average (rho) and half-width (hw) were determined to be as follows: rho = 75 A, hw = 24 A for IgE in solution; rho = 71 A, hw = 12 A for IgE bound to Fc epsilon RI; and rho = 100 A, hw = 88 A for IgG in solution.(ABSTRACT TRUNCATED AT 400 WORDS)

We use electrospray ionization mass spectrometry to quantify100 phospholipid (PL) components in detergent-resistant membrane (DRM) domains that are related to ordered membrane compartments commonly known as lipid rafts. We previously compared PL compositions of DRMs with plasma membrane vesicles and whole cell lipid extracts from RBL mast cells, and we made the initial observation that antigen stimulation of IgE receptors (FcepsilonRI) causes a significant change in the PL composition of DRMs [Fridriksson, E. K., et al. (1999) Biochemistry 38, 8056-8063]. We now characterize the signaling requirements and time course for this change, which is manifested as an increase in the recovery of polyunsaturated PL in DRM, particularly in phosphatidylinositol species. We find that this change is largely independent of tyrosine phosphorylation, stimulated by engagement of FcepsilonRI, and can be activated by Ca(2+) ionophore in a manner independent of antigen stimulation. Unexpectedly, we found that inhibitors of actin polymerization (cytochalasin D and latrunculin A) cause a similar, but more rapid, change in the PL composition of DRMs in the absence of FcepsilonRI activation, indicating that perturbations in the actin cytoskeleton affect the organization of plasma membrane domains. Consistent with this interpretation, a membrane-permeable stabilizer of F-actin, jasplakinolide, prevents antigen-stimulated changes in DRM PL composition. These results are confirmed by a detailed analysis of multiple experiments, showing that receptor and cytochalasin D-stimulated changes in DRM lipid composition follow first-order kinetics. Analysis in terms of the number of double bonds in the fatty acid chains is valid for total PL of the major headgroups and for headgroups individually. In this manner, we show that, on average, concentrations of saturated or monounsaturated PL decrease in the DRM, whereas concentrations of PL with two or more double bonds (polyunsaturated PL) increase due to cytoskeletal perturbation. We find that these changes are independent of fatty acid chain length. Our mass spectrometric analyses provide a detailed accounting of receptor-activated alterations in the plasma membrane that are regulated by the actin cytoskeleton.

The bivalent ligand N,N'-bis[[epsilon-[(2,4- dinitrophenyl)amino]caproyl]-L-tyrosyl]cystine [(DCT)2-Cys] binds and cross-links anti-dinitrophenyl (DNP) immunoglobulin E (IgE)-receptor complexes on the cell surface of rat basophilic leukemia cells. The rate of dissociation of this bound ligand was monitored by using a fluorescence method under two different conditions. In one case the monovalent ligand DCT was added in large excess to prevent the dissociating ligand from rebinding to unoccupied antibody combining sites. Under these conditions, dissociation of the bivalent ligand from IgE-sensitized cells proceeds to completion with kinetics that are well described by two rate constants that are independent of the time of preincubation of the bivalent ligand with the cells. In the second case, dissociation of (DCT)2-Cys from cell-bound anti-DNP IgE was monitored in the presence of a large excess of anti-DNP IgE in solution that acts as a sink to absorb the dissociated ligand. Under these conditions, the bivalent ligand becomes more resistant to dissociation as the preincubation time of the bivalent ligand with the cells is increased. An increasing fraction of the bound ligand does not dissociate on a measurable time scale in the presence of this sink. The results indicate that cell-associated IgE-receptor complexes undergo a time-dependent change that facilitates the reformation of the cross-linked state when one end of the ligand dissociates to break up the existing cross-link. The possible physical basis and functional implications of these results are discussed.

Time-resolved fluorescence anisotropy measurements have been used to examine the segmental flexibility of anti-dansyl immunoglobulin E (IgE) bound to its high-affinity receptor on membrane vesicles from rat basophilic leukemia cells. Although IgE in this complex exhibits only a restricted angular range of segmental motion, much of this restricted motion occurs on a relatively rapid time scale. A fast component of motion with a rotational correlation time of 15-35 ns may correspond to the twisting of Fab segments about their major axis. Intermolecular cross-linking by a short bivalent ligand, N,N'-didansylcadaverine, results in complete loss of this segmental motion. Solubilization of monomeric IgE-receptor complexes using a zwitterionic detergent results in a time-dependent anisotropy decay that exhibits both a fast component and a slower component that is intermediate between the decay for soluble and membrane-bound forms of IgE at long times after excitation. These results are discussed in terms of a model in which binding of IgE to its membrane-bound receptor restricts not only its global rotation but also its slower modes of segmental flexibility as well, while allowing its Fab segments to undergo rapid reorientation within a limited angular range.

