Alteration of the Fc RIIa Dimer Interface Affects

0 downloads 0 Views 328KB Size Report
binding chains and their associated ITAM-containing chains, the latter being covalent ... with 18 U.S.C. Section 1734 solely to indicate this fact. 1 This work was ...
The Journal of Immunology

Alteration of the Fc␥RIIa Dimer Interface Affects Receptor Signaling but Not Ligand Binding1 Maree S. Powell,* Nadine C. Barnes,* Tessa M. Bradford,* Ian F. Musgrave,† Bruce D. Wines,* John C. Cambier,‡ and P. Mark Hogarth2* The aggregation of cell surface FcRs by immune complexes induces a number of important Ab-dependent effector functions. However, despite numerous studies that examine receptor function, very little is known about the molecular organization of these receptors within the cell. In this study, protein complementation, mutagenesis, and ligand binding analyses demonstrate that human Fc␥RIIa is present as a noncovalent dimer form. Protein complementation studies found that Fc␥RIIa molecules are closely associated. Mutagenesis of the dimer interface, as identified by crystallographic analyses, did not affect ligand binding yet caused significant alteration to the magnitude and kinetics of receptor phosphorylation. The data suggest that the ligand binding and the dimer interface are distinct regions within the receptor, and noncovalent dimerization of Fc␥RIIa may be an essential feature of the Fc␥RIIa signaling cascade. The Journal of Immunology, 2006, 176: 7489 –7494.

C

lustering of cell surface Fc␥RIIa, a low affinity receptor for IgG, by immune complexes initiates a number of key Ig-dependent effector functions, including Ab-dependent cellular cytotoxicity, the uptake and destruction of pathogens, and also leads to the release of major inflammatory cytokines (1). Fc␥RIIa is the most broadly expressed activating leukocyte FcR and is unique to higher primates. Fc␥RIIa transgenic mice develop spontaneous autoimmune disease and are hypersensitive to Abinduced inflammatory reactions (2), demonstrating that this receptor plays a critical role in the modulation of inflammatory reactions in vivo. However, despite the studies addressing the biological function of this receptor, very little is known about the molecular organization of Fc␥RIIa (and FcRs generally) on the cell membrane. The major immunoreceptors involved in the activation of leukocytes such as the T and B Ag cell receptors, NK receptors, and most of the FcRs are multichain immune recognition receptors (MIRR),3 wherein ligand binding and signal transduction functions are performed by separate polypeptides (3, 4). The activation of leukocytes via these MIRR involves the aggregation of the ligand binding chains and their associated ITAM-containing chains, the latter being covalent, and therefore constitutive ITAM-containing dimers. For the activating FcRs, including the high-affinity IgE receptor Fc␧RI, the IgG receptors Fc␥RI and Fc␥RIII, and the IgA

*The Macfarlane Burnet Institute for Medical Research and Public Health Limited, Austin Health, Heidelberg, Victoria, Australia; †Department of Clinical and Experimental Pharmacology, Adelaide University, Adelaide, Australia; and ‡Integrated Department of Immunology, University of Colorado Health Sciences Centre, and National Jewish Medical Center, Denver, CO 80206 Received for publication March 22, 2005. Accepted for publication March 28, 2006. 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 National Health and Medical Research Council Project Grant 18165 and Prima Biomed. 2 Address correspondence and reprint requests to Dr. P. Mark Hogarth, The Macfarlane Burnet Institute for Medical Research and Public Health Limited, Austin Health, Studley Road, Heidelberg, Victoria, 3084 Australia. E-mail address: [email protected] 3 Abbreviations used in this paper: MIRR, multichain immune recognition receptor; HAGG, heat-aggregated ␥-globulin; HT, hypoxanthine and thymidine; DHFR, dihydrofolate reductase; TR-MTX, Texas Red-conjugated methotrexate; s, soluble.

Copyright © 2006 by The American Association of Immunologists, Inc.

receptor Fc␣RI, Ig interacts with the appropriate ␣ subunit, which noncovalently associates with the common FcR-␥ chain, which itself is a covalent dimer containing ITAMs (1, 3). Fc␥RIIa is the exception to this, because both IgG binding and the signal transducing ITAMs are present in the one polypeptide (5, 6). Furthermore, given the otherwise universal requirement for MIRR ITAMbased signaling to be mediated by covalent signaling dimers, it is surprising that Fc␥RIIa does not exist as a covalent dimer. However, crystallographic studies of Fc␥RIIa extracellular regions suggest that noncovalent dimerization may be possible (7). In the study presented herein, protein complementation assays, mutagenesis, and functional analyses demonstrate that Fc␥RIIa can exist as a constitutive dimer in cells and mutation of the dimer interface results in diminished receptor phosphorylation and calcium mobilization, but does not affect ligand binding. These data suggest that organized association of Fc␥RIIa monomers is essential for optimal receptor signaling.

