Intracellular Expression of FcγRIII (CD16) and Its Mobilization by Chemoattractants in Human Eosinophils¹

FcγRIII (CD16) is a receptor for complexed IgG, and its presence has been used extensively for immunomagnetic separation of CD16⁻ eosinophilic from CD16⁺ neutrophilic granulocytes. Two very homologous genes code for this receptor and are expressed in a cell type-specific way (1). The FcγRIIIA gene is expressed as a transmembrane protein by NK cells (FcγRIIIa^NK) and macrophages (FcγRIIIa^mφ), whereas the polymorphic FcγRIIIB gene is constitutively expressed only by neutrophils as a glycosyl-phosphatidylinositol (GPI)³-linked protein. The polymorphism of the FcγRIIIB gene results in the codominant biallelic NA1/NA2 system (2). FcγRIIIa^NK (3), FcγRIIIa^mφ (4), and FcγRIIIb (5, 6) are all spontaneously released from the cell surface in experiments in vitro. Activation of the cells by phorbol esters or FMLP enhances the shedding of FcγRIIIa and -b (6, 7, 8, 9), and the release of CD16 also occurs in vivo (6, 9, 10). Ligand-dependent (11) and independent (12, 13) CD16 internalizations also have been found in human neutrophils.

Although constitutively present on the surface of neutrophils, estimations of the degree and mode of surface expression of CD16 on eosinophils differ substantively (14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25). The explanation for differing results has not been elucidated previously. In this investigation, we examined the existence of CD16 both on cell surface and intracellularly by specific mAb and flow cytometry. We found that CD16, which is expressed on the surface membrane of <7% of quiescent eosinophils, was constitutively expressed intracellularly in nearly all eosinophils. Eosinophil CD16 was of the isoform B type as determined by its positive reactivity to CLB-gran11, its sensitivity to phosphotidylinositol-specific phospholipase C (PI-PLC), and its identity with neutrophil CD16 on Western blots. During activation by chemoattractants, intracellular CD16 is expressed transiently at the surface and secreted rapidly into the perfusate. Disappearance from the perfusate corresponded to increase in total eosinophil CD16, suggesting reabsorption of CD16 after secretion.

Materials for eosinophil isolation were obtained from Miltenyi Biotec (Sunnyvale, CA). C5a, FMLP, platelet-activating factor (PAF), cytochalasin B (CB), mouse IgM, IgG2a, horseradish peroxidase-conjugated goat anti-mouse IgG and IgM Ab, horseradish peroxidase conjugated goat anti-mouse IgG Ab, parabenzoquinone (PBQ), p-nitrophenyl phosphate substrate tablet sets, and n-octyl-β-d-glucopyranoside (OG) were purchased from Sigma (St. Louis, MO). Polystyrene microtiter plates for ELISA were obtained from Costar (Cambridge, MA). Anti-CD16 mAb, 3G8, 3G8-phycoerythrin (PE), mouse IgG1-PE, mouse IgG1-FITC, and anti-very late Ag(VLA)-4 mAb HP2/1-FITC were purchased from Immunotech (Westbrook, ME). Anti-CD16 mAb, Leu11b, mouse IgG1, and FITC-conjugated goat anti-mouse Ig was purchased from Becton Dickinson (Mountain View, CA). CLB-gran11 and GRM1 mAb were purchased from Accurate Chemical and Scientific (Wesbury, NY). DJ130c mAb was purchased from Dako (Carpinteria, CA). The mAbs, 3G8, DJ130c, and Leu11b recognize both CD16-A and CD16-B, but bind to different epitopes of CD16 (26, 27). The binding epitope for 3G8 is located on the FG loop of the membrane-proximal Ig-like domain, and epitope for DJ130c is on the first membrane-distal domain (26). CLB-gran11 reacts with the NA1⁺ B isoform only, whereas GRM1 reacts with both the A and NA2⁺ B isoforms (26). PI-PLC was purchased from Boehringer Mannheim (Indianapolis, IN). Human B lymphoblastoid cell line JY was kindly donated by Dr. G. van Seventer (University of Chicago).

