The Cytoplasmic Domain of Human FcγRIa Alters the Functional Properties of the FcγRI·γ-Chain Receptor Complex*

The γ/ζ-chain family of proteins mediate cell activation for multiple immunoglobulin receptors. However, the recognition that these receptors may have distinct biologic functions suggests that additional signaling elements may contribute to functional diversity. We hypothesized that the cytoplasmic domain (CY) of the ligand binding α-chain alters the biological properties of the receptor complex. Using macrophage FcγRIa as a model system, we created stable transfectants expressing a full-length or a CY deletion mutant of human FcγRIa. Both receptors functionally associate with the endogenous murine γ-chain. However, we have established that the CY of FcγRIa directly contributes to the functional properties of the receptor complex. Deletion of the FcγRIa CY leads to slower kinetics of receptor-specific phagocytosis and endocytosis as well as lower total phagocytosis despite identical levels of receptor expression. Deletion of the CY also converts the phenotype of calcium independent FcγRIa-specific phagocytosis to a calcium-dependent phenotype. Finally, deletion of the CY abrogates FcγRIa-specific secretion of interleukin-6 but does not affect production of interleukin-1β. These results demonstrate a functional role for the CY of FcγRIa and provide a general model for understanding how multiple receptors that utilize the γ-chain can generate diversity in function.

Fc␥ receptors play a central role in the handling of immune complexes, regulation of inflammatory responses, antibody secretion, and T cell activity (1)(2)(3)(4). Common to each of these functions is the initiation of tyrosine phosphorylation following receptor cross-linking (5) and the involvement of the ␥/ subunits leading to the view that Fc receptors subserve redundant signaling functions. However, recent evidence suggests that these receptors are not redundant. For example, Fc␥RIIIa appears necessary for initiating the Arthus inflammatory reaction (6,7), while Fc␥RIa and Fc␣RI can down-regulate inflam-matory responses by initiating the secretion of IL 1 -10 and IL-1ra, respectively (4,8). The basis for these differences are unknown.
Fc␥RI is expressed on the cell surface in association with the ␥-chain (9,10). This association is not a prerequisite for transient receptor expression but is necessary for stable expression (11,12). The ␥-chain cytoplasmic domain contains an immunoreceptor tyrosine activation motif (ITAM) and current data suggest that the ␥-chain cytoplasmic domain is both necessary and sufficient for Fc␥RIa induced functions (13)(14)(15). Biochemical studies have shown that cross-linking of the Fc␥RIa⅐␥chain complex results in activation of a Src family kinase(s) and the tyrosine kinase p72 Syk (2,5). Activation of these kinases results in tyrosine phosphorylation of the ␥-chain and the initiation of a signaling cascade that can culminate in the induction of degranulation, phagocytosis, an oxidative burst, ADCC activity and the induction of gene transcription. The association between Fc␥RIa and ␥-chain may also be important in the formation of a higher affinity receptor complex through the recruitment of two ligand binding chains to the ␥ homodimer (16).
Unlike the ␥-chain, the cytoplasmic domain of Fc␥RI does not contain an ITAM or other tyrosine containing signaling motifs. Nonetheless, murine Fc␥RI on J774 cells is constitutively phosphorylated on serine and after phorbol 12-myristate 13-acetate stimulation the level of phosphorylation increases (17). The cytoplasmic domain of Fc␥RI may also associate with actinbinding protein-280 (ABP-280, also known as non-muscle filamin) in the absence of ligand (18). Receptor engagement by ligand apparently abrogates this association, although its functional significance is not clear. Both of these observations suggest that the cytoplasmic domain of Fc␥RIa may be actively involved in the biologic phenotype of Fc␥RI. Furthermore, Fc␥RIa in the absence of the ␥-chain can signal for calcium in COS-1 cells and the transmission of this calcium signal requires the Fc␥RIa cytoplasmic domain (19). Based on these observations, and the observations that several ␥-chain associated Fc receptors initiate functionally distinct cell programs, we hypothesized that the Fc␥RIa cytoplasmic domain may serve to modify the signaling of the Fc␥RI⅐␥-chain receptor complex. By directly comparing wild type human Fc␥RIa with a cytoplasmic domain deletion mutant of Fc␥RIa expressed at comparable levels in stable transfectants of the murine macrophage cell line P388D1, we have established that the cytoplas-mic domain of Fc␥RIa alters the functional properties of the receptor complex. These observations provide a general framework for understanding the unique properties of the family of Fc␥ receptors which associate with the ␥-chain.

