Functional Association between the Human Myeloid Immunoglobulin A Fc Receptor (CD89) and FcR [IMAGE] Chain

FcR γ chain has previously been shown to interact with the TCR-CD3 complex, the IgE Fc receptor I (FcεRI), and the class I and IIIA IgG receptors (FcγRI and FcγRIIIa). Here, we demonstrate that the Fc receptor γchain associates with FcαR in transfected IIA1.6 B lymphocytes. FcαR could be expressed at the surface of IIA1.6 B cells by itself, but was devoid of signaling capacity. Upon co-expression of FcR γchain, a physical interaction with FcαR could be demonstrated. This association proved crucial for the triggering of both proximal (intracellular calcium increase and tyrosine phosphorylation), as well as distal (IL-2 release), signal transduction responses. We next tested the hypothesis that a positively charged arginine residue (Arg209) within the transmembrane domain of FcαR promotes association with FcR γchain. We therefore constructed FcαR molecules where Arg209 was mutated to either a positively charged histidine, a negatively charged aspartic acid, or an uncharged leucine. A functional association between FcαR and FcR γchain was observed only with a positively charged residue (Arg209 or His209) present within the FcαR transmembrane domain. These data show that transmembrane signal transduction by the FcαR is mediated via FcR γchain, and that FcαR requires a positively charged residue within the transmembrane domain to promote functional association.

FcR ␥ chain has previously been shown to interact with the TCR-CD3 complex, the IgE Fc receptor I (Fc⑀RI), and the class I and IIIA IgG receptors (Fc␥RI and Fc␥RIIIa). Here, we demonstrate that the Fc receptor ␥ chain associates with Fc␣R in transfected IIA1.6 B lymphocytes. Fc␣R could be expressed at the surface of IIA1.6 B cells by itself, but was devoid of signaling capacity. Upon co-expression of FcR ␥ chain, a physical interaction with Fc␣R could be demonstrated. This association proved crucial for the triggering of both proximal (intracellular calcium increase and tyrosine phosphorylation), as well as distal (IL-2 release), signal transduction responses. We next tested the hypothesis that a positively charged arginine residue (Arg 209 ) within the transmembrane domain of Fc␣R promotes association with FcR ␥ chain. We therefore constructed Fc␣R molecules where Arg 209 was mutated to either a positively charged histidine, a negatively charged aspartic acid, or an uncharged leucine. A functional association between Fc␣R and FcR ␥ chain was observed only with a positively charged residue (Arg 209 or His 209 ) present within the Fc␣R transmembrane domain. These data show that transmembrane signal transduction by the Fc␣R is mediated via FcR ␥ chain, and that Fc␣R requires a positively charged residue within the transmembrane domain to promote functional association.
IgA is the primary immunoglobulin in bodily secretions and plays a critical role in protection against the constant environmental challenges at mucosal sites. Although the protective mechanisms are incompletely defined, a significant role in IgAmediated immune defense has been proposed for IgA Fc receptors. These molecules have been detected on most populations of phagocytic cells in blood and mucosal tissues. Engagement of these molecules can trigger phagocytosis, degranulation, oxidative burst, inflammatory mediator release, and antibody-dependent cellular cytotoxicity (1,2). Fc␣R on monocytes/macrophages and neutrophils has been defined as a 55-75-kDa glycoprotein (3,4), whereas the eosinophil Fc␣R is more heavily glycosylated (70 -100 kDa) (5). Both types of myeloid receptors are recognized by the CD89 mAb 1 panel (4,5) and bind both IgA1 and IgA2 via their Fc regions (1). The cDNA encoding the myeloid Fc␣R has been characterized and was found to encode a 30-kDa peptide, with two extracellular Ig-like domains, a hydrophobic transmembrane region and a cytoplasmic tail devoid of recognized signaling motifs (2,6). Additionally, we have recently isolated and characterized the human gene encoding the CD89 molecule. The gene structure indicates Fc␣R to represent a more distantly related member of the immunoglobulin receptor gene family (7).