Materials and Methods Antibodies Heat-aggregated ␥-globulin (HAGG) was produced by heating Sandoglobulin (Novartis) at 63°C for 1 h before separation of monomeric IgG from aggregates using Superdex G200 gel filtration chromatography (Amersham Biosciences).

DNA expression constructs and transfection The human Fc␥RIIa-S129P and Fc␥RIIa-S129A mutant cDNAs were generated by QuickChange site-directed mutagenesis (Stratagene) of the HFc3.0 Fc␥RIIa cDNA template (8). For expression in IIA1.6 cells, the cDNAs encoding wild-type Fc␥RIIa, Fc␥RIIa-S129P, or Fc␥RIIa-S129A were subcloned into the retroviral vector pMXI-egfp (9). For the production of recombinant soluble Fc␥RIIa-S129P protein, a termination codon was introduced at Val171 (10). The cDNA was subcloned into baculovirus transfer vector pFASTbac (Invitrogen Life Technologies). Receptor:dihydrofolate reductase (DHFR) expression constructs used in the protein complementation assays were generated as follows: the cDNA encoding the extracellular and transmembrane domains of Fc␥RIIa or Fc␣RI (aa 1–171 and 1–226, respectively) were ligated into the zip-DHFR [1, 2] or zip-DHFR [3] expression plasmids (11, 12). The final expression constructs Fc␥RIIa-DHFR [1, 2], Fc␣RI-DHFR [1, 2], and Fc␥RIIa-DHFR [3] are linked to the DHFR subunit via a 10-aa poly-glycine linker. High-titer retroviruses were generated in the Phoenix packaging cell line (9, 13) for the expression of FcRs in IIA1.6 (FcR-deficient) cells (14) 0022-1767/06/$02.00

Fc␥RIIa DIMERIZATION AND FUNCTION

7490 according to the manufacturer’s specifications. For protein complementation studies, Chinese hamster ovary cells, DG44 (15), were seeded 24 h before transfection with 105 cells per well in a six-well plate in ␣-MEM (Invitrogen Life Technologies) supplemented with 10 ␮g/ml hypoxanthine and thymidine (HT). The cells were transfected with the Fc␥RIIa-DHFR [1, 2] or Fc␣RI-DHFR [1, 2] and/or Fc␥RIIa-DHFR [3] chimeric cDNAs using LipofectAMINE 2000 (Invitrogen Life Technologies) according to the manufacturer’s specifications. Cells were twice sorted using a MoFlo (DakoCytomation), and Fc␥RIIa or Fc␣RI expression was examined using mAb 8.2 or A59, respectively (10, 16).

DHFR complementation assays Proliferation assay. DHFR transfectants were seeded in quadruplicate at 2 ⫻ 103 cells in HT-deficient ␣-MEM, containing 0.5 ␮Ci of [3H]thymidine (Amersham Biosciences). The cultures were pulsed with [3H]thymidine for 8 h at 37°C and 10% CO2, harvested, and [3H]thymidine incorporation was counted. Fluorescence microscopy. Reassembly of DHFR was evaluated by fluorescence microscopy using Texas Red-conjugated methotrexate (TRMTX; Molecular Probes) as described elsewhere (11, 12). Briefly, FcRDHFR transfectants (5 ⫻ 105 cells/ml) were incubated with TR-MTX (5 ␮M final) for 2 h, washed, and reincubated for 30 min in ␣-MEM (HT deficient), before phenol red-free medium supplemented with 10% FCS and 1 ␮g of bisbenzimide H33342 fluorochrome (Calbiochem) was added to visualize nuclei. Fluorescent microscopy was performed on live cells using a Leica DC200 microscope and ⫻50/0.75 objective. The exposure and gain settings on the DC200 microscope were identical for all images.