EXPERIMENTAL PROCEDURES
Cell Culture and Reagents-The murine macrophage cell line P388D1 stably transfected with a cDNA encoding human Fc␥RIa or a mutant form of Fc␥RIa containing a stop codon after the first amino acid of the cytoplasmic domain (Lys 315 3 Stop 315) were prepared as described previously (13). P388D1 cells transfected with human Fc␥RIIa were previously described (20). All cell lines were maintained as adherent cultures (Corning Tissue Culture Dishes) in RPMI 1640 as described previously (20). All tissue culture reagents were from Life Technologies, Inc. (Grand Island, NY).
Human and mouse IgG were obtained from Sigma. Mouse F(abЈ) 2 fragments and F(abЈ) 2 goat anti-mouse IgG (G␣M) were obtained from Jackson ImmunoResearch (West Grove, PA). F(abЈ) 2 fragments of the anti-Fc␥RIa mAbs 22.2 and 32.2 were obtained from Medarex (Annandale, NJ). IgM anti-H-2D d (clone 3-25.4) was obtained from Pharmingen (San Diego, CA). The hybridoma line expressing the rat anti-murine Fc␥RII/Fc␥RIII mAb 2.4G2 was obtained from ATCC (Manassas, VA). All other reagents were from Sigma. Quantitative huFc␥RI expression was matched for cells expressing the wild type (WT) and the cytoplasmic domain deletion mutant (MUT) by fluorescence activated cell sorting using anti-Fc␥RI mAb 22.2-FITC (Medarex).
A polyclonal anti-␥-chain Ab (666) was kindly provided by Dr. Jean-Pierre Kinet (21). In addition, polyclonal anti-␥-chain Abs were prepared in rabbits immunized with a C-terminal peptide sequence that is shared by both human and murine ␥-chain exactly as described (21). To verify the specificity of the polyclonal antibodies, ␥-chain from U937 cells was immunoprecipitated with protein G-agarose bound anti-␥chain mAb (mAb 4D8) (kindly provided by Dr. J. Kochan) (22) followed by immunoblotting with the polyclonal Abs (see below).
For immunoblotting analysis, immunoprecipitates were separated by SDS-polyacrylamide gel electrophoresis and blotted onto nitrocellulose membranes (24). Membranes were blocked with 10% non-fat milk followed by incubation with either polyclonal anti-␥-chain Ab or antiphosphotyrosine mAb 4G10 (UBI). Blots were washed 3 times with PBS, 0.1% Tween 20 and bound mAb or Ab was detected with horseradish peroxidase-conjugated anti-mouse IgG or anti-rabbit IgG (Amersham Pharmacia Biotech or Jackson ImmunoResearch). Following 3 more washes, bound Ab was detected using ECL (Amersham Pharmacia Biotech) according to the manufacturer's directions. Membranes were stripped by incubation with Tris-HCl, pH 2.3, for 30 min at room temperature and then re-probed as described above.
Analysis of [Ca 2ϩ ] i -Fura-2 (Molecular Probes, Eugene, OR), a fluorescent dye with spectral properties that change with the binding of free Ca 2ϩ , was used to measure changes in intracellular calcium concentrations as we have described (25). P388D1 cells, adhered to 25-mm diameter round glass coverslips at 5 ϫ 10 5 cells/ml, were incubated at 37°C for 15 min with 2 M fura-2 AM. During the last 5 min, anti-Fc␥RIa mAb 22.2 F(abЈ) 2 was added. After incubation, the cells were washed once with modified PBS (PBS prepared with 5 mM KCl and 5 mM glucose) and then re-warmed to 37°C for 5 min in modified PBS plus 1.1 mM Ca 2ϩ and 1.6 mM Mg 2ϩ prior to analysis. The coverslips were transferred to the stage of a Nikon Diaphot and the ratio of fluorescence emission of fura-2 was monitored. After establishment of a baseline, F(abЈ) 2 goat anti-mouse IgG was added at a final concentration of 35 g/ml. Analysis was continued for an additional 5 min. Quantitation of intracellular [Ca 2ϩ ] before and after treatment of cells with BAPTA-AM was performed using Indo-1 (Molecular Probes) in an SLM Spectrofluorometer (Spectronics Instruments, Rochester, NY) exactly as we have previously described (20,25).