To explore the capacity of Fc␣R to trigger biological functions, we have now generated different transfectants in the mouse IIA1.6 B cell line. This line, derived from the A20 B cell lymphoma, lacks the 5Ј-end of the Fc␥RII gene and, consequently, is Fc receptor-negative (8). Previous work showed this line to represent an excellent model for assaying Fc␥R-mediated functioning (9 -11). Following transfection, Fc␣R was expressed at the surface of IIA1.6 cells by itself, but lacked signaling capacity. We therefore hypothesized that Fc␣R may associate with a specialized signaling molecule. The FcR ␥ chain was known previously to associate with all three classes of Fc␥R, Fc⑀RI, and the TCR-CD3 complex (12)(13)(14)(15). FcR ␥ chain is responsible for coupling these receptors to intracellular signaling pathways (16). By co-transfection experiments, we tested whether FcR ␥ chain could mediate signal transduction via Fc␣R. Our results show that co-expression of Fc␣R and FcR ␥ chain in IIA1.6 B cells conferred both proximal and distal signaling capacity to Fc␣R. During the preparation of this manuscript, it was shown that, in the U937 cell line, Fc␣R was associated with FcR ␥ chain and that ␥ chain was phosphorylated on tyrosine residues following Fc␣R cross-linking (17). The data presented here indeed confirm that Fc␣R associates with FcR ␥ chain in transfected IIA1.6 B cells and, furthermore, suggest that Fc␣R and FcR ␥ chain can associate in normal blood PMN. The present experiments demonstrate FcR ␥ chain to be critical for Fc␣R-mediated transmembrane signal transduction.
We, furthermore, explored the molecular basis for Fc␣R/FcR ␥ chain association. The transmembrane (TM) domain of Fc␣R is unusual as it contains a single positively charged arginine (Arg 209 ) residue. Since the TM domain of the FcR ␥ chain contains a single negatively charged aspartic acid residue, we hypothesized that these oppositely charged residues may promote association. A similar mechanism involving charged TM amino acids has previously been shown to be involved in the assembly of the TCR-CD3 complex (18 -21). Using PCR, we mutated the Arg 209 found in the wild type Fc␣R (R209) to either a positively charged histidine (R209H), a negatively charged aspartic acid (R209D), or an uncharged leucine (R209L). Mutated Fc␣R cDNAs were transfected together with FcR ␥ chain to IIA1.6 cells and assessed for their ability to trigger an increase in [Ca 2ϩ ] i following Fc␣R cross-linking. Our data show a positively charged residue within the TM domain of Fc␣R to be required for functional association with FcR ␥ chain.
Expression Vectors and Transfections-The human Fc␣R cDNA contained within the pCAV vector was a kind gift from Dr. C. Maliszewski (6). The murine FcR ␥ chain cDNA (23) was cloned into the pNUT expression vector (pNUT-␥), allowing selection of transfectants with methotrexate (24). Transfection of IIA1.6 B cells was performed by electroporation as described previously (10). IIA1.6 B cells were stably transfected with Fc␣R alone (Fc␣R ϩ ) or with Fc␣R and ␥ chain (Fc␣R ϩ / ␥ ϩ ) as follows. Fc␣R ϩ cells were generated by co-transfection of Fc␣R cDNA in pCAV, using pRC-CMV.neo as selection marker (10). Two days following transfection, cells were transferred to medium containing G418 and seeded in 24-well plates. After 14 days, each well was tested for Fc␣R expression by FACS analysis using anti-Fc␣R mAb A77 (murine IgG1; Ref. 4) or My43 (murine IgM; Ref. 25). Fc␣R ϩ cells were isolated (two rounds of selection) using mAb A77 and Dynabeads coated with rat anti-mouse IgG1 (Dynal, Oslo, Norway) (10). Fc␣R ϩ /␥ ϩ cells were generated by co-transfection of pCAV, pNUT-␥, and pRC-CMV-.neo. Two days following transfection, cells were transferred to medium containing G418 and 1 M methotrexate. After 2 weeks, the concentration of methotrexate was increased to 10 M. All cells surviving this selection procedure were found to be Fc␣R-positive. Expression of FcR ␥ chain was analyzed by RT-PCR, immunoprecipitation, and Western blotting.