Results Protein complementation assay indicates Fc␥RIIa self-association The possible physical association between Fc␥RIIa molecules as suggested by our previous crystallography studies (7) was investigated using the DHFR protein-fragment complementation assay wherein the receptor-dependent reassembly of complementary, nonfunctional fragments of DHFR reconstitutes enzyme function (11, 12). The reassembly of DHFR was monitored in vitro by fluorescence microscopy and cellular proliferation assays. Importantly, the reassembled DHFR binds with high-affinity TR-MTX in a 1:1 complex. The TR-MTX is only retained in cells where the DHFR has been reconstituted, whereas unbound TR-MTX is actively transported from the cell (12). The extracellular and transmembrane domains of Fc␥RIIa were fused to the complementary fragments of DHFR (DHFR [1, 2] and/or DHFR [3]), to create Fc␥RIIa-DHFR [1, 2] and Fc␥RIIa-DHFR [3], respectively. These were transfected individually or jointly into the DHFR-deficient DG44 fibroblasts. As a control for nonspecific enzyme reassembly, the extracellular and transmembrane domains of Fc␣RI were fused to the DHFR [1, 2] fragment (Fc␣RI-DHFR [1, 2]) and transfected jointly with Fc␥RIIa-DHFR [3] into DG44 fibroblasts. Expression

Detection of receptor phosphorylation Retroviral transduced IIA1.6 cell lines (1 ⫻ 107 cells/ml) expressing Fc␥RIIa, Fc␥RIIa-S129P, or Fc␥RIIa-S129A were stimulated with aggregated IgG (0 –100 ␮g/ml) and then lysed for 10 min on ice in 0.5 ml of lysis buffer containing 1% Brij-96, 150 mM NaCl, 10 mM Tris-HCl (pH 7.5), containing 1 mM pervanadate (Sigma-Aldrich) and protease mixture (Roche Molecular Biochemicals). The lysate was clarified by centrifugation at 10,000 ⫻ g for 10 min at 4°C, and the receptor was immunoprecipitated with anti-Fc␥RII mAb, 8.2-coated Sepharose at 4°C for 2 h. The 8.2 mAb beads were washed, and bound proteins were analyzed by Western blotting following SDS-PAGE and transfer of the protein to Immobilon polyvinylidene difluoride membranes (Munktell) according to the manufacturer’s specifications. Receptor phosphorylation was detected using the anti-phosphotyrosine Ab 4G10 conjugated with HRP. Detection of receptor was achieved using rabbit anti-Fc␥RIIa polyclonal Ab followed by antirabbit Ig/HRP (DakoCytomation).

Calcium mobilization Intracellular Ca2⫹ mobilization was examined in IIA1.6 cell lines expressing Fc␥RIIa, Fc␥RIIa-S129P, or pMXI vector only. Cells were washed and resuspended at 107 cells/ml in loading buffer (138 mM NaCl, 6 mM KCl, 1 mM MgSO4, 1 mM CaCl2, 1 mM Na2HPO4, 5 mM NaHCO3, 5.5 mM glucose, 20 mM HEPES, supplemented with 1% (w/v) BSA (pH 7.4; CSL Biosciences)). Cells were loaded with fura 2-AM (Molecular Probes), washed twice before resuspension in loading buffer supplemented with 1 mM Ca2⫹, to a final concentration of 2 ⫻ 106 cells/ml. Cells were preincubated for 2 min at 37°C, and basal calcium levels were monitored for 1 min before adding stimuli: 10 ␮g/ml rabbit anti-mouse IgG (RAM-IgG; Zymed Laboratories), or 0 –100 ␮g/ml HAGG (17). Fluorescence was measured using a Hitachi F2000 spectrophotometer and [Ca2⫹]i values were calculated according to Grynkiewicz et al. (18).

Analysis of Fc␥RIIa:IgG interaction The interaction between recombinant soluble (s)Fc␥RIIa or sFc␥RIIaS129P and monomeric IgG1 and IgE (irrelevant binding control) (SigmaAldrich) was investigated by surface plasmon resonance using a BIAcore 2000 biosensor (Amersham Biosciences) at 22°C in HEPES-buffered saline (10 mM HEPES (pH 7.4), 150 mM NaCl, 3.4 mM EDTA, 0.005% Surfactant P20) (Amersham Biosciences) as described elsewhere. Briefly, monomeric IgG1 or IgE were coupled over three or one channel, respectively (19). Recombinant sFc␥RIIa protein (0.001–1.2 mg/ml) were injected over the sensor-chip surface for 1 min at 20 ␮l/min followed by a 3-min dissociation phase. Equilibrium binding constants (KD) were obtained by nonlinear curve fitting of a single-site binding equation. Nonspecific binding responses (IgE channel) were subtracted from binding to the IgG1 channels. At the highest concentration of sFc␥RIIa, no more than 20 response units were detected over the IgE channel.