Endocytosis and Phagocytosis-Endocytosis of transfected huFc␥RIa was determined by monitoring the disappearance of cell surface-associated anti-Fc␥RI mAb 32.2 F(abЈ) 2 (Medarex) upon cross-linking with F(abЈ) 2 G␣M (26). Similarly, endocytosis of murine Fc␥RIa on nontransfected cells was determined using mIgG2a (Sigma) and F(abЈ) 2 G␣M. Cells (50 l, 5 ϫ 10 6 /ml) were incubated with a saturating concentration of mAb for 15 min at 4°C. Following two washes in PBS, 1% bovine serum albumin, F(abЈ) 2 G␣M was added, and cells were incubated for an additional 15 min at 4°C. Cells were then placed at 37°C for varying periods of time, rapidly pelleted, and washed with PBS, 1% bovine serum albumin containing azide at 4°C. Remaining cell surface-associated receptor was quantitated with FITC-conjugated F(abЈ) 2 donkey anti-goat IgG by flow cytometry. Phagocytosis by transfected P388D1 cells was determined in an adherent assay system (20). Biotinylated mAb 22.2 F(abЈ) 2 and biotinylated bovine erythrocytes were prepared as we have previously described (20). Biotinylated erythrocytes were saturated with streptavidin and washed. The resulting erythrocytes were coated with biotinylated mAb and the level of mAb binding was verified by flow cytometry.
P388D1 cells, adhered to round glass coverslips at 5 ϫ 10 5 cells/ml, were incubated with anti-Fc␥RIa mAb 22.2 F(abЈ) 2 -coated erythrocytes (E-22.2) in RPMI, 20% fetal calf serum (50 l at 5 ϫ 10 7 E/ml) for 1 h at 37°C. Alternatively, erythrocytes coated with an IgM anti-H2D d (Pharmingen) were used. Non-internalized erythrocytes were lysed by brief immersion of the coverslip in dH 2 O followed by immersion in buffer. Phagocytosis was quantitated by light microscopy and expressed as a phagocytic index (number of erythrocytes internalized per 100 P388D1 cells).
Treatment of cells with BAPTA-AM (Molecular Probes) to quench intracellular Ca 2ϩ levels was performed as described previously (20). Briefly, coverslip adherent cells were incubated with varying concentrations of BAPTA-AM in RPMI, 20% fetal calf serum for 30 min at 37°C followed by two washes. E-22.2 in RPMI, 20% fetal calf serum were then added and handled as described above. Controls included loading cells with the BAPTA-AM solvent (1% dimethyl sulfoxide) for the same period of time.
The kinetics of transfected Fc␥RIa-specific phagocytosis was performed using a flow cytometric based assay (27). In this assay, the E-22.2 were labeled with the PKH26 Red Fluorescence Cell linker Kit (Sigma). Transfected P388D1 cells were mixed in suspension with labeled E-22.2 at a ratio of 50:1 (E:P388D1)(both in RPMI, 20% fetal calf serum), pelleted, and incubated at 37°C for varying periods of time. At each time point, the supernatant was removed and non-internalized erythrocytes were rapidly lysed in hypotonic saline for 30 s followed by 3 washes in PBS, 1% bovine serum albumin at room temperature. Samples were analyzed immediately by flow cytometry. Results are expressed as a phagocytic capacity (mean fluorescence intensity of phagocytic cells with one or more internalized erythrocytes ϫ % of cells with one or more internalized erythrocytes) as we have described (27).
Flow Cytometry-Aliquots of cells at 5 ϫ 10 6 cell/ml were incubated with saturating concentrations of primary mAb for 30 min at 4°C followed by two washes. For indirect immunofluorescence, the cells were then incubated with saturating concentrations of FITC-conjugated goat anti-mouse IgG F(abЈ) 2 at 4°C for another 30 min. After washing, the cells were analyzed immediately for immunofluorescence using a FACScan (Becton Dickinson Immunocytometry Systems, San Jose, CA).