Construction of Mutant Fc␣R cDNAs-Overlap extension PCR (26) was used to construct mutant Fc␣R cDNAs where the codon CGC (encoding Arg 209 in the wild type Fc␣R) was mutated to either CAC (histidine), GAC (aspartic acid), or CTA (leucine). For each mutant, a first round of PCR was performed, as follows, generating two DNA fragments with overlapping ends. 5Ј fragments were generated using a sense oligonucleotide primer FCAL (5Ј-ATGGACCCCAAACAGACC-3Ј) and either antisense primer H2 (5Ј-TGCCACGGCCATGTGGAT-CAAGTT-3Ј), D2 (5Ј-TGCCACGGCCATGTCGATCAAGTT-3Ј), or L2 (5Ј-TGCCACGGCCATTAGGATCAAGAT-3Ј). Similarly, 3Ј DNA fragments were generated using the antisense primer AS1 (5Ј-TCCAGGT-GTTTACTTGCAGACAC-3Ј) and either the sense primer H1 (ACCTT-GATCCACATGGCCGTGGCA-3Ј), D1 (5Ј-ACCTTGATCGACATGGC-CGTGGCA-3Ј), or L1 (5Ј-ACCTTGATCCTAATGGCCGTGGCA-3Ј). Mismatch nucleotides responsible for the introduction of the desired mutation are underlined. First round PCR amplification was carried out by using 50 ng of Fc␣R cDNA as template, in a reaction mixture containing 50 mM KCl, 20 mM Tris⅐Cl, pH 8.3, 2.5 mM each dNTP, 2.5 mM MgCl 2 , 0.1 mg/ml bovine serum albumin, 10 pmol of each primer, and 2 units of Vent DNA polymerase (New England Biolabs) in a final volume of 100 l. Samples were overlaid with light mineral oil, denatured at 94°C for 5 min, before 30 cycles of amplification in a DNA thermal cycler 480 (Perkin Elmer) using a step program (94°C, 1 min; 55°C, 1 min; 72°C, 2 min) followed by a final extension at 72°C for 7 min. A second round of PCR was subsequently used to generate mutated cDNAs encoding the entire Fc␣R coding region. PCR conditions were as above, except that 25 ng of each overlapping DNA fragment was added as template along with 10 pmol of primers FCAL and AS1, and 1 unit of AmpliTaq DNA polymerase (Perkin Elmer) was used to amplify mutated cDNA fragments. PCR products were subcloned into pGEM-T (Promega) vectors, and the integrity of all mutants was confirmed by sequence analysis. Mutant cDNAs were cloned into the eukaryotic expression vector pSG5 (27) prior to transfection as described above.
Immunofluorescence-Cells were incubated with either mAb A77 or My43 culture supernatant for 30 min at 4°C. Cells were washed twice with PBS, 1% bovine serum albumin, 0.1% NaN 3 , and subsequently incubated with either FITC-conjugated goat anti-mouse (GAM) IgG1 or FITC-conjugated GAM IgM, respectively (Southern Biotechnology). Following 30 min incubation at 4°C, cells were washed twice and analyzed on a FACScan flow cytometer (Becton Dickinson). For intracellular staining, cells were permeabilized by incubation in FACS Lysing Solution (Becton Dickinson) for 10 min at room temperature. After washing, FcR ␥ chain expression was detected by incubation with a rabbit antiserum against FcR ␥ chain (generously provided by Dr J.-P. Kinet; Ref. 28) followed by FITC-conjugated goat anti-rabbit F(abЈ) 2 fragments (Southern Biotechnology).
Calcium Mobilization Assays-Intracellular free calcium levels were analyzed using a FACScan (10). Briefly, cells were loaded simultaneously with SNARF-1 (2.8 M) and Fluo-3 (1.4 M) (Molecular Probes) by incubation for 30 min at 37°C. After washing, cells were resuspended in calcium mobilization buffer at a concentration of 1 ϫ 10 7 cells/ml. A flow rate of ϳ140 cells/s was used for calcium mobilization studies with the initial 24 s of each run used to establish a baseline value for the intracellular free calcium concentration, [Ca 2ϩ ] i . The ability of Fc␣R (with or without FcR ␥ chain) to trigger [Ca 2ϩ ] i increases was determined by incubating SNARF-1/Fluo-3-loaded cells with either mAb A77 or My43 for 20 min at room temperature (in the dark). After washing, cells were run on a flow cytometer for 24 s, and GAM IgG1 or GAM IgM Ab was added to cross-link Fc␣R.
