HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Chuang, F. Y. S. Articles by Unkeless, J. C. Search for Related Content PubMed PubMed Citation Articles by Chuang, F. Y. S. Articles by Unkeless, J. C.

Convergence of Fc Receptor IIA and Fc Receptor IIIB Signaling Pathways in Human Neutrophils¹

Frank Y. S. Chuang^*,, Massimo Sassaroli^2, and Jay C. Unkeless^3,

Departments of ^* Biochemistry and Physiology and Biophysics, and Immunobiology Center, Mount Sinai School of Medicine, New York, NY 10029

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
Human neutrophils (PMNs) express two receptors for the Fc domain of IgG: the transmembrane FcRIIA, whose cytosolic sequence contains an immunoreceptor tyrosine-based activation motif, and the GPI-anchored FcRIIIB. Cross-linking of FcRIIIB induces cell activation, but the mechanism is still uncertain. We have used mAbs to cross-link selectively each of the two receptors and to assess their signaling phenotypes and functional relation. Cross-linking of FcRIIIB induces intracellular Ca²⁺ release and receptor capping. The Ca²⁺ response is blocked by wortmannin and by N,N-dimethylsphingosine, inhibitors of phosphatidylinositol 3-kinase and sphingosine kinase, respectively. Identical dose-response curves are obtained for the Ca²⁺ release stimulated by cross-linking FcRIIA, implicating these two enzymes in a common signaling pathway. Wortmannin also inhibits capping of both receptors, but not receptor endocytosis. Fluorescence microscopy in double-labeled PMNs demonstrates that FcRIIA colocalizes with cross-linked FcRIIIB. The signaling phenotypes of the two receptors diverge only under frustrated phagocytosis conditions, where FcRIIIB bound to substrate-immobilized Ab does not elicit cell spreading. We propose that FcRIIIB signaling is conducted by molecules of FcRIIA that are recruited to protein/lipid domains induced by clustered FcRIIIB and, thus, are brought into juxtaposition for immunoreceptor tyrosine-based activation motif phosphorylation and activation of PMNs.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
Polymorphonuclear leukocytes (PMNs⁴ or neutrophils) play a central role in Ab-mediated cellular immunity. Interaction of FcR on the surface of PMNs with the Fc domains of IgG molecules in immune complexes or on opsonized targets elicits a pleiotropic response, which includes phagocytosis, degranulation, and an oxidative burst. Receptor cross-linking upon binding to multivalent ligands, rather than ligand binding per se, constitutes the critical event leading to intracellular signaling and cell activation. Human PMNs are unique for their constitutive expression of two atypical FcR isoforms, FcRIIA (CD32) and FcRIIIB (CD16B). Unlike other multichain Fc receptors, in which Fc binding and signaling domains are segregated to different subunits, the cytosolic sequence of FcRIIA contains a slightly modified immunoreceptor tyrosine-based activation motif (ITAM) consisting of two YXXL (where X denotes any amino acid) repeats separated by 12 aa (1, 2). Upon cross-linking, the receptors are brought into juxtaposition, and Src family kinases phosphorylate the conserved ITAM tyrosine residues. Phosphorylated ITAMs then function as docking sites for proteins containing tandem Src homology 2 (SH2) domains, such as tyrosine kinases of the Syk family (3) or the p85 subunit of phosphatidylinositol 3-kinase (PI3K) (4), leading to downstream signaling events.

The second FcR isoform, FcRIIIB, expressed exclusively on human PMNs, is anchored to the plasma membrane via a C-terminus-linked GPI moiety and thus lacks any obvious means of signal transduction upon cross-linking. However, with a 10-fold higher abundance (135,000 vs 10,000 receptors/cell) (5) and a higher affinity for IgG than FcRIIA, it may play a predominant role in PMN binding of immune complexes.

Because both FcR isoforms are likely to be engaged by immune complexes, the questions of whether and how the GPI-anchored receptor may complement FcRIIA function have been subject to debate. Although one view is that FcRIIIB serves merely to enhance immune complex binding for presentation to FcRIIA, clear evidence supports an active role for the GPI-anchored isoform in signaling and PMN activation. Thus, without FcRIIA ligation, FcRIIIB cross-linking induces a rise in the intracellular free calcium concentration ([Ca²⁺]_i) and triggers degranulation and the respiratory burst (6, 7). Co-cross-linking of both FcRs also leads to synergistic enhancement of [Ca²⁺]_i transients and the phagocytic response (8, 9). In this study we have investigated PMN activation by specific Ab-mediated cross-linking of each of the two FcRs to compare their signaling phenotypes and to assess their functional relation.

Whereas the essential role of PI3K in FcR signaling is known, here we show that the PI3K inhibitor wortmannin blocks with identical efficacy the [Ca²⁺]_i transients elicited by cross-linking FcRIIIB or FcRIIA. Contradictory evidence exists about the role of phospholipase C and the amount of IP₃ generated upon FcR engagement (10, 11, 12, 13, 14). Prompted by a report that the [Ca²⁺]_i rise upon clustering of FcRI in a rat mast cell line was mediated by sphingosine-1-phosphate (S1P), the product of sphingosine kinase (SK) (15), we have examined whether this pathway is used by FcRs in PMNs.

