FcR γ-Chain Dependent Signaling in Immature Neutrophils Is Mediated by FcαRI, but Not by FcγRI¹

Immune cell signaling can be initiated by binding of Abs to FcRs. After activation, FcRs trigger effector functions such as Ab-dependent cellular cytotoxicity (ADCC),³ release of oxygen radicals, endocytosis, phagocytosis, degranulation, Ag presentation, and release of inflammatory mediators (for review, see Ref. 1). In humans, one class of IgA FcR (FcαRI, CD89) and three classes of leukocyte IgG FcRs, i.e., FcγRI (CD64), FcγRII (CD32), and FcγRIII (CD16), have been described (2, 3). FcγRIIIb is a GPI-linked protein (4). FcγRIIa bears a so-called activatory ITAM signaling motif in its cytoplasmic tail, which is required for induction of effector functions, whereas FcγRIIb harbors an inhibitory ITIM-signaling motif (5, 6). The α-chains of FcαRI, FcγRI, and FcγRIIIa lack such signaling motifs. Therefore, these latter receptors associate with the common ITAM-containing FcR γ-chain to mediate effector functions (7, 8).

Of all FcγR, cell distribution of FcγRI most closely resembles FcαRI expression. Both receptors are selectively expressed on cells of the myeloid lineage, including monocytes, macrophages, and dendritic cells (2, 9). Furthermore, neutrophils constitutively express FcαRI, whereas FcγRI expression can be induced on these cells by addition of either IFN-γ or G-CSF (10, 11). Neutrophils represent the most populous cytotoxic effector cell subset within the blood; neutrophil numbers can be increased by treatment with G-CSF (12). They play a prominent role in bacterial infections (13), but exert well-documented antitumor properties as well because neutrophils have been shown to play a role in tumor rejection in the presence of antitumor Ab, both in vitro and in vivo (14, 15, 16, 17). Therefore, we—and others—studied the potential of FcαRI (18, 19) and FcγR (20, 21) on these immune cells to induce tumor cell lysis.

FcγRI was reported to represent the most potent neutrophil FcγR for induction of ADCC (20, 21). However, compared with FcγRI, tumor cell lysis was more efficient via targeting of neutrophil FcαRI (18, 19). This difference was most evident in bone marrow neutrophils, in which FcγRI-initiated tumor cell killing was hampered (19). Interestingly, cell surface expression of either receptor depends on association with the common FcR γ-chain, which was shown in FcαRI- and FcγRI-transgenic mice, in which surface expression was lost when these transgenic mice were crossed with FcR γ-chain-deficient mice (22, 23). Furthermore, for initiation of most immune effector functions, both receptors depend on signaling via the FcR γ-chain as well (24, 25). Therefore, in the present work, we studied the underlying mechanisms behind the discrepancies between FcαRI- and FcγRI-mediated effector functions in more detail.

Ab A77 (murine (m) IgG1 anti-FcαRI), m22 (mIgG1 anti-FcγRI), and 520C9 (mIgG1 anti-Her-2/neu) were produced from hybridomas (Medarex). FcγRIxHer-2/neu bispecific Ab (BsAb) (22 × 520C9; MDX-H210) was also obtained from Medarex. FcαRIxHer-2/neu BsAb (A77 × 520C9) was produced by chemically cross-linking F(ab′) of mAb 520C9 with F(ab′) of mAb A77 as described (26). For neutrophil staining, FITC-conjugated anti-CD11b mAb (Immunotech), PE-conjugated anti-CD16 mAb (BD Biosciences), and FITC-conjugated anti-human CD66b mAb (Serotec) were used. FITC-conjugated and unconjugated F(ab′)₂ of goat anti-mouse IgG1 mAb were purchased from Southern Biotechnology Associates. F(ab′)₂ of FITC-labeled rabbit anti-goat IgG (H + L) Ab were obtained from Jackson ImmunoResearch Laboratories. For Western blot analyses, rabbit anti-FcR γ-chain Ab (Upstate Biotechnology), rabbit anti-phospho-p44/42 MAPK, or rabbit anti-total MAPK Ab (Cell Signaling Technology), and peroxidase (PO)-conjugated goat anti-rabbit Ab (Pierce Biotechnology) were used.