Immunoprecipitations-Cells (1 ϫ 10 7 per precipitation) were washed three times in PBS before surface radioiodination by the lactoperoxidase method. Cells were lysed in 1% digitonin, 150 mM NaCl, 10 mM triethanolamine, pH 7.4, containing the protease inhibitors phenylmethylsulfonyl fluoride, N ␣ -p-tosyl-L-lysine chloromethyl ketone, soybean trypsin inhibitor, and leupeptin (15) for 45 min at 4°C. Insoluble material was removed by centrifugation at 13,000 ϫ g for 30 min. Lysates were then precleared four times with Protein G-coated Sepharose CL-4B beads (Pharmacia) and once with beads coated with either an irrelevant mouse IgG1 mAb directed against plant allergens (CLB, Amsterdam) (A77 control) or rabbit serum (FcR ␥ chain antiserum control). Specific precipitations were then performed with beads coated with either mAb A77 or a rabbit antiserum against FcR ␥ chain. The beads were washed four times with digitonin lysis buffer, and precipitates were analyzed by reducing SDS-PAGE on 10% polyacrylamide gels followed by autoradiography.
Tyrosine Phosphorylation Assays-Transfectants were incubated with mAb A77 for 30 min at room temperature, washed twice with RPMI 1640 medium, and resuspended in separate tubes (2 ϫ 10 5 cells/30 l). GAM IgG1 (final concentration 20 g/ml) was added for the indicated time periods at 37°C to cross-link Fc␣R. Reactions were stopped by the addition of 70 l of reducing SDS-PAGE sample buffer (50 mM Tris/HCl (pH 6.8), 10% glycerol, 4.3% SDS, 0.05% bromphenol blue, 4% ␤-mercaptoethanol). Tyrosine phosphorylation signals resulting from cross-linking sIgG2a by addition of GAM F(abЈ) 2 fragments for 2 min at 37°C served as a positive control. Background levels of tyrosine phosphorylation were determined by omitting the cross-linking GAM IgG1. After boiling for 3 min, 15-l aliquots were separated on 7.5% SDS-PAGE gels and electrotransferred to nitrocellulose membranes (0.45 m; Schleicher & Schuell, Dassel, Germany) using a Transblot cell system (Bio-Rad). Membranes were blocked in PT buffer (PBS, 0.1% Tween 20) containing 5% nonfat dry milk (Nutricia, Zoetermeer, The Netherlands) and probed with anti-phosphotyrosine mAb (4G10; UBI, Lake Placid, NY) in PT, 0.05% nonfat dry milk for 1.5 h at room temperature. Following washing three times in PT and once in PT, 0.5% nonfat dry milk for 20 min, membranes were incubated with peroxidase-conjugated GAM IgG (Dako, Glostrup, Denmark) for 1.5 h at room temperature. After washing twice with PT and once with PT, 0.5% NaCl, bound antibodies were detected using the ECL detection system (Amersham, Buckinghamshire, UK).
IL-2 Production-Transfectants were incubated with mAb A77 for 30 min at room temperature. Following washing, cells were seeded into 96-well plates and GAM IgG1 (30 g/ml) added to the culture supernatant for 24 h at 37°C. Alternatively, transfectants were added to wells pre-coated with either human serum IgA (Organon Teknika, Belgium) or human polymeric IgA, isolated from the serum of myeloma patients (a gift from Dr. M. Daha), and incubated for 24 h at 37°C. As control, IL-2 production was triggered via surface IgG2a cross-linking by addition of GAM Ab to culture supernatants for 24 h at 37°C. IL-2 production was assessed by culturing 1 ϫ 10 4 CTLL-2 IL-2-dependent cells with the transfectant culture supernatants in 96-well plates. After a 24-h incubation at 37°C, 1 Ci of [ 3 H]thymidine was added to each well and cultured for 4 h, and cells were harvested onto glass fiber filters (Wallac, Turku, Finland) for liquid scintillation counting.