The molecular basis for signaling by the GPI-anchored receptor remains unclear. FcRIIIB is just one of a large group of unrelated proteins anchored via GPI to the PMN surface. No common functional theme has been found for this elaborate post-translational modification. In T lymphocytes, cross-linking any GPI-anchored protein was shown to lead to cell activation mediated by the TCR/CD3 complex (16, 17). Several models for signaling by GPI-anchored proteins have been proposed, invoking a role for either the glycosidic or the lipidic components of the GPI moiety (18, 19). Using immunofluorescence microscopy, we have examined the effect that specific Ab-mediated cross-linking of FcRIIIB has on the surface distribution and colocalization of FcRIIA with aggregated FcRIIIB. In analogy with the "signaling raft" model (19), we propose that aggregation of FcRIIIB leads to signal transduction via formation of protein/lipid domains to which signaling-competent molecules, such as FcRIIA and protein tyrosine kinases, are recruited.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Cells

A murine macrophage cell line, P388D₁, was transfected with wild-type (wt) or mutant human FcRIIA cDNA as previously described (20). Cells transfected with wt (designated PW16) or truncated (233 and 264) FcRIIA express 1.1–1.8 x 10⁶ receptors/cell. Human PMNs were isolated from buffy coat (Leukopac) preparations obtained from the Blood Donor Center of the Mount Sinai Hospital. PMNs were collected from the 1.119 g/ml interface of a Histopaque (Sigma, St. Louis, MO) density gradient and washed in DMEM (Sigma) containing 2% heat-inactivated FCS and 20 mM HEPES (pH 7.4). Smaller scale preparations were obtained from whole venous blood by centrifugation using Polymorphprep (Life Technologies, Gaithersburg, MD). PMNs were held at room temperature in incubation buffer (150 mM NaCl, 5 mM KCl, 1 mM CaCl₂, 10 mM glucose, and 20 mM HEPES (pH 7.5)), unless otherwise noted. In both protocols, contaminating RBCs were removed by a 30-s hypotonic lysis.

Labeling of Abs

Anti-FcRIIIB 3G8 mAb was obtained from Rhone Poulenc (Antony, France). The anti-FcRIIA IV.3 monoclonal cell line was from American Type Culture Collection (Manassas, VA). IV.3 IgG and Fab were prepared as previously described (20). Fab were labeled with amine-reactive probes (fluorescein, rhodamine and Texas Red isothiocyanate or succinimidyl esters; Molecular Probes, Eugene, OR) in 0.15 M bicarbonate buffer (pH 8.5) for 4 h at 4°C or 1 h at room temperature. Excess probes were blocked with 0.15 M ethanolamine and removed by passage over a Sephadex G-25 column. Absorbance measurements showed conjugation ratios of 2–3 fluorophores/Fab. Biotinylated Fab were prepared by reacting 1 mg protein with 1.2 mg sulfosuccinimidyl-6-(biotinamido)hexanoate (LC-biotin; Pierce, Rockford, IL) in 1 ml of 0.1 M bicarbonate buffer (pH 8.2) for 2 h on ice. Approximately 2 µg/ml IV.3 or 5 µg/ml 3G8 Fab were added to saturate labeling of FcRIIA or FcRIIIB in PMN suspensions (10⁷ cells/ml).

Spectroscopic determination of [Ca²⁺]_i

PMNs (6 x 10⁶ cells/ml) were incubated with 1.5 µM indo-1/AM Ca²⁺ indicator (Molecular Probes) and 3G8 or IV.3 Fab for 20–30 min at room temperature. Where indicated, wortmannin (Biomol Research, Plymouth Meeting, PA) was also added during this interval. Thereafter, the PMNs were gently centrifuged, suspended (0.5–2 x 10⁶ cells/ml) in incubation buffer or balanced salt solution (135 mM NaCl, 4.5 mM KCl, 0.5 mM MgCl₂, 1 mM CaCl₂, 5.6 mM glucose, and 10 mM HEPES (pH 7.4)), and placed in a 1 cm fluorescence cuvette. DMS (Biomol Research) was added in the final resuspension together with 0.1 mg/ml fatty acid-free albumin. Fab-labeled FcRs were cross-linked by adding about 30 µg/ml F(ab')₂ of goat anti-mouse IgG (GaM; Jackson ImmunoResearch, West Grove, PA). Dual wavelength ratiometric measurement of indo-1 fluorescence, calibration and calculation of [Ca²⁺]_i were performed as previously described (21). As a positive control and to assess cell viability, 100 nM fMLP (Sigma) was added to the cell suspension after the fluorescence ratio had returned to baseline.