The breast carcinoma cell line SK-BR-3, which overexpresses the tumor-associated Ag Her-2/neu, was obtained from the American Type Culture Collection. Cells were cultured in RPMI 1640 medium (Invitrogen Life Technologies) supplemented with 10% heat-inactivated FCS and antibiotics (RPMI 1640/10%). SK-BR-3 cells were harvested using trypsin-EDTA (Invitrogen Life Technologies).

A peripheral blood sample (30 ml) was drawn from healthy donors receiving recombinant human G-CSF (Neupogen; 5 μg/kg body weight, twice daily for 5 days; Amgen). Bone marrow samples were obtained from cardiac patients undergoing surgery. Studies were approved by the Medical Ethical Committee of Utrecht University (Utrecht, The Netherlands), in accordance with the Declaration of Helsinki. All donors gave informed consent.

Neutrophils from healthy G-CSF-stimulated donors (G-CSF neutrophils) were used as positive control for FcγRI-mediated effector functions and were isolated from heparin anticoagulated peripheral blood samples by standard Ficoll-Histopaque (Sigma-Aldrich) density gradient centrifugation.

Bone marrow neutrophils were isolated as described previously (19). In short, erythrocytes were removed by incubation for 5–10 min at 4°C with a lysis solution of pH 7.4 (0.16 M ammonium-chloride, 0.01 M potassium bicarbonate, and 0.1 mM sodium edetate), after which cells were separated by discontinuous Percoll gradient centrifugation (successively 81, 62, 55, 50, and 45% of Percoll). Percoll layers 1 and 5 in the gradient contained nonmyeloid cells, lipids, cellular debris, and remaining erythrocytes, respectively. Percoll layers 2, 3, and 4 comprised different neutrophil maturation stages.

Maturation status and purity of isolated human bone marrow cells was confirmed by cytospin preparations and staining with FITC-conjugated anti-CD11b mAb and PE-conjugated anti-CD16 mAb, as described previously (19, 27). Purity of isolated blood neutrophils was confirmed by cytospin preparation and staining with FITC-conjugated anti-human CD66b mAb. Neutrophil surface expression of FcαRI and FcγRI was determined with mAb A77 (FcαRI) or m22 (FcγRI), respectively (10 μg/ml), followed by incubation with FITC-conjugated F(ab′)₂ of goat anti-mouse IgG1 mAb. Cells were analyzed on a FACScan (BD Biosciences). In all experiments, neutrophil purity exceeded 95% and cell viability was >98%, as determined by trypan blue exclusion.

⁵¹Cr-release assays were performed as previously described (21). Briefly, SK-BR-3 target cells were incubated for 2 h at 37°C with ⁵¹Cr (100 μCi/1 × 10⁶ cells; Amersham), washed, and plated in 96-well round-bottom microtiter plates (5 × 10³ cells/well), together with 4 × 10⁵ neutrophils/well (E:T ratio of 80:1) in the presence of 2 μg/ml BsAb. After a 4-h incubation period at 37°C, ⁵¹Cr release in the supernatant was measured as cpm. The percentage of tumor cell lysis was calculated as follows: (experimental cpm − basal cpm)/(maximal cpm − basal cpm) × 100%.

Neutrophils were incubated with anti-FcαRI (A77), anti-FcγRI (m22), or, as an isotype control, irrelevant (520C9) mAb (10 μg/ml) for 30 min at 4°C. After washing, F(ab′)₂ of goat anti-mouse IgG1 mAb were added to cross-link neutrophil FcRs, and tubes were placed in a 953 LB Biolumat (Berthold). Luminol (150 mM) was injected in all tubes and light emission was recorded continuously for 30 min at 37°C. Addition of fMLP to neutrophils was used as a positive control.

Neutrophils were labeled with SNARF-1 (2.8 μM) and Fluo-3 (1.4 μM) (Invitrogen Life Technologies) for 30 min at 37°C. After washing, cells were incubated with anti-FcαRI (A77) or anti-FcγRI (m22) mAb (10 μg/ml) for 30 min at 4°C. Cells were washed twice and resuspended in calcium mobilizing buffer. FcRs were cross-linked with F(ab′)₂ fragments of goat anti-mouse IgG1 mAb; intracellular-free calcium levels were measured by FACScan. The first 20 s of each run, before cross-linking, were used to establish baseline intracellular calcium levels. fMLP was added to neutrophils as a positive control. To correct for variations between donors, levels of calcium flux were determined by the area under the curve (AUC), corrected for background levels by secondary Ab only.