FIGURE 1. Analysis of DG44 fibroblasts expressing FcR-DHFR chimeras by flow cytometry (A–D), by TR-MTX binding and fluorescence microscopy (E and F), and by cellular proliferation assays (G). Flow-cytometric analysis of DG44 cells: cells cotransfected with Fc␥RIIa-DHFR [1, 2] and Fc␥RIIa-DHFR [3] were stained with FITC-conjugated antimouse Ig alone (A); cells cotransfected with Fc␥RIIa-DHFR [1, 2] and Fc␥RIIa-DHFR [3] and expression detected by anti-Fc␥RIIa (CD32) mAb, 8.2 (B); cell surface expression of Fc␣RI-DHFR [1, 2] as measured by anti-Fc␣RI (CD89) PE staining (C); cell surface expression of cotransfected Fc␣RI-DHFR [1, 2] and Fc␥RIIa-DHFR [3] measured by antiFc␣RI (CD89) PE and anti-Fc␥RIIa (CD32) mAb, 8.2 FITC staining (D). Fluorescence microscopy: DG44 cells stained with TR-MTX and expressing either Fc␥RIIa-DHFR [1, 2] and Fc␥RIIa-DHFR [3] (E) or Fc␣RIDHFR [1, 2] and Fc␥RIIa-DHFR [3] (F). Cell proliferation: Reconstitution of DHFR was monitored by proliferation in nucleotide-free medium (G) in cells cotransfected with Fc␥RIIa-DHFR [1, 2] and Fc␥RIIa-DHFR [3], or Fc␣RI-DHFR [1, 2] and Fc␥RIIa-DHFR [3] or with one Fc␥RIIa-DHFR fragment; Fc␥RIIa-DHFR [1, 2]. The values represent the mean ⫾ SD of three independent experiments.

The Journal of Immunology of the FcR-DHFR fusion proteins on the cell membrane of the transfected fibroblasts was confirmed using anti-Fc␥RIIa (Fig. 1, B and D) or anti-Fc␣RI (C and D) mAbs. Successful DHFR reconstitution was monitored by two methods: first, fluorescence assays by binding of TR-MTX to cells (Fig. 1E), and second, by cellular proliferation assays in nucleotide-free medium (Fig. 1G). In the fluorescence assays, binding of the TR-MTX was only evident in cells cotransfected with both Fc␥RIIa-DHFR [1, 2] and Fc␥RIIaDHFR [3] where the distribution of staining is evident throughout the cell (Fig. 1E). Importantly, cells cotransfected with Fc␣RIDHFR [1, 2] plus Fc␥RIIa-DHFR [3] failed to bind the fluorescent MTX (Fig. 1F). In the second approach, proliferation of cells in nucleotide-free medium was only evident in the cells cotransfected with both Fc␥RIIa-DHFR [1, 2] and Fc␥RIIa-DHFR [3] (20,000 cpm [3H]thymidine incorporated), and not in those transfected with only one fusion partner: Fc␥RIIa-DHFR [1, 2]. Most importantly, there was no proliferation of the control cotransfection, i.e., Fc␣RI-DHFR [1, 2] plus Fc␥RIIa-DHFR [3] (Fig. 1G). This is a sensitive control for as few as 25 reconstituted molecules per cell are required for cell survival in this system (11, 12). Thus, the substantial expression for the F␥RIIa and Fc␣RI fusions on the cell surface (Fig. 1) makes it extremely unlikely that random intramembrane associations are responsible for the reconstitution observed in the Fc␥RIIa cotransfections (Fig. 1, E and G). Together, the lack of TR-MTX binding or the proliferation of cells cotransfected with Fc␣RI-DHFR [1, 2] plus Fc␥RIIa-DHFR [3] indicates that reconstitution of DHFR was dependent on self-association between Fc␥RIIa rather than random or nonspecific association between surface molecules. Mutation within the Fc␥RIIa dimer interface disrupts signal transduction In the Fc␥RIIa dimer observed in x-ray studies, the C⬘CFG strands and C⬘C loop of the second domains of each receptor monomer