A series of synthetic ligands consisting of poly(ethylene glycol) (PEG), capped on one or both ends with the hapten 2,4-dinitrophenyl (DNP), were previously shown to be potent inhibitors of cellular activation in RBL mast cells stimulated by a multivalent antigen [Baird, E. J., Holowka, D., Coates, G. W., and Baird, B. (2003) Biochemistry 42, 12739-12748]. In this study, we systematically investigated the effect of increasing length of the PEG scaffold on the binding of these monovalent and bivalent ligands to anti-DNP IgE in solution. Our analysis reveals evidence for an energetically favorable interaction between two monovalent ligands bound to the same receptor, when the PEG molecular mass exceeds approximately 5 kDa. Additionally, for ligands with much higher molecular masses (>10 kDa PEG), the binding of a single ligand apparently leads to a steric exclusion of the second binding site by the bulky PEG scaffold. These results are further corroborated by data from an alternate fluorescence-based assay that we developed to quantify the capacity of these ligands to displace a small hapten bound to IgE. This new assay monitors the displacement of a small, receptor-bound hapten by a competitive monovalent ligand and thus quantifies the competitive inhibition offered by a monovalent ligand. We also show that, for bivalent ligands, inhibitory capacity is correlated with the capacity to form effective intramolecular cross-links with IgE.

Architectural features of synthetic ligands were systematically varied to optimize inhibition of mast cell degranulation initiated by multivalent crossing of IgE-receptor complexes. A series of ligands were generated by end-capping poly(ethylene glycol) (PEG) polymers and amine-based dendrimers with the hapten 2,4-dinitrophenyl (DNP). These were used to explore the influence of polymeric backbone length, valency, and hapten presentation on binding to anti-DNP IgE and inhibition of stimulated activation of RBL cells. Monovalent MPEG(5000)-DNP (IC(50) = 50 nM), bivalent DNP-PEG(3350)-DNP (IC(50) = 8 nM), bismonovalent MPEG(5000)-DNP(2) (IC(50) = 20 nM), bisbivalent DNP(2)-PEG(3350)-DNP(2) (IC(50) = 3nM) and DNP(4)-dendrimer ligands (IC(50) = 50 nM) all effectively inhibit cellular activation caused by multivalent antigen, DNP-bovine serum albumin. For different DNP ligands, we provide evidence for more effective inhibition due to (i) preferential formation of intra-IgE cross-links by bivalent ligands of sufficient length, (ii) self-association of monovalent ligands with longer tails, and (iii) higher probability of binding for bisvalent ligands. We also show that larger DNP(16)-dendrimers of higher valency trigger degranulation by cross-linking IgE-receptor complexes, whereas smaller DNP-dendrimers are inhibitory. Thus, features of synthetic ligands can be manipulated to control receptor occupation, aggregation, and inhibition of the cellular response.

Specialized plasma membrane domains known as lipid rafts participate in signal transduction and other cellular processes, and their liquid ordered (L(o)) phase appears to be important for their function. To quantify ordered lipids in biological membranes, we investigated steady-state fluorescence anisotropy of two lipid probes, 2-[3-(diphenylhexatrienyl)propanoyl]-1-hexadecanoyl-sn-glycero-3-phosphocholine (DPH-PC) and N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)-1,2-dihexadecanoyl-sn-glycero-3-phosphoethanolamine (NBD-PE). We show using model membranes with varying amounts of cholesterol that steady-state fluorescence anisotropy is a sensitive measure of cholesterol-dependent ordering. The results suggest that DPH-PC is a more sensitive probe than NBD-PE. In the presence of cholesterol, ordering also depends on the degree of saturation of the phospholipid acyl chains. Using DPH-PC, we find that the plasma membrane of RBL-2H3 mast cells is substantially ordered, roughly 40%, as determined by comparison with anisotropy values for model membranes entirely in a liquid ordered (L(o)) phase and in a liquid disordered (L(alpha)) phase. This result is consistent with the finding that approximately 30% of plasma membrane phospholipids are insoluble in 0.5% Triton X-100. Furthermore, detergent-resistant membranes isolated by sucrose gradient fractionation of Triton X-100 cell lysates are more ordered than plasma membrane vesicles, suggesting that they represent a more ordered subset of the plasma membrane. Treatment of plasma membrane vesicles with methyl-beta-cyclodextrin resulting in 75% cholesterol depletion leads to commensurate decreases in lipid order as measured by anisotropy of DPH-PC and NBD-PE. These results demonstrate that steady-state fluorescence anisotropy of DPH-PC is a useful way to measure the amount of lipid order in biological membranes.