Neutrophils were preincubated with 20% pooled human serum to prevent aspecific binding to IgG FcRs (30 min at 4°C). Thereafter, cells were incubated for 30 min at 4°C with anti-FcαRI (A77), anti-FcγRI (m22) or, as an isotype control, irrelevant (520C9) mAb. After that, cells were washed and incubated with F(ab′)₂ of goat anti-mouse IgG1 mAb (30 min at 4°C). Samples were then split. One sample was kept at 4°C to measure total surface expression of the FcRs. The other sample was put at 37°C for the indicated time points to allow internalization, which was stopped by adding ice-cold FACS buffer. Remaining FcαRI and FcγRI surface expression was visualized by staining for 30 min at 4°C with F(ab′)₂ of FITC-labeled rabbit anti-goat IgG (H + L) Ab. Internalization of FcRs at 37°C was calculated as percentage of initial FcR surface expression, which was determined in 4°C samples.

Neutrophils were incubated with anti-FcαRI (A77) or anti-FcγRI (m22) mAb (10 μg/ml) for 30 min at 4°C. After washing, FcRs were cross-linked with F(ab′)₂ of goat anti-mouse IgG1 mAb at 37°C for the indicated time points. Ice-cold PBS was added to stop reactions, after which samples were boiled in reducing Laemmli sample buffer, run on 10% SDS-PAGE gels, and electrotransferred to nitrocellulose membranes (0.45 μm; Millipore). Membranes were blocked with 5% BSA (Roche Diagnostics) and probed with rabbit anti-phospho-p44/42 MAPK, or rabbit anti-total MAPK Ab for 2 h. Following washing, membranes were incubated for an additional hour with PO-conjugated goat anti-rabbit Ab. Staining was visualized using the ECL detection system (Amersham).

Interaction of FcαRI and FcγRI with the FcR γ-chain was measured by lysing neutrophils (5 × 10⁷) with RIPA buffer (20 mM Tris (pH 7.4), 150 mM NaCl, 5 mM EDTA, 1% Nonidet P-40 (NP40), 0.5% deoxycholate, and 0.1% SDS) or NP40 buffer (10 mM Tris (pH 7.4), 137 mM NaCl, 13.5 mM sodium pyrophosphate, 51.9 mM 4-fluro-3-nitrophenyl azide, 10% glycerin, 0.5% NP40), both supplemented with 1 mM PMSF, 250 μM sodium orthovanadate, 10 mM DTT, and a protease inhibitor mixture (Roche Diagnostics) for 30 min at 4°C. Homogenates were spun for 20 min at 13,000 rpm. Supernatants were precleared head over head for 30 min at 4°C with protein A/G beads (Santa Cruz Biotechnology). After that, beads were removed, and anti-FcαRI (A77), anti-FcγRI (m22) or, as an isotype control, irrelevant (520C9) mAb was added overnight (4°C, head over head). Protein A/G beads were added for 3 additional hours (4°C, head over head), after which beads were washed three times with lysis buffer and boiled in reducing Laemmli sample buffer. Samples were run on 15% SDS-PAGE gels, electrotransferred to nitrocellulose membranes (0.45 μm; Millipore), and membranes were blocked with 5% low-fat milk powder in PBS. Membranes were probed with rabbit anti-FcR γ-chain Ab, followed by incubated with PO-conjugated goat anti-rabbit Ab. Staining was visualized using the ECL detection system (Amersham). Films were scanned with a GS-700 Imaging Densitometer and analyzed with Quantity One Software (both Bio-Rad).

Data are shown as mean ± SD. Group data are shown as mean ± SEM. Statistical differences were determined using the two-tailed unpaired Student’s t test or ANOVA. Significance was accepted when p < 0.05.