FIGURE 2. Fc␥RIIa dimerization. A, Ribbon diagram of the crystallographic Fc␥RIIa dimer viewed down the 2-fold axis. Ser129 is located within the C⬘ strand of D2 located at the interface between the two receptor monomers (solid green circles). B, Details of the Fc␥RIIa dimer interface. Hydrogen bonds are shown as dashed lines (magenta) with distances shown in angstroms. Amino acid Ser129 forms a main chain contact with Leu162 from the neighboring receptor and Gly127 interacts with Ser164 from the neighboring receptor. C, Model of predicted disruption of dimer contacts of Fc␥RIIa. Mutagenesis of Ser129 to Pro is predicted to inhibit the formation of both main chain contacts with the adjacent receptor monomer at amino acids Ser129 and Gly127, because the polypeptide backbone is predicted to adopt a configuration that limits these interactions. The dashed black line indicates a distance of 3.7 Å that is greater than normally allowed for hydrogen bond formation (⬍3.2 Å). The model using Fc␥RIIa dimer structure was subjected to restrained energy minimization and molecular dynamics simulations using the Insight II program (Accelrys).

7491 (Fig. 2A) are interacting surfaces (7), wherein Ser129 and Gly127, located in the C⬘ strand form hydrogen bonds with Leu162 and Ser164, respectively, in the adjacent monomer (Fig. 2B). Mutation to change Ser129 to Pro (S129P) is predicted to disrupt main chain contacts between adjacent receptor molecules especially around residues Gly127, Lys128, and Ser129, thereby disrupting intermolecular interactions between the Fc␥RIIa monomers (Fig. 2C). Cells expressing the mutated receptor (Fc␥RIIa-S129P) were established using retroviral transduction, and the effect of S129P mutation on IgG binding and signal transduction was examined. Flow-cytometric analyses using the anti-Fc␥RIIa monoclonal 8.2 confirmed equivalent levels of wild-type and mutant receptor expression on the cell surface (Fig. 3A). Importantly, the Fc␥RIIaS129P mutation did not affect the binding of aggregated IgG with similar levels of HAGG bound to cells expressing Fc␥RIIa and Fc␥RIIa-S129P (Fig. 3B). Quantitative equilibrium dissociation experiments confirmed this result by demonstrating that recombinant soluble forms of Fc␥RIIa and Fc␥RIIa-S129P bound to immobilized monomeric IgG with equivalent affinities (Fig. 3C) (KD ⫽ 1.00 ⫾ 0.01 and 1.03 ⫾ 0.03 ␮M, respectively). These data are consistent with the crystallographic data that identify the dimer interface and IgG binding surfaces as distinct regions of the receptor (7). Cellular effects of alterations to the dimer interface, including ligand-induced receptor phosphorylation as an early signaling event and calcium mobilization as a late signaling event were also examined. Mutation at the dimer interface affected both the magnitude and kinetics of receptor phosphorylation (Fig. 4, A–C). Over a broad range of concentrations of aggregated IgG (up to 100 ␮g/ ml), phosphorylation of Fc␥RIIa-S129P was consistently lower than that of wild-type Fc␥RIIa (Fig. 4A). Interestingly, kinetic analysis of receptor phosphorylation using a fixed concentration (50 ␮g/ml) of aggregated IgG demonstrated that Fc␥RIIa-S129P phosphorylation was similar to wild-type Fc␥RIIa at the early time

7492

Fc␥RIIa DIMERIZATION AND FUNCTION

FIGURE 3. Mutagenesis to the dimer interface does not affect the Fc␥RIIa ligand binding site. Cell surface expression of Fc␥RIIa and Fc␥RIIa-S129P in IIA1.6 B cells was equivalent as determined by the surface expression of Fc␥RIIa (solid line), Fc␥RIIa-S129P (broken line), and pMXI only (solid histogram) using antireceptor mAb 8.2 (A); ligand binding of Fc␥RIIa (solid line), Fc␥RIIa-S129P (broken line), or pMXI only (solid histogram) using HAGG (B); and Langmuir plots of Fc␥RIIa (F) and Fc␥RIIa-S129P (䡺) binding to immobilized IgG1 (C). Mutagenesis at the dimer interface does not affect IgG binding because equivalent affinities for ligand were demonstrated for Fc␥RIIa (KD, 1.00 ⫾ 0.01 ␮M; n ⫽ 5) and Fc␥RIIa-S129P (KD, 1.03 ⫾ 0.03 ␮M; n ⫽ 3). The Langmuir plots are representative of several experiments.