Frustrated phagocytosis assay

Proteins were covalently coupled to glass as previously described (22). Dry, acid-washed glass dishes or coverslips were derivatized with 3-aminopropyltriethoxysilane (Sigma) for 4 min, rinsed, reacted with 0.25% glutaraldehyde (Sigma) for 30 min, rinsed, and incubated with 5 µg/ml GaM in 0.1 M Na₂CO₃ (pH 10) for 1 h. After a final rinse with PBS, residual reactive sites were quenched with 2% FCS. Cells (1 x 10⁶ cells/6-cm diameter dish) in DMEM with 2% FCS and 20 mM HEPES were plated onto the derivatized surfaces and kept for 20 min at 37°C. FcRIIA or FcRIIIB was ligated to the surface by adding IV.3 or 3G8 Fab (1–5 µg/ml), respectively, and the cells were maintained at 37°C for various time intervals. Where indicated, wortmannin was added 10 min before cell activation.

To quantify the phagocytic response by image-based cytometry, adherent cells were fixed with 0.2% glutaraldehyde in PBS, stained for 15 min with 0.2% Coomassie blue R-250 in 20% methanol, washed with 5% acetic acid, dried, and mounted in glycerol. Digital images were acquired using a Zeiss Axiovert microscope (Carl Zeiss, Thornwood, NY), a x10, 0.25 normal aperture (NA) or a x40, 0.75 NA objective, a CCD camera (OMA Vision, EG&G PARC, Princeton, NJ), and 560 nm transillumination light selected by a 40-nm bandpass filter (Omega Optical, Brattleboro, VT). Image analysis was performed using Image-1 (Universal Imaging, West Chester, PA). Each image was flat-fielded to correct for uneven illumination and sensitivity, median-filtered to reduce noise, and thresholded to separate dark cells from bright background (23). Individual cell areas were measured after calibration of pixel dimensions with a stage micrometer. Cell fragments and aggregates were rejected based on their size or irregular shape.

FcR internalization assay

F(ab')₂ of rabbit anti-goat IgG (RaG; Jackson ImmunoResearch) were radiolabeled with NA¹²⁵I using Iodogen as previously described (20). Triplicate aliquots of PMNs (100 µl, 1 x 10⁶ cells/ml) were labeled with 3G8 or IV.3 Fab (and wortmannin where indicated) in thin-walled microtubes and held at 4°C on a programmable thermal cycler (MJ Research, Watertown, MA). After addition of GaM (30 µg/ml) and incubation for 10 min, the cells were warmed to 37°C for 0, 2, or 10 min to allow receptor internalization. They were then returned to 4°C, centrifuged, and resuspended in ice-cold buffer containing 5 µg/ml ¹²⁵I-conjugated RaG. After 20 min, they were washed twice by centrifugation in 1.5 ml of ice-cold PBS containing 5% FCS. The pellets were dispersed in 100 µl of PBS, and the radioactivity was measured using a gamma counter (1271 Riagamma, LKB Wallac, Turku, Finland).

Immunofluorescence microscopy

PMNs were labeled in suspension with rhodamine-3G8 or fluorescein-IV.3 Fab and transferred to chambers with a coverslip bottom. After addition of 30 µg/ml cross-linking GaM, fluorescence image sequences documenting the receptor aggregation were acquired using the OMA Vision CCD camera, a Zeiss Axiovert microscope, a Plan-Neofluar x100, 1.3 NA objective and appropriate optical filters (Omega Optical).

To investigate the colocalization of FcRIIA and FcRIIIB, PMNs were labeled with fluorescein-IV.3 and LC-biotin-3G8 Fab, which was then cross-linked with 15 µg/ml Texas Red streptavidin (Molecular Probes). After 7 min, the cells were centrifuged onto glass slides (Cytospin 2, Shandon Southern Instruments, Sewickley, PA), promptly fixed in -20°C methanol, air-dried, and mounted in glycerol with 1 mg/ml 1,4-phenylenediamine (Sigma-Aldrich, Milwaukee, WI) as antifading agent. Fluorescence photomicrographs were obtained with an Olympus BX60 microscope, a UPlanFI x100, 1.30 NA objective, and automatic exposure control (PM-30 Exposure Control Unit, Olympus Instruments, New Hyde Park, NY). Leakage of Texas Red fluorescence through the FITC filter and vice versa was negligible, as shown by singly labeled controls. Double-labeled cells were also imaged with a Leica confocal laser scanning microscope (Leica Microsystems, Exton, PA) equipped with a krypton-argon laser, a Plan Apo x100, 1.3 NA objective, and fluorescein and rhodamine filter sets. Pinholes were selected to obtain optical sections of 0.5-µm thickness.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
Frustrated phagocytosis: response of P388D₁ cells to ligation of transfected FcRIIA

Following reports about the T cell response upon ligation of membrane Ags to plastic surfaces (24), we set up a similar assay to study FcR signaling. The GaM-coated glass did not induce spreading of resting PMNs or macrophages (Fig. 1A). However, upon addition of IV.3 Fab, surface ligation of FcRIIA resulted in dramatic cell spreading (Fig. 1B). This response was maximal after 10 min and persisted for at least 4 h. Similar results were obtained when cells, plated on streptavidin-derivatized glass, were exposed to LC-biotin-IV.3 Fab (not shown).