RESULTS AND DISCUSSION
Expression of Fc␣R and ␥ Chain in Transfected IIA1.6 B Cells-We transfected IIA1.6 B lymphoma cells with cDNAs encoding either Fc␣R alone, or Fc␣R and FcR ␥ chain. The Fc␣R was surface-expressed at a high level in both Fc␣R ϩ and Fc␣R ϩ /␥ ϩ transfectants (Fig. 1A). In Fc␣R ϩ /␥ ϩ transfectants, the presence of a ␥ chain message was detected by specific RT-PCR (Fig. 1B). Expression of ␥ chain at the protein level was confirmed by FACS analysis of permeabilized cells using an anti-␥ chain serum (Fig. 1C). Fc␣R expressed on IIA1.6 cells were reactive with previously described CD89 mAb A77 (Fig.  1A), My43, and A59 and were capable of binding human IgA (data not shown). The increased level of Fc␣R expression seen in Fc␣R ϩ /␥ ϩ cells was presumably not due to co-expression of the FcR ␥ chain since no increase in Fc␣R expression was observed following transfection of FcR ␥ chain cDNA to Fc␣R ϩ IIA1.6 cells. Furthermore, these latter cells were capable of all signaling processes attributable to Fc␣R ϩ /␥ ϩ cells (data not shown).
We found untransfected IIA1.6 cells not to express endogenous FcR ␥ chain (Fig. 1B). This observation suggests that, in contrast to Fc␥RIIIa and Fc⑀RI (12,16), mere cell surface expression of Fc␣R in IIA1.6 cells is not dependent on the presence of FcR ␥ chain. Previously, it was also shown that Fc␣R could be expressed on the surface of COS cells in the absence of FcR ␥ chain (6). Similarly, the high affinity IgG receptor Fc␥RI (CD64), which also associates with FcR ␥ chain (13,14), can be expressed by itself in COS (13) and 3T3 transfectants (30).
Physical Association between Fc␣R and ␥ Chain in Transfected IIA1.6 B Cells and in PMN-We next performed experiments to determine whether there is physical association between Fc␣R and FcR ␥ chain. Cells were surface-labeled with 125 I and lysed in 1% digitonin prior to immunoprecipitation with anti-Fc␣R mAb A77 or anti-FcR ␥ chain antiserum ( Fig.  2A). Immunoprecipitation with CD89 mAb A77 resulted in the isolation of one major band of ϳ60 kDa (arrowheads) from both Fc␣R ϩ (lane 2) and Fc␣R ϩ /␥ ϩ (lane 3) transfectants, but not from untransfected IIA1.6 cells (lane 1). The molecular weight of the observed band is consistent with the predicted size of Fc␣R (3)(4)(5). Immunoprecipitation of radiolabeled cell lysates with an anti-␥ chain-specific antibody precipitated one band of ϳ60 kDa only from the Fc␣R ϩ /␥ ϩ transfectants ( Fig. 2A).
Western blotting analysis of immunoprecipitated proteins further supported the presence of a physical interaction between Fc␣R and FcR ␥ chain in transfectants and suggest that Fc␣R and FcR ␥ chain can also associate in peripheral blood PMN. Proteins were precipitated from digitonin-solubilized cells with beads coated with either A77, anti-FcR ␥ chain serum, or control serum (see "Materials and Methods"), transferred to nitrocellulose, and probed with anti-FcR ␥ chain serum. In IIA1.6 cells, a specific band of ϳ20 kDa, corresponding to the expected size of FcR ␥ chain homodimers (12), was co-precipitated by anti-Fc␣R mAb A77 from Fc␣R ϩ /␥ ϩ cells only (Fig. 2B, lane 3). No bands of this size were precipitated by A77 from either untransfected IIA1.6 or Fc␣R ϩ transfectant cell lysates (lanes 1 and 2). Similarly sized bands were detected in PMN cell lysates following immunoprecipitation with either A77 (lane 4) or anti-␥ chain serum (lane 5). No specific bands of this size were seen following immunoprecipitation with either an irrelevant mouse IgG1 antibody (Fig. 2C, lanes 1-4; A77 control) or with normal rabbit serum (Fig. 2C, lane 5; anti-␥ chain control). The 28-kDa bands seen in Fig. 2B, lanes 1-5, are considered to be nonspecific since these bands were also seen in the control immunoprecipitations (Fig. 2C, lanes 1-5). We, furthermore, observed a slight difference in mobility of the FcR ␥ chain between transfectants (murine ␥ chain) and PMN (human ␥ chain) (Fig. 2B, lane 3 versus lanes 4 and 5). This different migration profile was surprising in view of the high homology between murine and human ␥ chains (12). Therefore, we next precipitated murine ␥ chain from Fc␣R ϩ /␥ ϩ transfectants and murine peritoneal macrophages, and human ␥ chain from PMN and U937 cells. Following SDS-PAGE and transfer to nitrocellulose, the blots were probed with anti-␥ chain serum. Results presented in Fig. 2D, clearly showed a slight, albeit significant, difference in mobility between murine (lanes 1 and 2) and human (lanes 3 and 4) FcR ␥ chains. The observed difference in the mobilities of human and mouse ␥ chains may be explained by the slightly different amino acid sequences of these two chains. However, the phosphorylation state of FcR ␥ chain has also been shown to affect its mobility in SDS-PAGE (31); therefore, differential phosphorylation between human and murine cells may also explain the observed size difference.