Statistical Analysis-Analysis of flow cytometry listmode data was done using CellQuest (Becton Dickinson Immunocytometry). Statistical comparisons were performed with the paired t test. A probability of 0.05 was used to reject the null hypothesis that there is no difference between the samples. Human Fc␥RIa expressed on monocytes and the myelomonocytic cell line U937 noncovalently associates with the ␥-chain of the Fc⑀RI receptor complex (9,10). The transmembrane regions of Fc␥RIa and the ␥-chain mediate receptor complex assembly and 20 of the 21 amino acids in the transmembrane region are identical in murine and human ␥-chain with one conservative difference (I 3 V). As predicted, comparable amounts of ␥-chain was co-immunoprecipitated from the WT and MUT lines after cell lysis in buffer containing 1% digitonin (Fig. 2). As a positive control, we co-immunoprecipitated ␥-chain with endogenously expressed murine Fc␥RII and Fc␥RIIIa using mAb 2.4G2. In accordance with the ␥-chain association data, cross-linking of the transfected WT and MUT huFc␥RIa with mAb 22.2 F(abЈ) 2 ϩ F(abЈ) 2 G␣M resulted in tyrosine phosphorylation of the ␥-chain (results not shown) demonstrating that both WT and MUT huFc␥RIa associate with the endogenous murine ␥-chain and initiate tyrosine phosphorylation of the ␥-chain.

Assembly of Fc␥RIa Receptor
The huFc␥RIa CY Domain Alters the Magnitude and Kinetics of Fc␥RIa Internalization-While devoid of tyrosine residues, the ␣-chain of murine Fc␥RIa has been shown to be phosphorylated on serine and/or threonine residues (17) and human Fc␥RIa has been shown to bind to ABP under some conditions (18). Furthermore, Fc␥RIa in the absence of the ␥-chain can signal for calcium in COS-1 cells and the transmission of this calcium signal requires the Fc␥RIa cytoplasmic domain (19). Accordingly, we considered the possibility that the CY domain of huFc␥RIa may contribute to the functional properties of the receptor complex. Using erythrocytes coated with the antihuman Fc␥RIa mAb 22.2 F(abЈ) 2 , both WT and MUT huFc␥RIa mediated receptor-specific phagocytosis (Fig. 3). However, the WT construct consistently displayed a higher phagocytic index despite identical levels of receptor expression (Fig. 3). There was no internalization of E-22.2 by parental non-transfected P388D1 cells and no phagocytosis of erythrocytes coated with an IgM anti-H-2Dd mAb (clone 3-25.4) by any cell type (Fig.  3A), despite comparable binding of the E-3-25.4 probe to the transfected cells when compared with E-22.2.
We considered the possibility that engagement of the ligandbinding site of Fc␥RIa, which can augment receptor function (28), might alter the difference in phagocytic capacity between the WT and MUT Fc␥RI. However, saturation of Fc␥RIa with mIgG2a did not alter the magnitude of Fc␥RIa-specific phagocytosis and did not abrogate the quantitative difference in phagocytosis between WT and MUT Fc␥RIa (Table I).
Phagocytosis by WT huFc␥RIa also displayed more rapid kinetics (Fig. 3B). Similarly, while both WT and MUT huFc␥RIa were capable of endocytosis, endocytosis by the WT receptor was more rapid than that mediated by the MUT receptor (Fig. 4). Endocytosis of endogenous murine Fc␥RIa, assessed on non-transfected P388D1 cells, was indistinguishable from the transfected huFc␥RIa. Since we have matched the WT and MUT cell lines for receptor expression, the differences in phagocytic capacity and the more rapid kinetics of phagocytosis and endocytosis by WT Fc␥RI provide the first evidence that the cytoplasmic domain of Fc␥RIa, in association with the ␥-chain, can affect receptor function.