To study the mechanisms underlying the differences in FcαRI and FcγRI-mediated effector functions in neutrophil precursor subsets, cells were isolated from bone marrow and separated into three populations as described (19). In short, Percoll layer 2 (P2 neutrophil precursors) comprised the earliest neutrophil precursors, which are characterized by intermediate CD11b and low CD16 expression, as well as a round- to kidney-shaped nucleus. Percoll layer 3 (P3 neutrophil precursors) contained “intermediate” neutrophil precursors (defined by intermediate CD11b expression, heterogeneous CD16 expression, and a horseshoe-shaped nucleus). Percoll layer 4 (P4 neutrophil precursors) consisted of the most mature neutrophil precursors (high CD16 expression levels and a segmented nucleus) (19, 27).

First, FcαRI and FcγRI expression levels were determined (Fig. 1, A and B). In bone marrow, P2 neutrophil precursors expressed a high level of FcγRI, which was down-regulated during differentiation. An ∼3- ± 0.08-fold decrease in FcγRI expression was observed in P4 neutrophil precursors compared with P2 precursors in each donor (n = 4). On mature blood neutrophils, the level of FcγRI was low to absent (data not shown, n = 5), but could be up-regulated by G-CSF treatment. As such, experiments were only performed with G-CSF-stimulated blood neutrophils. FcαRI expression levels were low on P2 neutrophil precursors and were up-regulated during differentiation. In each donor, a 2.7- ± 0.4-fold increase in FcαRI expression was observed in P4 compared with P2 neutrophil precursors (n = 4). In blood, unstimulated as well as G-CSF neutrophils expressed high FcαRI levels.

The capacity of bone marrow neutrophils to lyse tumor cells via triggering of FcαRI or FcγRI was determined in ⁵¹Cr-release assays, in which SK-BR-3 tumor cells were used as targets. G-CSF-stimulated blood neutrophils were able to lyse tumor cells via triggering of FcγRI, but lysis was more efficient by targeting of FcαRI (Fig. 1 C). P2 neutrophil precursors were unable to mediate tumor cell killing via either FcR. P3 neutrophil precursors, which coexpressed FcαRI and FcγRI, triggered tumor cell killing via FcαRI, but not via FcγRI. P4 neutrophil precursors, with low expression of FcγRI, did not induce tumor cell lysis by targeting to FcγRI either, whereas lysis was mediated after triggering FcαRI.

We next investigated whether FcR-mediated tumor cell lysis was the sole effector function in which FcγRI activation was absent, or whether other neutrophil functions were impaired as well. Therefore, FcR-mediated respiratory burst activity was analyzed. Cross-linking of FcαRI or FcγRI on G-CSF neutrophils induced respiratory burst activity, albeit the activity mediated via FcγRI was consistently lower, compared with FcαRI (Fig. 2,A). FcαRI triggering on P2, P3, and P4 neutrophil precursors activated a respiratory burst as well, although the maximal burst activities and durations were lower, compared with respiratory bursts observed in G-CSF neutrophils (1.5–4.0 × 10⁶ cpm compared with 11.5 × 10⁶ cpm, respectively; Fig. 2, B–D). In P2 neutrophil precursors, which expressed the highest FcγRI level, cross-linking of FcγRI induced some respiratory burst activity. This burst level, however, was minimal and reached only 0.5 × 10⁶ cpm (Fig. 2,B). Cross-linking of FcγRI in P3 and P4 neutrophil precursors did not induce any respiratory burst activity (Fig. 2, C and D).

Furthermore, FcR γ-chain-mediated signaling was investigated by triggering either receptor. As P2 neutrophil precursors express low FcαRI levels, and P4 neutrophil precursors express low FcγRI levels, we used P3 precursors, which coexpress both FcRs, for additional experiments. FcR cross-linking induces tyrosine phosphorylation of ITAM, which triggers activation of the PI3K pathway, leading to calcium mobilization (28). In G-CSF neutrophils, cross-linking of either FcαRI or FcγRI induced an increase in intracellular-free calcium levels (Fig. 3,A). Although the total levels of FcαRI and FcγRI-mediated calcium release were similar (Fig. 3,C), FcγRI-mediated calcium mobilization was consistently slower compared with FcαRI-mediated mobilization. On average, FcαRI-mediated calcium mobilization peaked 29.8 ± 2.3 s after cross-linking, whereas FcγRI-mediated calcium mobilization peaked after 48.5 ± 2.9 s (n = 4, p < 0.05, data not shown). In P3 neutrophil precursors, however, a rise in intracellular-free calcium was only observed after cross-linking FcαRI (Fig. 3,B). FcγRI cross-linking did not result in release of intracellular calcium (Fig. 3 C).