FIGURE 4. Tyrosine phosphorylation and calcium mobilization responses following aggregated IgG stimulation of Fc␥RIIa and Fc␥RIIa-S129P. A, The Fc␥RIIa and Fc␥RIIa-S129P cell lines were stimulated with 0 –100 ␮g/ml aggregated IgG for 2 min, receptor immunoprecipitated, and analyzed for phosphotyrosine and Fc␥RIIa staining using anti-phosphotyrosine mAb 4G10 and an anti-receptor polyclonal Ab. B, Using 50 ␮g/ml aggregated IgG, the phosphorylation pattern of Fc␥RIIa and Fc␥RIIa-S129P was examined over time (0, 0.5, 1, 5, 10 min). Data shown are representative of seven experiments. C, Densitometry values for the kinetic analyses of Fc␥RII (solid line) and Fc␥RIIa-S129P (dashed line) phosphorylation over time. D, Peak Ca2⫹ mobilization responses by Fc␥RIIa (⽧), Fc␥RIIa-S129P (f), or pMXi transduction of IIA1.6 cells (Œ) following stimulation with 100 ␮g/ml aggregated IgG. E, The Fc␥RIIa-S129P cell line (u) was unable to reach equivalent peak levels of intracellular Ca2⫹ mobilization compared with the Fc␥RIIa cell line (f), over a 2 log range of IgG concentrations. F, RAM-IgG stimulation of Fc␥RIIa-S129P (f) and wild-type receptor (⽧) transfectants. The values represent the mean ⫾ SD of four independent experiments. The SD was generally 10% of the mean, and SD bars have been omitted for clarity.

The Journal of Immunology

FIGURE 5. Tyrosine phosphorylation following aggregated IgG stimulation of Fc␥RIIa and Fc␥RIIa-S129A. The Fc␥RIIa and Fc␥RIIa-S129A cell lines were stimulated with 50 ␮g/ml aggregated IgG over time (0, 0.5, 1 min), receptor immunoprecipitated, and analyzed for phosphotyrosine and Fc␥RIIa staining using anti-phosphotyrosine mAb 4G10 and an antireceptor polyclonal Ab. Data shown are representative of three experiments.

points (Fig. 4, B and C). However, Fc␥RIIa-S129P phosphorylation was rapidly diminished with minimal phosphorylation detected at 5–10 min. This differs substantially from the kinetics of Fc␥RIIa phosphorylation, which peaked at 5 min and slowly declined to low levels by 10 min (Fig. 4, B and C). A more conservative mutation of Ser129 to Ala (Ala being present in the equivalent position in Fc␥RI and Fc␧RI) found that ligand-induced receptor phosphorylation was similarly affected (Fig. 5), further highlighting the importance of this amino acid position within the dimer interface. Intracellular calcium responses were also diminished over a 2 log range of IgG concentrations in Fc␥RIIa-S129P-expressing cells compared with the wild-type receptor (Fig. 4, D and E). At the highest aggregated IgG concentration, the peak calcium response from Fc␥RIIa was 414 ⫾ 17 nM Ca2⫹ compared with 307 ⫾ 6 nM Ca2⫹ for the Fc␥RIIa-S129P mutant (Fig. 4, D and E). The difference in release of calcium was significant at lower concentrations of ligand, e.g., at 12.5 ␮g/ml aggregated IgG, there was a 70% reduction in mobilization (i.e., 374 ⫾ 15 nM for Fc␥RIIa vs 122 ⫾ 16 nM for Fc␥RIIa-S129P) (Fig. 4E). The reduction in calcium mobilization by Fc␥RIIa-S129P was not due to a generalized deficiency within these cells, because activation of the B cell Ag receptor with anti-IgG resulted in robust calcium responses from both cell lines. Indeed, the cells expressing Fc␥RIIa-S129P gave a greater Ag receptor-dependent calcium response than stimulation of IIA1.6 cells expressing the wild-type receptor (Fig. 4F).