View larger version (91K): [in this window] [in a new window]   FIGURE 1. Frustrated phagocytosis response of P388D1 cells transfected with wild-type or truncated FcRIIA. Cells plated on GaM-derivatized coverslips were stimulated by adding 1 µg/ml anti-FcRIIA IV.3 Fab. A, Wild-type FcRIIA, unstimulated (U); B, wt FcRIIA, stimulated (S); C, 264, U; D, 264, S; E, 233, U; F, 233, S; G, wt FcRIIA, loaded with Ca2+ chelator BAPTA, S; H, wt FcRIIA, preincubated with 10 µg/ml of the protein tyrosine kinase inhibitor genistein, S.

View larger version (44K): [in this window] [in a new window]   FIGURE 2. Histograms of the projected areas of P388D1cells expressing wt or truncated FcRIIA during frustrated phagocytosis. Transfected P388D1 cells were stimulated for 0, 10, and 60 min. Analysis was performed by image-based cytometry as described in Materials and Methods. Ligated wt FcRIIA evoked a rapid and sustained spreading response. No response was observed upon ligation of truncated FcRIIA mutants, 233 and 264.

Frustrated phagocytosis: response of human PMNs to ligation of FcRIIA or FcRIIIB

In PMNs, too, ligation of FcRIIA to GaM-conjugated glass via IV.3 Fab resulted in enhanced spreading relative to unstimulated controls (Fig. 3). As in PW16 cells, this response was blocked by BAPTA and genistein (data not shown). In contrast to FcRIIA, ligation of FcRIIIB to the GaM-derivatized surface by 3G8 Fab did not elicit any morphological change in PMNs (Fig. 3). This is the only instance in which we found a discrepancy between the two FcRs in a cellular response to a stimulus or an inhibitor. Its significance will be discussed later in the context of our model of FcRIIIB signal transduction.

View larger version (24K): [in this window] [in a new window]   FIGURE 3. Tethering of FcRIIA, but not FcRIIIB, stimulates frustrated phagocytosis in PMNs. Histogram analysis of cell spreading on GaM-coated coverslips, 15 min after stimulation by either anti-FcRIIA IV.3 or anti-FcRIIIB 3G8 Fab, compared with the unstimulated control.

Wortmannin blocks the FcRIIA-mediated spreading response in phagocytes

Incubation of PW16 macrophages or PMNs with 10 nM wortmannin inhibited cell spreading upon FcRIIA ligation to the derivatized glass, indicating that this response requires PI3K activity. The dependence of the spreading response on the concentration of inhibitor was quantified by image analysis (Fig. 4). We found that the wortmannin IC₅₀ is 2 nM in PMNs and 23 nM in PW16 cells. The value for PMNs is in excellent agreement with that reported for inhibition of purified PI3K (25). The higher IC₅₀ measured in macrophages may reflect an enhanced capacity of these cells to sequester or excrete wortmannin or a lower sensitivity of the murine PI3K to the inhibitor.

View larger version (18K): [in this window] [in a new window]   FIGURE 4. Dose-dependent inhibition of FcRIIA-mediated Ca2+release and frustrated phagocytosis by wortmannin. Data were acquired by indo-1 Ca2+ indicator fluorometry (, ) or by image-based cytometry of spread cells (•, ) in either PMNs (, •) or P388D1 cells transfected with wt FcRIIA (, ) and plotted as a percentage of the response measured in the absence of wortmannin. The efficacy of wortmannin was consistent for each cell type, with estimated IC50 of 2 and 23 nM in PMNs and macrophages, respectively.

Cross-linking of FcR triggers intracellular Ca²⁺ release

The FcRIIA-mediated phagocytosis of opsonized erythrocytes (26) and the spreading response triggered by frustrated phagocytosis are blocked in cells loaded with the Ca²⁺ chelator BAPTA. To determine whether wortmannin inhibits the phagocytic response by interfering with the intracellular Ca²⁺ mobilization, we measured [Ca²⁺]_i in PMNs loaded with the fluorescent indicator indo-1. Cross-linking of IV.3 Fab-bound FcRIIA or 3G8 Fab-bound FcRIIIB by GaM triggered rapid (<2 min) and large (up to 1 µM) [Ca²⁺]_i transients (Fig. 5A), approaching in magnitude those stimulated by 100 nM fMLP. Interestingly, we noticed a consistent lag (30 s) in Ca²⁺ release by FcRIIIB relative to that by FcRIIA (Fig. 5A), which may be relevant to the mechanism of FcRIIIB signal transduction.

View larger version (19K): [in this window] [in a new window]   FIGURE 5. [Ca2+]i transients stimulated by cross-linking FcRIIA or FcRIIIB are inhibited equally by wortmannin. A, Representative [Ca2+]i transients in PMNs stimulated by GaM cross-linking of 3G8 Fab-bound FcRIIIB (solid lines) or IV.3 Fab-bound FcRIIA (dashed lines). B, Percent inhibition of peak [Ca2+]i after GaM cross-linking of FcRIIIB () and FcRIIA () or fMLP stimulation () as a function of wortmannin concentration.