Previously it has been demonstrated that FcR ␥ chain homodimers associate with TCR-CD3, Fc⑀RI, and all three classes of Fc␥R (12)(13)(14)(15). During the preparation of this manuscript, it was demonstrated that FcR ␥ chain may also be found in membrane complexes with Fc␣R in the U937 cell line (17). Our results confirm this observation in a transfectant model system and, furthermore, provide evidence to suggest that Fc␣R may associate with FcR ␥ chain in normal peripheral blood PMN.
FcR ␥ Chain Is Critical for Fc␣R Signaling in IIA1.6 B Cells-We next assessed Fc␣R signal transduction by both types of IIA1.6 B cell transfectants. These FcR-negative B lymphoid cells have been used extensively as a model system for studying Fc␥R functioning. IIA1.6 cells express endogenous surface IgG2a, which upon cross-linking triggers a calcium flux, protein tyrosine phosphorylation, and synthesis and release of IL-2 (9 -11). Changes in [Ca 2ϩ ] i and tyrosine kinase activation are important proximal signaling events and have been correlated with the presence of immunoreceptor tyrosinebased activation motifs (ITAMs) (32) within receptor cytoplasmic domains. Fc␣R is devoid of ITAMs (6) while the FcR ␥ chain homodimer contains two such motifs (16).
Our results show FcR ␥ chain to be required for generation of both proximal and distal signaling events via Fc␣R. Crosslinking of Fc␣R in Fc␣R ϩ /␥ ϩ transfectants leads to a rapid rise in [Ca 2ϩ ] i that was maintained for at least 2 min (Fig. 3A). In the absence of FcR ␥ chain, no [Ca 2ϩ ] i increase was observed. Following Fc␣R cross-linking, a rapid tyrosine phosphorylation of cellular proteins was also observed only in Fc␣R ϩ /␥ ϩ cells (Fig. 3B). Tyrosine phosphorylation was detected within 20 s and reached a maximum between 40 s and 2 min in all experiments. In all experiments, no detectable signaling responses were initiated when the cells were incubated with GAM IgG1 mAb alone, indicating that this antibody does not react with the surface IgG2a expressed by the IIA1.6 cells (as reported previously; Ref. 33).
Recently, it has been noted that Fc␣R cross-linking triggers phosphorylation of FcR ␥ chain in U937 cells (17), supporting the hypothesis that FcR ␥ chain is critically involved in the generation of Fc␣R signal transduction responses.
Very little is known concerning signal transduction pathways associated with Fc␣R. Interestingly, it has recently been proposed that Fc␣R surface expression may be regulated by [Ca 2ϩ ] i , since treatment of neutrophils with known calcium agonists leads to up-regulated Fc␣R expression (34). This observation, taken together with our data that Fc␣R/␥ chain cross-linking mediates a rapid rise in [Ca 2ϩ ] i , may explain the phenomena of IgA-induced Fc␣R expression-up-regulation, a function apparently unique to Fc␣R, at least among other FcRs (2).