FcR cross-linking furthermore activates the RAS-ERK pathway, thereby triggering phosphorylation of MAPK (P-MAPK) (29). We previously showed that cross-linking of either FcαRI or FcγRI on G-CSF neutrophils induced MAPK phosphorylation, albeit P-MAPK levels were higher following triggering of FcαRI, compared with FcγRI (19). In P3 neutrophil precursors, FcαRI cross-linking induced MAPK phosphorylation, which was detected within 15 min, and peaked after 60 min of cross-linking (Fig. 4). In contrast, cross-linking of FcγRI on these precursors did not result in P-MAPK (Fig. 4).

As all above-mentioned FcR γ-chain-dependent functions were induced by FcαRI but not by FcγRI, we next performed an internalization assay, which is an FcR γ-chain-independent function (24, 30). Cross-linking of either FcαRI or FcγRI on P3 neutrophil precursors induced efficient receptor internalization (Fig. 5, A and B). Within 5 min, both FcαRI and FcγRI levels decreased by 50–70% (Fig. 5 C). These data therefore suggest that the discrepancy between FcαRI and FcγRI-mediated functions was restricted to the FcR γ-chain pathway.

To evaluate association of both receptors with the FcR γ-chain, we established a coimmunoprecipitation assay. When P3 and P4 neutrophil precursors were lysed using mild conditions with RIPA buffer, FcR γ-chain was pulled down by FcαRI (81.4 arbitrary units (A.U.) for P3, and 84.9 A.U. for P4 precursors) as well as via FcγRI (68.9 for P3, and 76.4 A.U. for P4 precursors), indicating that both FcRs were associated with FcR γ-chain (Fig. 6,A). However, in a NP40 lysis buffer, FcR γ-chain was coimmunoprecipitated via FcαRI but not via FcγRI (A.U. for P3 neutrophils are 94.1 and 2.3, respectively; A.U. for P4 neutrophils are 92.9 and 15.6, respectively), indicating that association of FcγRI with FcR γ-chain was abrogated using this detergent (Fig. 6 B). In parallel experiments, similar results were observed with mature neutrophils (data not shown).

Triggering of neutrophil FcαRI induces more efficient tumor cell killing, compared with FcγRI (18, 19). Furthermore, immature neutrophils effectively lyse tumor cells via FcαRI, but not FcγRI (19). In the present study, we evaluated the mechanisms underlying the dissimilarities in FcR-mediated effector functions in different subsets of neutrophil precursors. We observed that ADCC, respiratory burst, and signaling functions like calcium mobilization and MAPK phosphorylation were effectively triggered by FcαRI, but not by FcγRI. FcR internalization, however, was readily induced by cross-linking either FcαRI or FcγRI. Formally, we cannot exclude the possibility that the observed differences are Ab-specific rather than receptor specific. However, this is less likely as the anti-FcαRI and anti-FcγRI Abs used in this study have similar affinities. Furthermore, FcαRI and FcγRI expression is similar on G-CSF stimulated as well as on P3 neutrophils. Targeting either FcαRI or FcγRI on G-CSF neutrophils induces signaling and function. However, when the same Abs are used to target P3 neutrophils, only FcαRI mediates signaling. Targeting FcγRI on P3 neutrophils with the same anti-FcγRI Ab that induces signaling in G-CSF neutrophils does not result in any functional activity, supporting that FcγRI signaling is hampered in immature neutrophils.