Eosinophils were isolated from donors having a history of mild atopy. These donors have been classified from previous studies, and we have shown previously that the stimulated secretory properties of eosinophils from these donors are identical to those of nonatopic donors (28). However, cell yield after negative immunoselection is improved vs the yield obtained from purely nonatopic individuals. Where neutrophils were used for comparison study, eosinophils from the same donors were used in each comparison.

The method for isolation of eosinophils used in this study was modified from that of Hansel et al. (29). Briefly, 120 ml whole blood was withdrawn from the antecubital vein and placed into containers containing 2 ml of 1:1000 heparin. Blood was diluted 1:1 with calcium-free HBSS, layered over 15 ml of 1.089 g/ml percoll and centrifuged for 20 min at 900 × g (all centrifugation done at 4°C). The supernatant and the mononuclear cells at the interface were carefully aspirated, and the inside wall of the tube was wiped with sterile gauze to remove mononuclear cells attached to the wall. To the pellet of granulocytes and erythrocytes, 20 ml of ice-cold sterile water was added and mixed gently for 30 s after which 20 ml of 2× HBSS was added. If erythrocytes remained, the procedure was again repeated. After erythrocyte lysis, granulocytes were washed once in HBSS/0.2% BSA, total cell numbers were counted by Coulter counter (Coulter Electronics, Hialeah, FL), and neutrophil percentage were calculated by differential counts of Wright-Giemsa stained cytospin preparation. The supernatant was carefully aspirated, leaving the pellet nearly dry. The pellet was cooled on ice, and 0.65 μl of CD16 beads (Miltenyi Biotec) per 10⁶ neutrophils was added. Granulocytes were incubated at 4°C for 30 min and then resuspend in 10 ml of HBSS/0.2% BSA. Granulocytes then were passed through a 1 × 10 cm column packed with steel wool and held within a 0.6 Tesla MACS magnet (Becton Dickinson). Cells were eluted with another 30 ml of HBSS/0.2% BSA. Neutrophils binding the Ab-magnetic bead were retained in the magnetized steel wool, whereas eosinophils passing through the column were collected, washed, and resuspended in HBSS/0.2% BSA. Count and purity were assessed as above. Eosinophils purity of >99% was routinely obtained. Cells were kept on ice until use.

Neutrophils were obtained as a by-product of eosinophil isolation. Before adding the CD16 beads, aliquots of granulocytes were taken as neutrophils; these preparations contained ≥95% neutrophils, with the remainder being eosinophils.

Granulocytes (mixed eosinophils and neutrophils) or purified eosinophils were double stained with directly conjugated 3G8-PE and HP2/1-FITC to determine whether eosinophils (HP2/1⁺ cells) expressed CD16 on the surface membrane. Aliquots of 10⁶ cells were incubated for 30 min at 4°C with 10 μl of specific mAb. Mouse IgG1-FITC and mouse IgG1-PE were used as isotype controls. The cells were washed twice, resuspended in PBS containing 0.1% NaN₃, and kept at 4°C until analyzed. Flow cytometry was performed on a FACScan (Becton Dickinson). VLA-4⁺/CD16⁺ subpopulations were determined with two-color flow cytometry of double stained cells. VLA-4 was used in studies of nonpurified granulocytes to separate eosinophils from neutrophils by two color flow cytometry. This method utilizes the absence of VLA-4 on neutrophils to identify eosinophils on which this integrin is expressed constitutively (30). Gates for positive and negative populations were set in a similar fashion for all groups. Isotype control mAbs were used to determine any nonspecific Ab binding.