Endocytosis of cross-linked FcRs is unaffected by wortmannin

Although the internalization of FcRIIA bound to immune complexes is well documented, the fate of similarly cross-linked FcRIIIB has not been characterized. Based on the amounts of ¹²⁵I-conjugated RaG bound to cross-linking GaM left on the surface of PMNs, we conclude that both Fab-labeled FcRs are sequestered from the cell surface within 3 min after cross-linking by GaM (30 µg/ml) at 37°C (Fig. 6). Internalization of both FcRs was inhibited only slightly or not at all by wortmannin concentrations as high as 100 nM, which completely blocked Ca²⁺ release and frustrated phagocytosis. Similarly, in PW16 cells incubated with 1 µM wortmannin there was no significant decrease in the rapid (<2 min) endocytosis of cross-linked IV.3 Fab-labeled FcRIIA (data not shown).

View larger version (14K): [in this window] [in a new window]   FIGURE 6. Internalization of Ab-cross-linked FcRIIIB or FcRIIA in human PMNs is unaffected by PI3K inhibition. The kinetics of sequestration of GaM-cross-linked IV.3 Fab-bound FcRIIA (A) or 3G8 Fab-bound FcRIIIB (B) in the presence (, ) or the absence (, •) of 100 nM wortmannin was measured by RIA, using 125I-conjugated F(ab')2 rabbit anti-goat IgG as described in Materials and Methods.

N,N-Dimethylsphingosine blocks FcR-mediated Ca²⁺ release

To determine whether SK and its product S1P participate in FcR signaling, we tested the effect of the SK inhibitor DL-threo-dihydrosphingosine, which blocked the FcRI-induced Ca²⁺ flux in rat mast cells (15), on the Ca²⁺ response of PMNs to FcR cross-linking. However, since we found this reagent to be poorly soluble and difficult to administer to live cells, we opted to use DMS, a more soluble and potent inhibitor of SK (28, 29). Incubation of PMNs with DMS suppressed the [Ca²⁺]_i transients triggered by specific Ab cross-linking of each FcR, without significantly affecting the response to fMLP (Fig. 7, A and B). From a rough dose dependence of this effect, we derived an IC₅₀ of about 0.5 nmol DMS/10⁶ cells (Fig. 7C). We use these units because the level of inhibition depended on the concentration of DMS and inversely on the density of PMNs. We believe that membrane partitioning of DMS results in its effective concentration being inversely proportional to the total membrane area in the cell suspension. For comparison to published values, because the cell density was usually about 10⁶ cells/ml, our IC₅₀ corresponds to 0.5 µM DMS.

View larger version (20K): [in this window] [in a new window]   FIGURE 7. DMS inhibits the [Ca2+]i rise triggered by cross-linking FcRIIA or FcRIIIB. [Ca2+]i transients measured fluorometrically from indo-1-loaded human PMNs upon GaM cross-linking of Fab-bound FcRIIA (A) or FcRIIIB (B) in the presence (dashed lines) or the absence (solid lines) of DMS (1 nmol/106 cells). C, DMS dose-dependent inhibition of Ca2+ release triggered by cross-linking FcRIIA (•) or FcRIIIB ().

Time course of FcR capping

The surface distribution of FcRs was examined on live PMNs using fluorescent Fab. FcRIIIB, labeled with fluorescein-3G8 Fab, appeared uniformly distributed on the surface. Upon addition of GaM, however, the receptors aggregated immediately into patches, which gradually grew in size and often coalesced into caps within 7–10 min at room temperature (Figs. 8, a–d). FcRIIA displayed a similar behavior (data not shown). Although capping results from events occurring downstream of FcR activation and Ca²⁺ mobilization, the kinetics of the initial clustering is consistent with this process initiating the signaling cascade in PMNs.

View larger version (167K): [in this window] [in a new window]   FIGURE 8. Aggregation of FcRIIIB upon cross-linking. Image sequence shows formation of fluorescein-3G8 Fab-labeled FcRIIIB caps in live PMNs at room temperature. Images were acquired at 0 (a), 1.5 (b), 4.5 (c), and 7.5 (d) min after addition of GaM (30 µg/ml). Similar results were obtained upon cross-linking fluorescein-IV.3 Fab-labeled FcRIIA (not shown).

Wortmannin disrupts FcR capping

In addition to blocking the [Ca²⁺]_i transients, wortmannin profoundly disrupts the FcR aggregation in PMNs. Confocal microscope images of PMNs labeled with rhodamine-3G8 or fluorescein-IV.3 Fab, fixed 7 min after cross-linking with GaM, are shown in Fig. 9, a and b. In PMNs incubated with 30 nM wortmannin, rather than coalescing into large patches, the clusters of FcRIIA and FcRIIIB remain dispersed over the entire cell (Fig. 9, c and d). Thus, the PI3K activity is necessary for the large scale redistribution of cross-linked FcRs.