The CY Domain of the ␣-Chain Determines Ca 2ϩ Sensitivity of Fc␥RIa Phagocytosis-Through the use of chimeric and mutant receptors, the ITAM has been shown to be both necessary and sufficient for Fc␥R phagocytosis and the Fc␥R Ca 2ϩ transient (14, 15, 20, 29 -31). The functional importance of the Ca 2ϩ transient has been demonstrated with huFc␥RIIa which incorporates an ITAM directly in the CY domain and requires elevations in intracellular Ca 2ϩ to mediate phagocytosis (20). In contrast, Fc␥RIa/␥-chain specific phagocytosis is independent of the receptor-induced Ca 2ϩ transient (20), and we considered the possibility that the CY domain of Fc␥RIa confers a Ca 2ϩindependent phenotype on Fc␥RIa-specific phagocytosis. When intracellular Ca 2ϩ levels were quenched with BAPTA (resulting in [Ca 2ϩ ] ϭ 57 Ϯ 9.3 nM with 20 M BAPTA treatment), receptor-specific phagocytosis induced by the WT huFc␥RIa was unaltered (Fig. 5A), as we have previously shown for Fc␥RI on human monocytes. In contrast, receptor-specific phagocytosis induced by the MUT huFc␥RIa was blocked by pretreatment of the cells with BAPTA in a dose-dependent manner (Fig. 5A).
Since the absolute level of MUT huFc␥RIa phagocytosis is lower than WT huFc␥RIa, we considered the possibility that the BAPTA sensitivity might be related to the quantitative level of phagocytosis. Accordingly, phagocytosis by the WT huFc␥RIa was performed with E-22. huFc␥RIa and E-22.2 prepared at the maximal conjugation level. WT huFc␥RI insensitivity to BAPTA was maintained under these reduced phagocytic conditions (Fig. 5A) indicating that the Ca 2ϩ insensitivity is a property of the CY domain of huFc␥RIa. As an additional control, Fc␥RIIa-specific phagocytosis by P388D1 cells expressing full-length huFc␥RIIa (with a phagocytic index of 167 Ϯ 32.6) was also shown to be blocked by pretreatment of the cells with BAPTA (Fig. 5B), as we have previously reported (20). Importantly, both WT and MUT huFc␥RIa receptor complexes induced indistinguishable Ca 2ϩ transients when cross-linked with anti-receptor mAb (results not shown). Thus, WT huFc␥RIa engages a Ca 2ϩ -insensitive phagocytic pathway while MUT huFc␥RIa with the associated ␥-chain engages a Ca 2ϩ -sensitive phagocytic pathway. These results provide additional evidence that the CY domain of the ligand-binding ␣-chain of huFc␥RIa alters functional properties of ␥-chain ITAM-dependent functions.
Requirement of the Fc␥RIa ␣-Chain for the Induction of IL-6 Secretion-In addition to its role in internalization, Fc␥RIa can also modulate the immune response through the induction of cytokine secretion. In particular, activation of monocytes/ macrophages by Fc␥RIa can result in the secretion of IL-6 and IL-1␤ (32,33). Accordingly, P388D1 expressing the WT and MUT forms of huFc␥RIa were stimulated with receptor-specific mAb bound to surface absorbed F(abЈ) 2 -G␣M. Quantitation of IL-1␤ secretion after cross-linking of huFc␥RIa demonstrated that both WT and MUT forms of the receptor were capable of eliciting comparable levels of secretion of this cytokine (Fig. 6). Cells incubated in the presence of G␣M alone were not stimulated to secrete IL-1␤ above the baseline control. In contrast, cross-linking WT huFc␥RIa, but not MUT Fc␥RIa, induced the secretion of IL-6 (8-h time point). We did detect IL-6 production above baseline at 24 h after MUT Fc␥RIa stimulation; however, neutralization of endogenously produced IL-1␤ prevented this induction of IL-6 secretion by the MUT Fc␥RIa after 24 h culture (253 Ϯ 52 pg/ml and 75 Ϯ 21 pg/ml in the absence and presence of a neutralizing anti-IL-1␤ mAb). In contrast, neutralizing anti-IL-1␤ mAb did not abrogate the IL-6 induction observed at the 4-or 8-h time points by WT huFc␥RIa. No significant difference in the ability of phorbol 12-myristate 13-acetate (100 ng/ml) or surface bound IgG, engaging endogenous murine Fc␥RIIa/Fc␥RIIIa and transfected human Fc␥RIa, to elicit IL-1␤ or IL-6 secretion was observed between the WT and MUT lines (results not shown). These results document the requirement for the ␣-chain of the Fc␥RIa receptor complex for the induction of the IL-6 response by the receptor complex and demonstrate that the pathways leading to IL-6 secretion and IL-1␤ secretion are distinct.