One-color flow cytometry was performed on eosinophils to detect surface CD16 expression. Eosinophils were fixed in 1% paraformaldehyde for 20 min at 4°C. Aliquots of 5 × 10⁵ eosinophils were incubated with 10 μl of specific mAb (3G8 or Leu 11b) or isotypic control for 30 min at 4°C. After two washes, the cells were incubated with an excess of FITC-conjugated goat anti-mouse Ig for 20 min at 4°C. The cells were washed twice, resuspended in PBS containing 0.1% NaN₃, and kept at 4°C until analyzed. Fluorescence intensity was determined on at least 5000 cells from each sample. The results were expressed as specific mean fluorescence intensity (MFI, control Ab fluorescence subtracted).

Granulocytes at a concentration of 5 × 10⁶/ml were incubated for 1 h at 37°C with HBSS or 0.5 U/ml PI-PLC from Bacillus cereus. PI-PLC digestion is specific for the CD16B isoform (GPI-linked) that is common to eosinophils and neutrophils (see Results). As an additional control, we also treated PBMCs isolated from the same preparations with PI-PLC. These mononuclear cells include NK cells, which contain the CD16A receptor, i.e., the transmembrane form of the CD16. All cells were then stained with 3G8-PE as above.

Fixation and permeabilization of eosinophils for indirect fluorescent staining of intracellular CD16 was performed by the FOG (fixation and membrane permeabilization with OG) method previously described by Krug et al. (31). The use of the fixative PBQ instead of paraformaldehyde reduces the nonspecific binding of FITC-labeled Abs to permeabilized eosinophils. Briefly, eosinophils were fixed in 0.4% PBQ in 10 mM PBS for 10 min. After a further wash in PBS, cells were permeabilized by incubation in 0.74% OG in PBS for 6 min. Permeabilized cells were washed in PBS, and mAbs against CD16 and mouse isotypic controls (IgG1, IgG2a, IgM) were added in optimal concentrations. After 30 min of incubation at 4°C, cells were washed in PBS and incubated with FITC-labeled goat anti-mouse Ig as a secondary Ab for 20 min. After a final wash in PBS, the cells were resuspended in PBS and kept at 4°C in the dark until analyzed. JY lymphocytes were treated in the same way and served as a negative control.

To demonstrate specificity of intracellular CD16 staining, the anti-CD16 mAb, 3G8 (20 μl at 200 μg/ml), was absorbed sequentially with three aliquots of pooled fixed neutrophils consisting of 2 × 10⁶ cells. Each absorption was done for 30 min at 4°C on a rocking platform. Cells were removed by centrifuging at 2000 × g for 5 min. Control absorptions were done with JY B cells. Neutrophil and JY B-absorbed samples were compared with an untreated aliquot of 3G8 for their reactivity with peripheral blood eosinophils.

The NA phenotype of neutrophils were determined by indirect immunofluorescence flow cytometry as described above with NA1- or NA2-specific mAbs (CLB-gran11 and GRM1, respectively).

Eosinophils (4–8 × 10⁶) were lysed in 1 ml of lysis buffer (20 mM Tris-HCl, 30 mM Na₄P₂O₇, 50 mM NaF, 40 mM NaCl, 5 mM EDTA, pH 7.4) containing 1% Nonidet P-40, 10 μg/ml leupeptin, 5 μg/ml aprotinin, 1 mM PMSF, 2 mM Na₃VO₄, and 0.5% deoxycholic acid. After 10 min on ice, the sample was centrifuged at 12,000 × g for 20 min to remove nuclear and cellular debris. The supernatants then were mixed with 50 μl of 3G8 and shaken for 90 min; 30 μl of protein A-Sepharose beads then was added, and incubation was continued for another 30 min. The immunoprecipitated proteins were washed four times with lysis buffer. Afterward, 50 μl of sample buffer without reducing agent was added and boiled for 5 min. The supernatant was collected and saved at −70°C. Similar procedures were used in human neutrophils as a positive control.