View larger version (183K): [in this window] [in a new window]   FIGURE 9. Wortmannin disrupts capping of cross-linked FcRs. Confocal fluorescence images of PMNs labeled with rhodamine-3G8 (a and c) or fluorescein-IV.3 (b and d) Fab, and fixed 7 min after cross-linking with 30 µg/ml GaM. FcR clusters remain dispersed in PMNs preincubated with 100 nM wortmannin (c and d), but form caps in its absence (a and b).

Colocalization of FcRIIA with cross-linked FcRIIIB

The results presented thus far indicate early convergence of the signaling pathways used by the two FcR isoforms in PMNs. We propose that FcRIIIB clustering triggers FcRIIA activation. However, this process requires physical interaction between FcRIIIB and FcRIIA. Indeed, in doubly labeled PMNs, we find that FcRIIA comigrates with FcRIIIB when the latter is cross-linked. Because both 3G8 and IV.3 are murine mAbs, GaM could not serve as selective cross-linker. Instead, we labeled PMNs with LC-biotin-3G8 Fab and fluorescein-IV.3 Fab and specifically cross-linked the LC-biotin-Fab with Texas Red-streptavidin. Fluorescence photomicrographs (Figs. 10, a–c) and confocal images (Figs. 10, d–e) demonstrate extensive colocalization of Texas Red (FcRIIIB) and fluorescein (FcRIIA). The two photomicrographs were digitized and superimposed using Photoshop (Adobe Systems, San Jose CA; Fig. 10b). Confocal images were also acquired under conditions of heterotypic cross-linking, in which both FcRs were engaged by adding GaM (30 µg/ml) to PMNs labeled with rhodamine-3G8 Fab and fluorescein-IV.3 Fab. As expected, the distributions of the two FcRs are perfectly correlated in these images (Figs. 10, f–g).

View larger version (121K): [in this window] [in a new window]   FIGURE 10. The FcRIIA colocalizes with cross-linked FcRIIIB. a—c, Fluorescence photomicrographs of PMNs labeled with both fluorescein-IV.3 and LC-biotin-3G8 Fab after addition of Texas Red-streptavidin (15 µg/ml) to cross-link FcRIIIB. Cells were cytospun onto coverslips and fixed about 7 min after cross-linking, as described in Materials and Methods. a, Texas Red image; c, fluorescein image; b, merged image obtained by superposition of scanned images a and c. Confocal images of similarly prepared cells: d, Texas Red; and e, fluorescein channel. Confocal images of PMNs labeled with rhodamine-3G8 and fluorescein-IV.3 Fab after heterotypic cross-linking by GaM (30 µg/ml): f, rhodamine; and g, fluorescein channel.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
Given their similar binding properties, it is likely that both FcR isoforms on PMNs interact with immune complexes under physiological conditions. However, to isolate the signaling function of each receptor, we have used isoform-specific mAbs to engage selectively FcRIIA or FcRIIIB. The major findings of this study are as follows. First, without binding of FcRIIA to ligand or cytokine priming of PMNs, cross-linked FcRIIIB delivers an intracellular signal that leads to a robust Ca²⁺ response and receptor capping. Second, the [Ca²⁺]_i transients triggered by cross-linked FcRIIIB are inhibited by wortmannin and DMS with identical efficacies as those stimulated by FcRIIA. Third, the physical disposition of cross-linked FcRIIIB, determined by immunofluorescence microscopy and internalization assay, closely parallels that of cross-linked FcRIIA under all conditions tested. Fourth, cross-linking of FcRIIIB induces the redistribution and colocalization of FcRIIA with aggregated FcRIIIB, as measured at the resolution afforded by optical microscopy. The only divergence between the two FcR phenotypes was the failure of PMNs to respond by spreading when FcRIIIB was ligated to substrate-immobilized Ab.

Based on these findings, we propose that FcRIIIB signaling is mediated by FcRIIA, which copartitions into membrane domains induced by cross-linked FcRIIIB and is thus brought into clusters for ITAM phosphorylation and activation.

Signaling by Ab-cross-linked FcRIIIB

Our results, obtained by GaM cross-linking of 3G8 Fab-bound FcRIIIB, confirm that GPI-anchored proteins are capable of intracellular signaling, as shown in T cells, and that FcRIIIB aggregation can trigger [Ca²⁺]_i transients in PMNs (31). Ca²⁺ signaling by FcRIIIB was reported to require priming of PMNs with TNF or GM-CSF (10). However, we find no such need for fully functional GPI-anchored receptors.