DISCUSSION
Three lines of evidence indicate that the Fc␥RIa⅐␥-chain complex induced responses are altered by the cytoplasmic domain of the ␣-chain. First, when matched for receptor expression, WT Fc␥RI has a higher phagocytic capacity than the cytoplasmic domain deletion mutant. The WT receptor complex also displays faster kinetics of phagocytosis and endocytosis despite comparable kinetics of tyrosine phosphorylation of the ␥-chain. Second, the cytoplasmic domain of Fc␥RI confers a calciumindependent phenotype on phagocytosis by the receptor complex. Third, and most importantly, we have observed that the

TABLE I Fc␥RIa-specific phagocytosis
Coverslip adherent cells were prepared as described under "Experimental Procedures." Murine IgG2a (20 g/ml) was added to appropriate coverslips, incubated for 10 min at room temperature followed by the addition of E-22.2 F(abЈ) 2 to all of the coverslips. After 60 min at 37°C, non-internalized E were lysed and the phagocytic index (the number of internalized erythrocytes per 100 P388D1 cells) was determined by light microscopy. induction of IL-6 secretion requires the cytoplasmic domain of the Fc␥RIa ␣-chain (Fig. 6). Taken together, these data provide the first direct demonstration of a functional role of the cytoplasmic domain of Fc␥RI in the Fc␥RIa⅐␥-chain complex.
Previous studies of Fc␥RI transiently transfected into COS cells have suggested that cross-linking of the ␣-chain may mediate changes in intracellular calcium and endocytosis (19,34,35). However, in the absence of ␥-chain, Fc␥RIa is not competent for phagocytosis (13)(14)(15). The central role of the ␥-chain and the tyrosine kinase p72 Syk , which docks via SH2 domains to the tyrosine-phosphorylated ␥-chain, has been demonstrated through a series of experiments using co-transfection of the Fc␥RIa⅐␥-chain complex with Syk kinase and transfection of chimeric receptor constructs incorporating either the ␥-chain cytoplasmic domain or Syk kinase directly (36,37). Furthermore, the observation that both 1) human monocytes/ macrophages in which Syk has been specifically ablated by antisense constructs (38) and 2) murine macrophages from Syk knockout mice (39,40) are incapable of Fc␥ receptor-mediated phagocytosis supports the conclusion that the ␥-chain induced activation of Syk is necessary and sufficient for a phagocytic response. Given these results, our observations that the cytoplasmic domain of Fc␥RIa modulates the kinetics of phagocytosis may reflect the association of cytoskeletal elements with the cytoplasmic domain and subsequent changes in the mobility of the receptor in the plasma membrane. Such a mechanism may explain the observation that Fc␥RI in the NOD mouse also shows altered kinetics of endocytosis (41).
The change in calcium independent signaling elements and in the induction of IL-6 synthesis indicates that the cytoplasmic domain of Fc␥RI also affects the nature of intracellular signals generated by the Fc␥RI⅐␥-chain receptor complex. Studies of the high affinity receptor complex for IgE comprised of ligand binding ␣-chain, a single ␤-chain, and a ␥-chain homodimer indicate that the ␤-chain is constitutively associated with the Src kinase Lyn and can recruit protein kinase C-␦ both of which can modulate the signaling capacity of the ITAM in the ␥-chain (42)(43)(44)(45). In this manner, the ␤-chain acts as an amplifying mechanism and, although the role of serine/threonine phosphorylation of ␥-chain is not yet established, the presence of serine/threonine phosphorylation targets within the ␥-chain provide an attractive target for regulation. Indeed, Fc␥RIa cross-linking on U937 cells results in the serine phosphorylation of the ␥-chain (46) and recruitment of protein kinase C-␤ and protein kinase C-⑀ to the membrane by Fc␥ receptors during phagocytosis has been documented (47). In a parallel fashion, perhaps, the cytoplasmic domain of Fc␥RIa may recruit signaling elements to the receptor complex. Although it contains no ITAM the Fc␥RI cytoplasmic domain is actively serine/threonine phosphorylated and dephosphorylated (17), suggesting active participation in receptor function. Thus, through regulation of phosphorylation and/or recruitment of other signaling molecules to the receptor complex, it is also possible that the Fc␥RIa cytoplasmic domain may directly transmit the signal for IL-6 release in the absence of participation of the ␥-chain.