Aliquots of immunoprecipitated protein were subjected to SDS-PAGE by the method of Laemmli (32), using 7.5% acrylamide gels under nonreducing condition. Electrotransfer of proteins from the gels to polyvinylidene difluoride membrane was achieved using a semidry system (400 mA, 1 h). The membrane was blocked with 1% BSA for 60 min, then incubated with 4 μg/ml of DJ-130c diluted in TBS-T for 1 h. The membranes then were washed three times for 20 min with TBS-T. Goat anti-mouse IgG-conjugated with horseradish peroxidase was diluted 1:2500 in TBS-T and incubated with membrane for 1 h. The membrane was again washed three times with TBS-T and assayed with ECL chemiluminesence system (Amersham, Burlington Heights, IL).

Eosinophils were preincubated with CB (5 μg/ml) for 2 min and then stimulated with different concentrations of C5a, FMLP, and PAF for various times. The addition of CB was used to promote both cellular degranulation and cytokine secretion as described previously (Refs. 33 and 34; see also Discussion). The reaction was stopped quickly by centrifugation at 10,000 × g for 10 s, and eosinophils then were fixed with 1% paraformadyhyde in PBS. Eosinophils then were resuspended in PBS buffer containing 0.5% BSA and incubated further for 10 min to block nonspecific binding before analysis by immunofluorescence flow cytometry.

This ELISA was modified from the method of Khayat et al. (35). Briefly, microtiter plates were coated with 100 μl of 30 μg/ml mouse IgG1 3G8 mAb overnight at 4°C. Wells were blocked by addition of 200 μl of 1% BSA in TBS (20 mM Tris-HCl, pH 7.40, 500 mM NaCl) for 30 min at 37°C, after which wells were washed four times with buffer (0.05% Tween-20 in TBS, pH 7.40). Plates then were stored at −20°C until use. Samples (100 μl) diluted in TBS/0.1% BSA were added and incubated for 1.5 h at room temperature with shaking. The plates were washed four times with wash buffer, followed by addition of 100 μl/well of 0.8 μg/ml IgM anti-Leu11b mAb and incubated for 1 h at room temperature. The plates were washed again followed by addition of 100 μl of 1:2500 dilution of goat anti-mouse IgM conjugated with alkaline phosphatase and incubated for 1 h at room temperature. The plates were washed again followed by the addition of 100 μl substrate/well (p-nitrophenyl phosphate, 1 mg/ml) and incubated for 30 min at 37°C, after which the plates were read at 405 nm using a microplate absorbance spectrophotometer (Thermomax, Molecular Devices, Menlo Park, CA). All assays were performed in duplicate, and values are given as means. Data storage and analysis were facilitated by the use of computer software interfaced with the reader (Softmax, Molecular Devices).

All data are expressed as mean ± SEM. Differences between groups were assessed by paired t test. Where more than two groups were compared, differences among groups were assessed by one way analysis of variance. Where differences were founds, comparisons among groups were made by Fisher’s least protected difference test. Statistical significance was claimed where p < 0.05.

Figure 1 shows results of a representative experiment in which the pattern of reactivity of Abs 3G8 and Leu11b on the surface of isolated peripheral blood eosinophils was determined by indirect immunofluorescence. Surface CD16 could not be detected above the background fluorescence for these eosinophils, which had been separated by negative immunoselection (see also Fig. 2 F below).