Posttranslational glycosylation is critical for biological function of many proteins, but its structural characterization is complicated by natural heterogeneity, multiple glycosylation sites, and different forms. Here, a top-down mass spectrometry (MS) characterization is applied to three constructs of the Fc segment of IgE: Fcepsilon(3-4) (52 kDa) and Fcepsilon(2-3-4)(2) (76 kDa) disulfide-bonded homodimers. Fourier transform MS of a reduced sample of Fcepsilon(2-3-4) gave molecular masses of 37 527, 37 689, 37 851, and 38 014 Da, directly characterizing multiple glycoforms (hexose = 162 Da) without chromatographic separation. Limited proteolysis of the nonreduced Fcepsilon(2-3-4)(2) protein yielded a peptide mixture with molecular weight values that agreed with those expected from the DNA sequence. The single glycosylation site in these constructs was identified, and quantities were determined of five glycoforms that agreed within +/-2% of the molecular ion values. The 2-D mass spectrum of two glycosylated peptides showed these to have high-mannose structures, -GlcNAc-(hex)(n)(), demonstrating that Fcepsilon(2-3-4) has a single such structure of n = 5-9. For a mutated sample of Fcepsilon(3-4), in addition to five glycoforms, MS showed a molecular discrepancy that could be assigned with proteolysis and 2-D mass spectra to the oxidation of two methionines and an additional residue difference.

To investigate IL-1-dependent interactions of IL-1 type I (IL-1 RI) receptors on intact cells, lateral and rotational mobilities and detergent insolubility were investigated. Lateral mobility was measured by fluorescence photobleaching recovery, using a Cy3-modified, noncompetitive mAb specific for IL-1RI (M5) bound to wild-type IL-1 RI or mutant IL-1 RI with a truncated cytoplasmic tail. Addition of IL-1 causes significant reduction in the mobile fraction of wild-type IL-1 RI for two different transfected cell lines. For the mutant IL-1 RI, no significant decrease in response to IL-1 is observed, indicating that the missing cytoplasmic segment is involved in IL-1-dependent interactions of IL-1 RI that lead to reduced lateral mobility on the cell surface. The rotational mobility of IL-1 RI was assessed with phosphorescence anisotropy decay measurements using erythrosin-labeled M5. IL-1 decreases the rotational mobility of cell surface IL-1 RI on the microsecond time scale and also increases the initial anisotropy, indicating loss in segmental motion. Measurements of resistance to solubilization by Triton X-100 showed that IL-1 binding increases the fraction of IL-1 RI sedimenting with cytoskeletal residues. The IL-1 receptor antagonist protein (IL-1ra) causes partial effects in reducing rotational mobility and increasing detergent insolubility of M5-lableled IL-1 RI, indicating that this ligand causes structural changes in the presence of the dimerizing M5 mAb. These ligand-dependent physical interactions of IL-1 RI on the cell surface may be related to signal initiation by this receptor.

The aggregation of IgE anchored to high-affinity Fc epsilon receptors on rat basophilic leukemia (RBL) cells by multivalent antigens initiates transmembrane signaling and ultimately cellular degranulation. Previous studies have shown that the rate of dissociation of bivalent and multivalent DNP ligands from RBL cells sensitized with anti-DNP IgE decreases with increasing ligand incubation times. One mechanism proposed for this effect is that when IgE molecules are aggregated, a conformational change occurs that results in an increase in the intrinsic affinity of IgE for antigen. This possibility was tested by measuring the equilibrium constant for the binding of monovalent DNP-lysine to anti-DNP IgE under two conditions, where the cell-bound IgE is dispersed and where it has been aggregated into visible patches on the cell surface using anti-IgE and a secondary antibody. No difference in the equilibrium constant in these two cases was observed. We also measured the rate of dissociation of a monovalent ligand from cell surface IgE under these two conditions. Whereas the affinity for monovalent ligand is not altered by IgE aggregation, we observe that the rate of ligand dissociation from IgE in clusters is slower than the rate of ligand dissociation from unaggregated IgE. These results are discussed in terms of recent theoretical developments concerning effects of receptor density on ligand binding to cell surfaces.