Discussion The combination of protein complementation assays, mutagenesis, and cell signaling indicate the Fc␥RIIa exists within the cell as a dimer, even in the absence of its IgG ligand, which is consistent with the previous x-ray crystallographic studies (7). The inability of Fc␥RIIa-S129P to signal efficiently (despite no loss of IgG binding) implies that Fc␥RIIa self-association is required for optimal function. The introduction of the S129P mutation did not detectably affect Fc␥RIIa-driven protein complementation assays (data not shown); however, in this system, we cannot rule out changes to the rate of dimerization or even qualitative changes to the nature of the dimerization. We cannot exclude the possibility of other regions within Fc␥RIIa that help coordinate the organization of this receptor. Indeed, for other immunoreceptors, the transmembrane domain has proven to be crucially important in the organization of receptor assemblies (20, 21) (B. D. Wines, H. M. Trist, P. A. Ramsland, and P. M. Hogarth, unpublished results). It remains to

7493 be determined whether this applies to the organization of Fc␥RIIa within the cell. The crystallographic dimer of Fc␥RIIa predicts that the cytoplasmic tails may be separated by 42 Å (7), thereby allowing possible transphosphorylation between the ITAM of each cytoplasmic tail (12) and the cooperative recruitment of nonreceptor tyrosine kinases achieving signaling functions that are common to the MIRR (22, 23). The noncovalent dimerization of Fc␥RIIa is an interesting evolutionary adaptation given that all other ITAM signaling subunits are covalent dimers. However, it is notable that normal signaling by engineered CD16:TCR ␨-chain chimeras was possible despite the absence of the disulfide bond within the ␨-chain sequence (24). The retention of the gpA-like motif within the ␨-chain transmembrane sequence would ensure that such chimeras were noncovalent dimers like Fc␥RIIa in this study. The data demonstrated that Fc␥RIIa forms a homodimer that is necessary for optimal signal transduction. However, the role of receptor dimerization remains unclear. Dimerization alone was not sufficient to induce signal transduction, which is consistent with the well-established dogma that ligand-induced aggregation of Fc␥RIIa (and FcR generally) is required for the full spectrum of cellular activation (23). Nonetheless, before ligand-induced aggregation, the organized association of Fc␥RIIa dimers is an essential component of the receptor signaling cascade, and aggregation of dimers is a speculative possibility. The extent to which similar self-associations exist in other FcRs is unknown. Because all other activating FcRs are MIRR associated with FcR-␥ chain dimers, receptor dimerization of the type seen for Fc␥RIIa may not be necessary for optimal signal transduction for these FcR; indeed, the Fc␥RIIa interface is not well conserved among other FcRs. However, Fc␥RIIb, the low-affinity inhibitory FcR shares an almost identical dimer interface with Fc␥RIIa, including Ser129, raising the possibility that receptor heterodimerization may act as a mechanism of regulating the cellular functions of these receptors.

Acknowledgments We thank Drs. Stephen Michnick and Paul Ramsland for valuable discussions and help with the preparation of this manuscript.

Disclosures The authors have no financial conflict of interest.

References 1. Daeron, M. 1997. Fc receptor biology. Annu. Rev. Immunol. 15: 203–234. 2. Tan Sardjono, C., P. L. Mottram, N. C. van de Velde, M. S. Powell, D. Power, R. F. Slocombe, I. P. Wicks, I. K. Campbell, S. E. McKenzie, M. Brooks, et al. 2005. Development of spontaneous multisystem autoimmune disease and hypersensitivity to antibody-induced inflammation in Fc␥ receptor IIa-transgenic mice. Arthritis Rheum. 52: 3220 –3229. 3. Isakov, N. 1998. Role of immunoreceptor tyrosine-based activation motif in signal transduction from antigen and Fc receptors. Adv. Immunol. 69: 183–247. 4. Dykstra, M., A. Cherukuri, and S. K. Pierce. 2001. Rafts and synapses in the spatial organization of immune cell signaling receptors. J. Leukocyte Biol. 70: 699 –707. 5. Hogarth, P. M. 2002. Fc receptors are major mediators of antibody based inflammation in autoimmunity. Curr. Opin. Immunol. 14: 798 – 802. 6. Indik, Z. K., J. G. Park, S. Hunter, and A. D. Schreiber. 1995. The molecular dissection of Fc␥ receptor mediated phagocytosis. Blood 86: 4389 – 4399. 7. Maxwell, K. F., M. S. Powell, M. D. Hulett, P. A. Barton, I. F. McKenzie, T. P. Garrett, and P. M. Hogarth. 1999. Crystal structure of the human leukocyte Fc receptor, Fc ␥RIIa. Nat. Struct. Biol. 6: 437– 442. 8. Hibbs, M. L., L. Bonadonna, B. M. Scott, I. F. McKenzie, and P. M. Hogarth. 1988. Molecular cloning of a human immunoglobulin G Fc receptor. Proc. Natl. Acad. Sci. USA 85: 2240 –2244. 9. Nolan, G. P., and A. R. Shatzman. 1998. Expression vectors and delivery systems. Curr. Opin. Biotechnol. 9: 447– 450. 10. Ierino, F. L., M. S. Powell, I. F. McKenzie, and P. M. Hogarth. 1993. Recombinant soluble human Fc␥RII: production, characterization, and inhibition of the Arthus reaction. J. Exp. Med. 178: 1617–1628. 11. Remy, I., and S. W. Michnick. 1999. Clonal selection and in vivo quantitation of protein interactions with protein-fragment complementation assays. Proc. Natl. Acad. Sci. USA 96: 5394 –5399.