We further examined surface CD16 expression on nonpurified and purified eosinophils by two-color flow cytometry. In these studies, eosinophils were separated from neutrophils by positive staining for VLA-4, which is absent on the surface of neutrophils (see Materials and Methods). Figure 2 is a representative dot plot of nonpurified granulocytes and of purified eosinophils stained for CD16 and VLA-4. Isotype control showed nonspecific staining (Fig. 1,A). Mixed granulocytes were separated into two population by anti-VLA-4 mAb, HP 2/1; the lower left quadrant is neutrophils and the lower right quadrant identified eosinophils (Fig. 1,B). Figure 2,C identifies mixed granulocytes according to CD16 positivity. CD16 was expressed on the surface of 6.5 ± 0.3% of all VLA-4⁺ eosinophils (n = 5 separate donors) (Fig. 2,D). Both eosinophils and neutrophil CD16 were the GPI-linked B isoform. PI-PLC treatment reduced surface CD16 on both eosinophils and neutrophils from these nonselected preparations (Fig. 2,E). MFI for neutrophil CD16 decreased from 988 ± 52.1 to 394 ± 15.6 (p < 0.001) after treatment with PI-PLC; MFI decreased comparably (after gating) for CD16⁺ eosinophil from 976 ± 51.9 to 372 ± 10.2 (p < 0.001). Control studies performed with PBMC that contained the NK CD16A transmembrane isoform of the receptor showed no reduction in CD16 expression after PI-PLC (data not shown). Figure 2,F is the two-color flow cytometry for eosinophil purified by negative immunoselection using anti-CD16 mAb. No CD16 positive eosinophils are demonstrated by anti-VLA-4 mAb after negative immunoselection (see also Fig. 1).

In contrast to the rare surface expression on eosinophils, CD16 was detected intracellularly in nearly all eosinophils from NA1/NA2 or NA1/NA1 donors by all four anti-CD16 mAb used in these studies, each of which recognizes a different CD16 epitope (Fig. 3). CLB-gran11 did not bind eosinophils from NA2/NA2 donors intracellularly (data not shown). By contrast, identically permeabilized JY B lymphocytes, which served as control cells, did not demonstrate CD16 expression for any of the four mAbs demonstrating intracellular CD16 in eosinophils (Fig. 3).

The specificity of CD16 mAb binding to eosinophils during intracellular flow cytometry was assessed by absorption studies. CD16 mAb recognized specifically CD16 Ag in eosinophils. After absorption of 3G8 with sufficient numbers of CD16⁺ neutrophils, binding to eosinophils was eliminated (n = 5, p < 0.05) (Fig. 4). JY cells shown above to be negative for CD16 expression were used for control absorption; this treatment with comparable cell numbers did not affect 3G8 reactivity with eosinophils. MFI was 17.1 ± 4.7 before treatment with JY cells and 16.8 ± 3.6 after JY B cells absorption (p = NS).

Figure 5 is a Western blot after immunoprecipitation with 3G8 mAb as detected by DJ130c. This blot demonstrates a smear of 65- to 80-kDa protein that is comparable qualitatively for both eosinophils and neutrophils from the same NA2/NA2 donor. The fainter pattern in eosinophils reflects the lesser amount of CD16 contained in these eosinophils compared with neutrophils. For all other donors, the m.w. of CD16 also was comparable for both eosinophils and neutrophils. In eosinophil lysate from donors identified as NA1/NA2 by CLB-gran11 and GRM1 reactivity (data not shown), DJ130c and 3G8 reacted with a broad band ranging from 50 to 80 kDa, whereas CLB-gran11 reacted with a narrow band ranging from 50 to 65 kDa, a pattern similar to that obtained with the same mAb in neutrophils from the same donor (not shown).

In these studies, all cells were incubated with CB initially for 2 min plus either C5a, FMLP, or PAF over a broad concentration range (see Materials and Methods). Stimulation of human eosinophils with C5a, FMLP, or PAF caused rapid surface expression of CD16 in a concentration-dependent manner (Fig. 6). C5a caused concentration-dependent increase in surface expression of CD16 from 5.41 ± 2.55 MFI to a maximum of 58.5 ± 7.35 MFI at 10⁻⁸ M (p < 0. 001). FMLP up-regulated CD16 expression from 0.79 ± 0.24 MFI to 47.3 ± 3.92 MFI at 10⁻⁷ M (p < 0.001) (Fig. 6). Similarly, PAF caused significant surface mobilization of CD16, but only at high concentrations (>1 μM). Comparable maximal MFI occurred with approximately 10⁻⁸ M C5a, 10⁻⁷ M FMLP, and 10⁻⁵ M PAF (p < 0.05 for all groups compared with nonstimulated controls; ANOVA).