The dependence of signaling on large scale aggregation appears to be more critical for FcRIIIB than for FcRIIA. Again, we stress that the failure of 3G8 Fab to stimulate frustrated phagocytosis in PMNs was the only instance in which the signaling phenotypes of the two FcRs differed. This seemingly anomalous result corroborates a report that while T cells were stimulated by ligation of the TCR/CD3 complex to anti-CD3 IgG-coated plastic, similar ligation of the GPI-anchored proteins TAP and Thy-1 did not elicit a response (24). We believe that this observation illustrates a key difference in the physical requirements for signal transduction by the GPI-linked FcRIIIB and the transmembrane FcRIIA. Phosphorylation of the FcRIIA ITAM is the critical common event, and, indeed, FcRIIA is phosphorylated upon FcRIIIB cross-linking (8). Formation of small clusters of FcRIIA upon cross-linking with IV.3 mAb is sufficient to generate a full PMN response. However, initiation of FcRIIA ITAM phosphorylation by FcRIIIB cross-linking requires the formation of protein/lipid domains large enough to recruit and bring into proximity an adequate number of FcRIIA molecules. This requirement accounts for both the lag of the Ca²⁺ response to FcRIIIB cross-linking (Fig. 5A) and the inability of individually tethered and immobilized GPI-anchored proteins to trigger frustrated phagocytosis. It also explains why bivalent ligands and low valency immune complexes are ineffective at activating PMNs via FcRIIIB (10, 32).

Role of PI3K in FcR signaling

The participation of PI3K in FcR signaling is well documented (33) and was indeed used to show the efficacy of the inhibitor wortmannin (4). We have used wortmannin to demonstrate the congruence of signaling by the two FcRs. The inhibition of FcRIIIB-mediated Ca²⁺ release by wortmannin implicates PI3K as a critical element for GPI-linked receptor signaling. However, because this cytosolic enzyme and FcRIIIB cannot interact directly, their coupling requires a transmembrane molecule. We believe that FcRIIA is likely to fulfill this function, not only because it colocalizes with clustered FcRIIIB, but also because wortmannin blocks signaling by both FcRs with identical efficacy.

The molecular mechanism for recruitment of PI3K to the FcR signaling cascade is not clear. The SH2 domains of p85, the regulatory subunit of PI3K, have been shown to recognize tyrosine-phosphorylated YXXM sequences (34, 35). Because FcRIIA lacks this motif, it may interact with PI3K via an adapter molecule. In platelets, the tyrosine kinase Syk was shown to associate with phosphorylated FcRIIA and was proposed to recruit PI3K to the activated receptors (36). However, direct interaction between p85 and FcRIIA cannot be ruled out, because p85 was found to bind to doubly phosphorylated ITAMs of the - and -chains of CD3 (37, 38). The PI3K from PMNs was also found to associate in vitro with a fusion protein consisting of GST and the cytosolic domain of FcRIIA (39).

In agreement with our observation that wortmannin is unable to block the endocytosis of cross-linked FcRs, PI3K inhibition does not block the internalization of PDGF receptors, but it interferes with their trafficking to lysosomal compartments (40). These findings suggest that the lipid products of PI3K may play a critical role in FcR signaling. Polyphosphorylated and, in particular, 3-hydroxyphosphorylated inositol lipids participate in regulating endocytic transport and membrane trafficking (41). Phosphatidylinositol 3,4,5-trisphosphate (PI(3, 4, 5)P₃) activates phospholipase C-1 by binding to its pleckstrin homology or its SH2 domain (42, 43). The PI(3, 4, 5)P₃ also functions in the activation of c-akt by binding to its pleckstrin homology domain, which is required for phosphorylation of c-akt, and by directly stimulating a specific kinase (44). Whether PI3K interacts with phosphorylated FcRIIA directly or via an intermediate, our results indicate that it has an identical functional role in both FcRIIA and FcRIIIB signaling.

Inhibition of [Ca²⁺]_i transients by DMS

Following a report of a much weaker IP₃ release after FcR cross-linking than after fMLP stimulation (11), later investigations have yielded conflicting results (10, 12, 13). The amount of IP₃ was recently confirmed to be much smaller in PMNs activated by FcRIIA aggregation than by fMLP stimulation and almost negligible after FcRIIIB cross-linking (14). Meanwhile, S1P has been found to mediate intracellular Ca²⁺ release in 3T3 fibroblasts (45) and the autocrine stimulation of platelets (28). In permeabilized fibroblasts, the Ca²⁺ response to S1P was not blocked by heparin, an IP₃ antagonist (46). In rat mast cells, inhibition of SK abolished the Ca²⁺ release following FcRI stimulation while leaving the IP₃ pathway intact (15). However, in U937 cells, FcRI mobilized Ca²⁺ by activating PC-PLD and SK, whereas FcRIIA triggered a substantial IP₃ production (30).

We have shown that DMS, a competitive inhibitor of SK, blocks Ca²⁺ release in PMNs stimulated by FcR cross-linking, but not by fMLP. In contrast to our results, DMS was recently reported to inhibit the PMN Ca²⁺ response to fMLP with an IC₅₀ of 5 µM (47). However we found the FcR response to be 10-fold more sensitive to inhibition, and interpreted any effect on the fMLP-triggered Ca²⁺ release at DMS concentrations >2 µM as nonspecific toxicity. The SK activation pathway in PMNs remains undefined, because the inhibition studies with the butanol isomers failed to confirm the participation of PC-PLD demonstrated in the case of FcRI (30). It also seems likely that FcRIIA uses distinct signaling pathways in different cell types, because IP₃ release is substantial in U937 cells (30) but only minimal in PMNs (14). However, as for wortmannin, the similar efficacy with which DMS inhibits PMN activation by both FcRs indicates that they share a common signaling pathway.