7494 12. Remy, I., I. A. Wilson, and S. W. Michnick. 1999. Erythropoietin receptor activation by a ligand-induced conformation change. Science 283: 990 –993. 13. Kitamura, T., M. Onishi, S. Kinoshita, A. Shibuya, A. Miyajima, and G. P. Nolan. 1995. Efficient screening of retroviral cDNA expression libraries. Proc. Natl. Acad. Sci. USA 92: 9146 –9150. 14. Jones, B., J. P. Tite, and C. A. Janeway, Jr. 1986. Different phenotypic variants of the mouse B cell tumor A20/2J are selected by antigen- and mitogen-triggered cytotoxicity of L3T4-positive, I-A-restricted T cell clones. J. Immunol. 136: 348 –356. 15. Urlaub, G., P. J. Mitchell, E. Kas, L. A. Chasin, V. L. Funanage, T. T. Myoda, and J. Hamlin. 1986. Effect of gamma rays at the dihydrofolate reductase locus: deletions and inversions. Somat. Cell Mol. Genet. 12: 555–566. 16. Wines, B. D., H. M. Trist, R. C. Monteiro, C. Van Kooten, and P. M. Hogarth. 2004. Fc receptor ␥ chain residues at the interface of the cytoplasmic and transmembrane domains affect association with Fc␣RI, surface expression, and function. J. Biol. Chem. 279: 26339 –26345. 17. Musgrave, I. F., R. Seifert, and G. Schultz. 1994. Maitotoxin activates cation channels distinct from the receptor-activated non-selective cation channels of HL-60 cells. Biochem. J. 301: 437– 441.

Fc␥RIIa DIMERIZATION AND FUNCTION 18. Grynkiewicz, G., M. Poenie, and R. Y. Tsien. 1985. A new generation of Ca2⫹ indicators with greatly improved fluorescence properties. J. Biol. Chem. 260: 3440 –3450. 19. Powell, M. S., P. A. Barton, D. Emmanouilidis, B. D. Wines, G. M. Neumann, G. A. Peitersz, K. F. Maxwell, T. P. Garrett, and P. M. Hogarth. 1999. Biochemical analysis and crystallisation of Fc␥RIIa, the low affinity receptor for IgG. Immunol. Lett. 68: 17–23. 20. Schamel, W. W., and M. Reth. 2000. Monomeric and oligomeric complexes of the B cell antigen receptor. Immunity 13: 5–14. 21. Garrity, D., M. E. Call, J. Feng, and K. W. Wucherpfennig. 2005. The activating NKG2D receptor assembles in the membrane with two signaling dimers into a hexameric structure. Proc. Natl. Acad. Sci. USA 102: 7641–7646. 22. Jouvin, M. H., M. Adamczewski, R. Numerof, O. Letourneur, A. Valle, and J. P. Kinet. 1994. Differential control of the tyrosine kinases Lyn and Syk by the two signaling chains of the high affinity immunoglobulin E receptor. J. Biol. Chem. 269: 5918 –5925. 23. Pribluda, V. S., C. Pribluda, and H. Metzger. 1994. Transphosphorylation as the mechanism by which the high-affinity receptor for IgE is phosphorylated upon aggregation. Proc. Natl. Acad. Sci. USA 91: 11246 –11250. 24. Romeo, C., M. Amiot, and B. Seed. 1992. Sequence requirements for induction of cytolysis by the T cell antigen/Fc receptor ␨ chain. Cell 68: 889 – 897.