To examine the kinetics of eosinophil-surface CD16 expression caused by C5a, FMLP, and PAF, CB-treated eosinophils were stimulated with optimal concentration of C5a, FMLP, or PAF (determined above) and compared with buffer controls at various time periods for surface CD16 expression. Increased surface CD16 expression was maximal at ∼30 s for all chemoattractants (p < 0.05 for all groups compared with nonstimulated control), and MFI decreased gradually over the next 10 min (Fig. 7). By 10 min, MFI for C5a was 21.8% of maximal, FMLP was 8.8% of maximal, and PAF was 21.6% of maximal surface expression. Eosinophils incubated with 5 μg/ml CB alone did not up-regulate surface CD16 expression.

To determine whether the rapid decrease in surface CD16 expression after activation resulted from secretion, degradation, or reabsorption, we measured the amount of total CD16 in stimulated eosinophils by intracellular flow cytometry. As shown in Figure 8, 30 s after stimulation by FMLP/CB, total (intracellular plus surface) CD16 in the cells was 47.3 ± 3.3% of that of unstimulated eosinophils (p < 0.01), indicating secretion into the extracellular medium (Fig. 8). Thereafter, the total amount of CD16 in the cells gradually increased. By 20 min after initial stimulation, total CD16 content was 90.9 ± 7.8% of that of unstimulated eosinophils, suggesting possible reabsorption of CD16 receptor.

To confirm further the secretion of the soluble CD16 in FMLP-stimulated eosinophils, soluble CD16 was measured in perfusate over the same time period as above using sandwich ELISA (Fig. 9). The amount of CD16 released from FMLP-stimulated eosinophils was directly related to the duration of exposure to this agent. CD16 was released maximally at 60 s and disappeared from the perfusate within 10 min. These data mirror the reappearance of total eosinophil CD16 after stimulation as shown in Figure 8.

The objective of this investigation was to elucidate the potential existence, cellular content, biochemical structure, and fate of the CD16 receptor of the eosinophil before and during cellular activation. Prior investigations have reported considerable numbers of FcγRIII receptors on eosinophil (>43,000 receptors/cell) using anti-CD16 mAbs, 3G8, and 4F7 (14, 15, 16, 17), whereas other studies have shown insignificant surface presence of the CD16 receptor on eosinophils using 3G8 or CLB-FcR-1 mAbs (18). Other investigators showed absent CD16 expression on the surface of freshly isolated hypodense or normodense eosinophils using mAbs, Leu 11b, 3G8, CLB-FcRgran1 and BW209/2 (22, 23, 24). The absence of this CD16 surface receptor from the eosinophil, but not the neutrophil, surface membrane has become the basis for immunomagnetic separation and purification of eosinophil.

We examined the potential mechanism by which CD16 receptor could be expressed transiently on the eosinophil surface and further characterized potential structural similarities between the CD16 receptor of the eosinophil and that of the neutrophil. We hypothesized that CD16 was not expressed on the surface in significant numbers of eosinophils in the resting (nonactivated) state. We further hypothesized that CD16 existing within the eosinophil could be expressed transiently as a surface receptor after activation (14, 15, 16, 17, 22, 23, 24). To test this hypothesis, we permeabilized eosinophils from mildly atopic humans and used a panel of CD16 mAbs to identify CD16 intracellularly. We examined the surface expression of eosinophil CD16 before and after negative immunoselection, which utilizes mAb directed against the CD16 receptor to separate neutrophils from eosinophils (22, 29). In these studies, two-color flow cytometry was used to identify eosinophils containing VLA-4 (an integrin not present on neutrophils) and CD16. Although expressions was ubiquitous for neutrophils, <7% of unsorted eosinophils expressed surface CD16 (Fig. 2,D). Two-color flow cytometry also confirmed the specificity of our data; eosinophils extracted by negative immunoselection using CD16 mAb did not express any surface CD16 (Fig. 2 F).