Positions of Fc␣R are marked by arrowheads. B, 1% digitonin cell lysates (as indicated) were immunoprecipitated with either A77 mAb (lanes 1-4) or anti-␥ chain serum (lane 5). Precipitates were separated by nonreducing SDS-PAGE, transferred to nitrocellulose, and probed with anti-␥ chain serum. C, control precipitations were also performed using beads coated with an irrelevant murine IgG1 Ab (mIgG1), as isotype control for A77 (lanes 1-4), or normal rabbit serum (Rbt serum) as a control for anti-␥ chain serum (lane 5). D, mouse and human FcR ␥ chains were precipitated from lysates of different mouse (lanes 1 and 2)  We, furthermore, examined the capacity of Fc␣R ϩ and Fc␣R ϩ /␥ ϩ transfectants to trigger the release of IL-2 from IIA1.6 B cells. Secretion of IL-2, triggered via Fc␣R crosslinking, also proved to be dependent on FcR ␥ chain co-expression (Fig. 3C). IL-2 was secreted following Fc␣R cross-linking by polymeric IgA, serum IgA, and CD89 mAb A77. The ability of polymeric IgA to trigger a higher level of IL-2 release than serum IgA may indicate Fc␣R to have higher affinity for this molecular species. IL-2 release represents a distal signaling event which can be triggered via cross-linking sIgG2a in IIA1.6 B cells (35). IIA1.6 sIgG2a is part of the B cell antigen receptor signaling complex which includes at least four ITAMs (36). In contrast, previous data from our laboratory utilizing a panel of Fc␥R IIA1.6 transfectants showed that cross-linking of Fc␥RIIa (CD32; with only one ITAM) was sufficient to induce a Ca 2ϩ response and tyrosine kinase activation but not IL-2 release (10,32). Taken together, these data suggest the number of ITAMs within signaling complexes to be of importance for determining the type of signals generated via receptor complexes in IIA1.6 B cells. This hypothesis is supported by recent work in Jurkat cells, where the number of ITAMs within the TCR chain was found to quantitatively affect T cell responses (37).
Several reports in the literature suggest Fc␣R to be capable of synergizing with Fc␥R in promoting ADCC, phagocytosis, and the respiratory burst (38 -40). Our results provide a model for such co-operation in which FcR ␥ chain mediates signal transduction via both types of receptors. The importance of FcR ␥ chain in triggering phagocytosis via Fc␥R has recently been demonstrated in FcR ␥ chain knock-out mice (41) and transfection studies (42,43). Since our data suggest that Fc␣R functioning also depends on FcR ␥ chain association, we hypothesize that FcR ␥ chain-deficient mice may display aberrations in IgA-mediated mucosal immune responses.
Molecular Basis for Fc␣R/FcR ␥ Chain Association-Earlier reports proposed the interaction of FcR ␥ chain with Fc⑀RI and Fc␥RIIIA to involve a conserved region (LFAVDTGL) within the TM domains of these receptors, containing a negatively charged aspartic acid residue (underlined) (12). Human Fc␥RI (which also associates with FcR ␥ chain) displays high homology with Fc⑀RI and Fc␥RIIIA within this region of the TM domain (MFLVNTVL), but lacks such an aspartic acid residue (13,14,44). The predicted TM domain of Fc␣R is 19 amino acids long and, unusually, contains a positively charged arginine residue at position 209 (Arg 209 ) (6; Fig. 4). Moreover, Fc␣R TM domain displays no obvious homology at the protein level to the TM domains of Fc⑀RI, Fc␥RIIIA, or Fc␥RI (13,14).
Although, in general, charged residues are uncommon within the TM domains of integral membrane proteins, they are not unknown. For example, the TCR-␣ and TCR-␤ chains have conserved positively charged residues within their TM domains while the invariant chains of the CD3 complex contain negatively charged TM residues. Elegant site-directed mutagenesis experiments have shown these charged residues to be of critical importance for surface expression of the TCR-CD3 complex (18 -21).
Therefore, based upon the predicted TM regions of Fc␣R and FcR ␥ chain (Fig. 4), we hypothesized the positively charged Arg 209 to be important for association with FcR ␥ chain. To test FIG. 4. Protein sequence within the region of the transmembrane domains of Fc␣R, Fc␥RI, and FcR ␥ chain. Predicted transmembrane spanning residues are underlined. Positively (ᮍ) and negatively (ᮎ) charged residues which may mediate association between Fc␣R, Fc␥RI, and FcR ␥ chain are indicated.

FIG. 3. Signalling events triggered by cross-linking Fc␣R in transfected IIA1.6 B cells.