Erythrosin 5'-thiosemicarbazide labeled immunoglobulin E (IgE) was used to monitor the rotational dynamics of monomeric and dimeric Fc epsilon RI receptors for IgE on rat basophilic leukemia (RBL) basophilic leukemia (RBL) cells using time-resolved phosphorescence anisotropy. Receptors were studied both on living RBL cells and on membrane vesicles derived from RBL cell plasma membrane. The un-cross-linked IgE-receptor complexes on cells and vesicles exhibit rotational correlation times that are consistent with those expected for freely rotating monomers, but a small fraction of these complexes on cells may be rotationally immobile. A comparison of the initial phosphorescence anisotropy values for erythrosin-labeled IgE-receptor complexes on cells and vesicles reveals a fast component of rotational motion that is greater on the vesicles and may be due to a site of segmental flexibility in the receptor itself. Dimers of IgE-receptor complexes formed with anti-IgE monoclonal antibodies appear to be largely immobile on cells, but they are mobile on vesicles with a 2-fold larger rotational correlation time than the monomeric complexes. The results suggest that dimeric IgE-receptor complexes undergo interactions with other membrane components on intact cells that do not occur on the membrane vesicles. The possible significance of these interactions to receptor function is discussed.

Previous resonance energy transfer studies suggested that murine immunoglobulin E (IgE) is bent near the junction of its Fc and Fab segments when bound to its high-affinity receptor (Fc epsilon RI) on RBL cells. To examine further the conformations of IgE, both bound to this receptor and in solution, a mutant recombinant IgE (epsilon/C gamma 3*) was prepared that has a cysteine replacing a serine near the C-terminal ends of the heavy chain. The introduced cysteine residues provide a means for specific modification of IgE, and the sulfhydryl groups were selectively labeled with fluorescein-5-maleimide (FM-epsilon/C gamma 3*). This IgE also binds a 5-(dimethylamino)naphthalene-1-sulfonyl (DNS) group in the antigen-binding sites. Resonance energy transfer experiments carried out on receptor-bound FM-epsilon/C gamma 3* yielded a distance of 53 A between fluorescein near the C-terminal end of the Fc segment and amphipathic acceptor probes at the membrane surface. The average distance between this C-terminal fluorescein and acceptor eosin-DNS in the antigen-binding sites at the N-terminal ends of the Fab segments was found to be 69 A. These results combine with those from previous structural studies to provide an unprecedented detailed description of the bent geometry of IgE bound to its receptor on the membrane. Energy transfer measured for FM-epsilon/C gamma 3* in solution between fluorescein near the C-terminal end of the Fc segment and eosin-DNS at the N-terminal ends of the Fab segments indicates that the average distance between these probes is about 71 A.(ABSTRACT TRUNCATED AT 250 WORDS)

We study the dissociation of preformed bivalent ligand-bivalent receptor aggregates in solution, where the ligand is a symmetric bivalent hapten with two identical 2,4-dinitrophenyl (DNP) groups and the receptor is a fluorescein-labeled monoclonal anti-DNP IgE. We promote dissociation in two ways: by the addition of high concentrations of a monovalent hapten that competes for IgE binding sites with the bivalent hapten and by the addition of high concentrations of unlabeled IgE that binds almost all ligand binding sites that dissociate from labeled IgE. We investigate both theoretically and experimentally the two types of dissociation and find them to be quite different. Theory predicts that their kinetics will depend differently on the fundamental rate constants that characterize binding and aggregation. Using monovalent ligand to promote dissociation, we find that the fraction of labeled IgE sites bound to bivalent ligand decays with a slow and fast component. The fast decay corresponds to the dissociation of a singly bound DNP hapten. The interpretation of the slow decay depends on the detailed way in which ligand-receptor aggregates break up. We show that one possible explanation of these data is that small stable rings form before the addition of monovalent ligand. Other possible explanations are also presented.