It is also interesting to note that subtle differences in the ITAM sequences used by Fc receptors may also contribute to distinct biological properties. The ITAM-like sequence in Fc␥RIIa differs from the ITAM in the ␥-chain and this difference, or other adjacent sequences, influence the relative dependence on intracellular calcium transients for phagocytosis (Fig. 5). This property and the differences in serine/threonine residues within the ITAM may allow for differences in some of the functions between Fc␥RIIa and Fc␥RIa/Fc␥RIIIa.
The mechanism(s) by which the cytoplasmic domain of Fc␥RIa alters receptor function is not clear. Murine Fc␥RI is phosphorylated on serine/threonine residues after phorbol 12myristate 13-acetate stimulation (17). However, the functional importance of this phosphorylation has not been examined. Differences in the association of WT and MUT Fc␥RIa with ␥-chain cannot be the mechanism for the observed functional differences between these receptors. Transfectants have been sorted for identical cell surface receptor expression and both WT and MUT associate equally with ␥-chain (by direct coimmunoprecipitation and functionally by receptor induced tyrosine phosphorylation). Recent studies have also demonstrated that association with ␥-chain is essential for stable expression of huFc␥RIa (11,12). A direct interaction between FIG. 5. Differential sensitivity to pretreatment with BAPTA. A, cells were pre-loaded with the intracellular Ca 2ϩ chelator BAPTA followed by the addition of E-22.2 F(abЈ) 2 to assess human Fc␥RIa-specific phagocytosis in WT and MUT P388D1 stable transfectants (n ϭ 8). WT (open bars) and MUT (hatched bars) P388D1 stable transfectants were incubated with E-22.2 were prepared at maximal mAb conjugation ratios (prepared as described in the legend to Fig. 3A). Alternatively, WT P388D1 stable transfectants (solid bars) were incubated with E-22.2 prepared at lower conjugation ratio to match the quantitative level of phagocytosis of the MUT huFc␥RIa (see text). B, as a control, Fc␥RIIa-specific phagocytosis in P388D1 cells transfected with human Fc␥RIIa (20) was determined using E-IV.3 Fab (n ϭ 6). Data are expressed as the mean Ϯ S.D. *, p Ͻ 0.01 relative to control (no BAPTA). N.D., not done.
FIG. 6. IL-6 release, but not IL-1␤ release, requires the cytoplasmic domain of Fc␥RIa. P388D1 stable transfectants were cultured for 8 h in tissue culture wells that had been pretreated with F(abЈ) 2 G␣M (XL) or mAb 22.2 F(abЈ) 2 ϩ XL as described under "Experimental Procedures." Data are expressed as the mean pg/ml cytokine produced Ϯ S.D. (n ϭ 6). *, p Ͻ 0.01 relative to the XL alone control.
human Fc␥RIa and ABP-280 (non-muscle filamin) has been reported (18). Although there are no know consequences of the interaction between Fc␥RIa and ABP, ligand engagement of Fc␥RIa results in both the dissociation of the receptor from ABP-280 and in the enhancement of the Fc␥RIa triggered oxidative burst in U937 cells (28). In the present study, we have used an anti-human Fc␥RIa mAb F(abЈ) 2 fragment that does not engage the ligand-binding site (23). In this transfection system, saturation of Fc␥RI with mIgG2a did not alter the magnitude of human Fc␥RIa-specific phagocytosis nor did it abrogate the quantitative difference in phagocytosis between WT and MUT Fc␥RIa ( Table I).
Demonstration that the cytoplasmic domain of the ␣-chain of Fc␥RIa alters the functional properties of the Fc␥RI⅐␥-chain complex provides a mechanism for unique biologic properties initiated by each receptor complex. Given the diversity in primary sequence of the cytoplasmic domains of the ␥/-chainassociated Fc receptors, the opportunity for these unique domains to confer distinct functions on each receptor complex is clear. These observations also suggest that single nucleotide polymorphism leading to missence mutations in the cytoplasmic domains of these receptors may have biological significance. We have recently described two single nucleotide polymorphisms in the cytoplasmic domain of Fc␥RIa proximate to putative phosphorylation sites (48). These polymorphisms may provide another level of functional variation which builds upon a general framework of the role of sequence variations in the cytoplasmic domains altering the functional properties of the receptor complex.