A, calcium mobilization triggered by Fc␣R in SNARF-1/Fluo-3-loaded transfectants. Fc␣R ϩ (dotted line) or Fc␣R ϩ /␥ ϩ (solid line) transfectants were incubated with CD89 mAb A77 for 20 min at room temperature, and Fc␣R was subsequently crosslinked with GAM IgG1 F(abЈ) 2 (arrow). [Ca 2ϩ ] i levels were analyzed by flow cytometry as described under "Materials and Methods." Data are representative of five individual experiments. B, tyrosine phosphorylation of cellular proteins upon cross-linking Fc␣R in Fc␣R ϩ (left panel) or Fc␣R ϩ /␥ ϩ (right panel) IIA1.6 B cells. Transfectants were incubated with mAb A77 for 30 min at room temperature and washed twice with RPMI 1640 medium. Cells were then incubated with GAM IgG1 Ab for the indicated time periods. As control, cells were also incubated either with A77 alone (A77) as negative control or with GAM IgG (GAM) to cross-link the surface IgG (positive control). Samples were separated by SDS-PAGE, transferred to nitrocellulose, and probed with an antiphosphotyrosine mAb. C, induction of IL-2 production following crosslinking of Fc␣R in transfected IIA1.6 B cells. Fc␣R ϩ (open boxes) and Fc␣R ϩ /␥ ϩ (filled boxes) transfectants were incubated for 24 h in wells coated with either serum IgA (IgA) or polymeric IgA (pIgA). Alternatively, transfectants were incubated with CD89 mAb A77 for 30 min at room temperature, washed, and seeded into wells. A cross-linking GAM IgG1 Ab was then added to the culture supernatant for 24 h (A77ϩGAM␥1). Control cells were incubated with culture medium alone (medium) or had GAM␥ Ab added to cross-link sIgG2a (GAM). Results shown are representative of data obtained for IL-2 release in three separate experiments. this hypothesis, we constructed mutant Fc␣R molecules using overlap extension PCR (see "Materials and Methods"), in which the wild type Arg 209 residue was replaced by either a positively charged histidine (Fc␣R-R209H), a negatively charged aspartic acid (Fc␣R-R209D), or an uncharged leucine (Fc␣R-R209L). Mutant Fc␣R cDNAs were transfected together with FcR ␥ chain to IIA1.6 B cells. Fc␣R and FcR ␥ chain mRNA transcripts were readily detectable by RT-PCR, and the resulting mutated Fc␣R proteins were well expressed at the cell surface as shown by FACS analysis (Fig. 5, A and B). Surprisingly, however, FcR ␥ chain protein was only observed in transfectants co-expressing an Fc␣R molecule possessing a positively charged residue within the TM domain, i.e. wild type Fc␣R and Fc␣R-R209H (Fig. 5C). We next assayed mutant Fc␣Rs for their ability to form functional Fc␣R/FcR ␥ chain signaling complexes, by measuring the increase in [Ca 2ϩ ] i triggered upon Fc␣R cross-linking (see "Materials and Methods"). Predictably, only the Fc␣R-R209H mutant resulted in intact functional integrity of the Fc␣R/FcR ␥ chain complex comparable to the wild type Fc␣R (Fig. 5D). These data demonstrate that a positively charged residue within the Fc␣R TM domain promotes functional association with FcR ␥ chain. Our studies suggest furthermore that, in IIA1.6 cells, FcR ␥ chain molecules unable to associate with Fc␣R are degraded. This occurs possibly via a mechanism recognizing the charged aspartic acid residue within the TM domain (Fig. 4). The presence of charged residues within the TM domain of some proteins can result in their retention and degradation in the endoplasmic reticulum (44).
A similar charge-based mechanism may also be operational for Fc␥RI-FcR ␥ chain association, since a positively charged histidine residue is located directly preceding the predicted TM domain (45). Charged residues located at or near the extracellular/TM boundary may also be able promote association between protein subunits as suggested for the ␣ and ␤ subunits of the major histocompatibility complex class II molecule (46). The fact that surface expression of both Fc␣R and Fc␥RI appears independent of FcR ␥ chain (in contrast to Fc⑀RI and Fc␥RIIIA) may, indeed, argue for a different type of FcR ␥ chain association between these two receptors and either Fc⑀RI or Fc␥RIIIa.
In conclusion, our data demonstrate that Fc␣R is capable of associating with the FcR ␥ chain in a transfectant model system and also provides evidence that these two proteins can associate in peripheral blood PMN. We, further, show Fc␣R signal transduction responses to be critically dependent upon co-expression of FcR ␥ chain, and that a positively charged residue within the TM domain of Fc␣R is involved in the functional association between these two molecules. Implications of these observations in terms of cooperation (and competition) between immunoglobulin receptors in cellular activation processes remain to be elucidated.