Interestingly, FcαRI signaling is believed to be mediated through signaling routes that are also used by FcγRI, requiring the common FcR γ-chain (7, 8, 22, 23). Earlier work showed that most effector functions, such as phagocytosis and cytokine production, by either FcγRI or FcαRI were dependent on the ITAM-signaling motifs within this subunit (24, 31). However, a few FcR γ-chain-independent functions have been described. Both FcαRI and FcγRI can mediate endocytosis without the FcR γ-chain. This was shown for FcγRI in transfection studies with COS cells (24, 32), and for FcαRI in colostral neutrophils (33), which express a population of FcαRI that is not associated with the FcR γ-chain, but can mediate IgA endocytosis (30). Because FcR internalization was the only effector function that was not hampered via FcγRI, our data suggest that the observed discrepancy between FcαRI and FcγRI-mediated function is the result of a difference between both FcRs in the FcR γ-chain-signaling pathway. Furthermore, because early signaling events, like calcium mobilization and MAPK phosphorylation, were not initiated by FcγRI, our data suggested that the interaction with the FcR γ-chain was affected.

This hypothesis was supported by coimmunoprecipitation studies, in which we found a less stable interaction between FcγRI and the FcR γ-chain, compared with FcαRI. This correlates well with earlier data in which FcαRI was found to bear a positively charged amino acid residue on position 209, which associates with a negatively charged amino acid of the FcR γ-chain, resulting in an electrostatic interaction within the transmembrane region (25). The positions of the positively charged amino acid residue was critical, as changing its position within the transmembrane region abrogated signaling and effector functions, except for FcR internalization, due to disturbance of the physical association between FcαRI and FcR γ-chain (34). FcγRI lacks such a positively charged amino acid in its transmembrane region, which might underlie a weaker association with the FcR γ-chain.

The FcR γ-chain is required for stable FcγRI expression, as expression was lost in the absence of the FcR γ-chain (23). Furthermore, FcεRI and FcγRIII were shown to compete for available FcR γ-chain in mast cells (35), a phenomenon that was also suggested to occur in neutrophils for FcαRI and FcγRI, when both FcRs were maximally engaged (36). In addition, neutrophils express relatively low levels of FcR γ-chain, compared with monocytes (37). Therefore, we postulate that due to a stronger association of FcαRI with the FcR γ-chain and limited availability of FcR γ-chain in immature neutrophils, FcαRI competes with FcγRI for available FcR γ-chain. This consequently leads to inability of immature neutrophils to signal via FcγRI, as well as to loss of FcγRI expression during maturation. As FcγRI triggering in P2 neutrophil precursors induced a small respiratory burst, some FcR γ-chain might have been available for FcγRI due to low FcαRI expression levels in these early precursors.

In G-CSF neutrophils, FcR γ-chain protein levels, measured by Western blotting, were considerably increased compared with levels in bone marrow and unstimulated blood neutrophils (data not shown, n = 3), indicating that more FcR γ-chain is available after G-CSF stimulation. This might lead to the observed up-regulation of FcγRI expression and function in G-CSF neutrophils (Fig. 1 and Ref. 20). We speculate that loss of FcγRI expression on mature neutrophils is favorable for maintaining homeostasis, as FcγRI is a high-affinity receptor for IgG (1) and, as such, may activate unwanted inflammatory responses. However, during bacterial infections, FcγRI-expressing immature neutrophils are recruited from the bone marrow, which may help to clear the infection (38). Neutrophils of patients with streptococcal pharyngitis were shown to express increased numbers of FcγRI (38).

Studies with FcαRI × FcγRI double transgenic mice supported the above-mentioned observations because, similar to human neutrophil precursors, mouse bone marrow neutrophils were unable to mediate ADCC via triggering of FcγRI, but not FcαRI (19). However, bone marrow neutrophils from G-CSF-treated mice could mediate ADCC via triggering of FcγRI, as well (data not shown, n = 4).

Taken together, our data suggest that inefficient signaling and effector functions via FcγRI on neutrophils is most likely induced by differences in interaction between FcαRI or FcγRI with the common FcR γ-chain. This may lead to competition between both FcRs for available FcR γ-chain in favor of FcαRI, hereby abrogating FcγRI function.

We thank Eefke Petersen and Faiz Ramjankhan for provision of blood and bone marrow samples, Martin Glennie and Allison Tutt for generation of FcαRI × Her-2/neu BsAb, and Paul Coffer, Erik-Jan Oudijk, Jantine Bakema, Jeffrey Beekman, and Olivier van Beekum for helpful discussions.

The authors have no financial conflict of interest.