In further studies, we examined the effects of PI-PLC on both eosinophil and neutrophil CD16. After treatment, reduction of MFI in cells containing CD16 was comparable (Fig. 2,E), suggesting that both neutrophil and eosinophil CD16 are GPI-linked isoforms. By contrast, mononuclear cells, which include CD16⁺ NK cells, treated identically did not demonstrate a GPI subunit. These results suggest biochemical similarity between eosinophil and neutrophil CD16. While CD16 did not exist constitutively on the surface membrane in substantial numbers of unstimulated eosinophils (Figs. 1 and 2), it was readily identified within normodense, unstimulated cells (Fig. 3). This is the first identification of this receptor in the human eosinophil.

The positive intracellular reactivity to CLB-gran11 from NA1/NA2 or NA1/NA1 donors indicates that eosinophil CD16 is not an A isoform. Western blot analysis with 3G8 mAb demonstrated a 65- to 80-kDa band from both eosinophils and neutrophils in NA2/NA2 donors (Fig. 5). These results further substantiate the biochemical identity of eosinophils CD16 as a GPI-linked B isoform.

Further studies were performed on the effects of chemoattractants that induce transient surface expression of eosinophil CD16. In eosinophils, chemoattractants have been shown to cause generation of oxygen intermediates (36), granular protein release (33, 37), lipid mediators (33), and cytokines including IL-8, IL-5, and granulocyte/macrophage colony-stimulating factor (34). In this report, we also demonstrate that chemoattractants cause transient expression of surface CD16, which is not present in substantial quantities in the quiescent state (Figs. 1 and 2). Our data show that chemoattractants induce transient membrane expression of CD16 (Fig. 7), concomitant decrease in the total amount of CD16 detectable in eosinophils (Fig. 8), and a concomitant release of soluble CD16 into the supernatant (Fig. 9). The amount of membrane CD16 (Fig. 7), of total cellular CD16 (Fig. 8), and of soluble CD16 in the supernatant (Fig. 9) gradually returned to baseline level within 5 to 20 min. The mechanism for this apparent reabsorption was not established in these investigations. Several possibilities may account for these findings including: 1) recylcing of membrane CD16, 2) binding of shed CD16 to the cell surface, which possibly then could be followed by 3) internalization.

The biologic role of surface-expressed or secreted eosinophil CD16 remains unknown. It is possible that secreted receptor could serve to down-regulate cellular responses to IgG complexes in inflammatory states by binding these complexes on the cell surface. Huizinga et al. (2) demonstrated that soluble CD16 in human plasma or serum is associated with monomeric IgG. It also is possible that CD16 may down-regulate IgG production in humoral immunity. Soluble CD16 released from human neutrophils has been shown to interfere with differentiation of peripheral blood B cells into Ig-secreting cells in vitro (38). Recombinant soluble CD16 containing the extracellular region of CD16B also has been shown to bind human IgG1 and IgG3 and to inhibit proliferation of IgG and IgM production of human PBMCs stimulated by pokeweed mitogen in vitro (39). Although the significance of in vitro reabsorption of CD16 by eosinophils after stimulated secretion is unknown, chemotactic stimulation of eosinophils can stimulate uptake of extracellular molecules (40).

We demonstrate for the first time the unequivocal presence of CD16 receptor in the human eosinophil. This receptor is expressed on the cell surface in significant numbers of eosinophils only transiently after activation by a variety of chemoattractants. Thereafter, surface expression rapidly decreases, and secreted CD16 appears to be taken up by the eosinophil by a mechanism that remains to be established. Although secreted CD16 may have several important immunoregulatory roles, most interestingly in B lymphocytes, further studies are required to define the role of this secreted IgG receptor in inflammatory disease.