Resonance energy-transfer methods have been used to investigate the structure of immunoglobulin E (IgE) bound to its high-affinity receptor on plasma membrane vesicles derived from rat basophilic leukemia cells. The structural mapping of receptor-bound IgE was initiated in an earlier study [Holowka, D., & Baird, B. (1983) Biochemistry 22, 3475], and it is based on measuring the minimal distance from IgE sites that are selectively labeled with donor probes to a plane of amphipathic acceptors at the membrane surface. This paper describes the use of monoclonal IgE specific for 5-(dimethylamino)naphthalene-1-sulfonyl (DNS) to place a donor probe, DNS-L-Lys, in the antibody combining sites. The distance from these sites to the membrane surface was determined to be greater than 100 A with two different amphipathic acceptor probes. Another site in the Fab segments of monoclonal IgE (anti-dinitrophenyl) could be labeled selectively with N-[4-[7-(diethylamino)-4-methylcoumarin-3-yl]phenyl]maleimide (CPM) in the absence of reducing agents [CPM(-)], and the reaction could not be blocked by prereaction with N-ethylmaleimide. The pattern of CPM(-)-labeled proteolytic fragments and the lack of fluorescence quenching by (trinitrophenyl)lysine in the antibody combining sites suggested the CPM(-)-labeled site to be in the C epsilon 1 domain of IgE. The distance between this site on receptor-bound IgE and the membrane surface was determined to be 75-87 A with two different amphipathic acceptors.(ABSTRACT TRUNCATED AT 250 WORDS)

The segmental flexibility of mouse immunoglobulin E (IgE) bound to its high-affinity receptor on membrane vesicles from rat basophilic leukemia cells was compared to that of IgE in solution by measuring the steady-state anisotropy as a function of temperature and viscosity. A monoclonal IgE was used to bind the fluorescent probe N-[5-(dimethylamino)naphthalene-1-sulfonyl]-L-lysine (DNS-Lys) rigidly and specifically in the antigen combining site at the tip of the Fab region. The average rotational correlation time, phi, of 74-89 ns for the receptor-bound IgE is only slightly longer than that for IgE in solution where phi = 54 ns. Another mouse monoclonal IgE was covalently labeled in the Fab region with N-(1-pyrenyl)maleimide. Anisotropy measurements with this derivative yielded results that are very similar to those found with anti-DNS IgE and DNS-Lys. These findings are strikingly different from that expected for a rigid IgE bound to its receptor since in this case phi is likely to be very much larger. Evidently, the segmental flexibility of IgE is not greatly altered upon binding to its receptor.

Previous resonance energy-transfer measurements have suggested that immunoglobulin E (IgE) may bend near the junction of its Fc and Fab segments in order to bind to its high-affinity receptor on rat basophilic leukemia cells. In order to test this possibility, two monoclonal antibodies were employed that bind specifically to rat IgE (IgER) when IgER is in solution and when it is bound to receptors on the plasma membrane. The F(ab')2 fragment of one monoclonal (B5) that is specific for the Fab region of IgER was labeled with donor probes and bound to IgER, and the quenching of the fluorescence of these donors due to simultaneous binding of the Fab' fragment of an anti-Fc monoclonal (A2) that was labeled with an acceptor probe at its interchain disulfide bond was measured. Significantly less energy transfer between these probes was observed when IgER was bound to its receptor on membrane vesicles than when it was free in solution, and this result is interpreted in light of other energy-transfer measurements using A2 and B5 that were preferentially labeled near their combining sites with donors and acceptors, respectively, as well as measurements of the distance of closest approach between these sites and the membrane surface. These results along with previous energy-transfer measurements and other biochemical information form the basis for a working model of the conformation and orientation of receptor-bound IgE. This study demonstrates the use of fluorescently labeled monoclonal antibodies as highly selective energy-transfer probes in assessing structures of macromolecular complexes on the plasma membrane.

In order to investigate the properties of the membrane-bound IgE-receptor complex, a simple procedure has been adapted for preparing large plasma membrane vesicles from rat basophilic leukemia cells. These vesicles pinch off from the adherent cells after treatment with 2 mM N-ethylmaleimide or 50 mM formaldehyde and 1 mM dithiothreitol, and they are isolated from the supernatant after two centrifugation steps with yields of 20-25% of the initial cell-bound 125I-IgE. With phase and fluorescence microscopy, micron-size vesicles are seen which are unilamellar and spherically shaped and devoid of intracellular organelles. On dextran gradients at least 70% of the 125I-IgE is bound to membranes which band at low density, indicating large, intact vesicles that are impermeable to macromolecules. Between 60 and 75% of the bound 125I-IgE is accessible to the external medium, showing the vesicles to be predominantly right side out. This preparation was found to be suitable for resonance energy-transfer measurements. We have determined that amphipathic, fluorescent donor and acceptor probes partition into the vesicle bilayer in a randomly distributed, noninteracting manner. The densities of the probes can be ascertained directly from the amount of energy transfer that is observed as a function of acceptor concentration. This experimental system will allow energy-transfer measurements to determine distances between sites on receptor-bound IgE and the membrane surface.