Because S1P and IP₃ mobilize Ca²⁺ from the same thapsigargin-sensitive stores (11, 46), differences in Ca²⁺-dependent PMN activation via Fc and fMLP receptors may arise from a different cellular compartmentalization of their respective second messengers. As opposed to the water-soluble IP₃ molecule, S1P is probably mostly membrane bound. Thus, SK activation by FcRs may generate a more localized response, suitable for mediating phagocytosis, than that produced by IP₃ release, which may mediate whole cell responses, such as chemotaxis.

Functional dependence of FcRIIIB on signaling by FcRIIA

Because the pattern of FcRIIA aggregation induced by FcRIIIB cross-linking is identical with that induced by direct FcRIIA cross-linking, we propose that a major component of the signal generated by FcRIIIB is transduced by FcRIIA. In our model, FcRIIIB relies on the ITAM of FcRIIA for signal transduction and, thus, disruption of this motif should impair the signaling capacity of both FcR isoforms. A suitable biological system to replicate the PMN FcR signaling machinery is not at hand. Human PMNs are short lived and not amenable to conventional molecular biological approaches, whereas FcRIIIB transfected into heterologous systems is often expressed in a nonfunctional or transmembrane form. Nevertheless, Green et al. expressed both FcRs in Jurkat T cells and showed that, as in PMNs (9), co-cross-linking of the two receptors elicited a synergistic enhancement of the [Ca²⁺]_i transient relative to that triggered by cross-linking of FcRIIA alone (48). This effect required expression of the ITAM of FcRIIA and of the GPI anchor of FcRIIIB. However, it is unclear whether this cell line is a reliable model of human PMNs, because the [Ca²⁺]_i transients triggered by cross-linking FcRIIIB or the more abundant endogenous CD59 were weak, slow rising, and completely abolished by chelation of extracellular Ca²⁺. Moreover, the TCR complex expressed in this cell line may contribute to signaling by the exogenous FcRIIIB. In T cells, the ITAM-bearing TCR mediates signaling by cross-linked GPI-anchored proteins (16, 49). Further work in this area is required.

Physical models of FcRIIIB function

In formulating a model for FcRIIIB signaling we note that in PMNs, as in T cells, Ca²⁺ is released upon cross-linking of various GPI-anchored proteins (50). The signaling capacity of these diverse proteins may derive from their common structural element, the GPI anchor, rather than from specific protein-protein interactions. Focusing on the glycan portion of GPI, Petty and colleagues found that FcRIIIB-triggered [Ca²⁺]_i transients and superoxide production were inhibited by high concentrations of D-mannose or N-acetyl-D-glucosamine, each part of the conserved core structure of GPI anchors (51). Moreover, N-acetyl-D-glucosamine disrupted cocapping of FcRIIIB with CD11b/CD18, a ß₂ integrin also known as complement receptor 3 (18). Because complement receptor 3 contains a lectin-like site that could recognize GPI, they proposed that FcRIIIB signaling results from binding to the ß₂ integrin (52). Indeed, integrins are known to modulate FcR-mediated PMN activity (53). FcRIIIB may also interact specifically with the formyl peptide receptor, because soluble immune complexes or 3G8 Fab block fMLP-induced chemotaxis, but not the response of PMNs to other chemotactic stimuli (54). However, based on its physical proximity, signaling capacity, and the evidence of functional interactions between GPI-anchored proteins and ITAM-bearing immunoreceptors, we propose that FcRIIA is the primary signal transducer for FcRIIIB in PMNs.

Aggregated FcRIIIB may recruit FcRIIA via interactions within the membrane hydrophobic core. GPI-anchored proteins are selectively enriched, together with sphingolipids, glycolipids, and cholesterol, in detergent-insoluble membrane complexes isolated from cold Triton X-100 cell lysates (55, 56). Doubly acylated Src family tyrosine kinases are also found in these complexes (57). First identified with caveolae, these complexes have since been isolated from cells lacking these structures (58). These and other similar findings led Simons and Ikonen to propose that these lipid-protein complexes form microdomains or functional "rafts" that may participate in trafficking and sorting of membrane components, and function as integrated signaling assemblies (19). Evidence of clusters of GPI-anchored proteins, obtained by fluorescence microscopic measurements of resonance energy transfer and by chemical cross-linking (59, 60), supports the existence of rafts in vivo. However, the long term stability of these domains is not absolutely required by the model. Indeed, rafts may be transient dynamic entities stabilized by cooperative interactions formed upon aggregation of GPI-anchored proteins (61). Thus, random dispersion of FcRIIIB on the surface of resting PMNs is compatible with raft formation after cross-linking.

Two fundamental questions remain outstanding. The first concerns the identity of the molecular interactions leading to formation of rafts and inclusion of lipid-linked proteins. The second, relevant to our model of FcRIIIB signaling, concerns the mechanism for recruitment of transmembrane proteins, such as FcRIIA, to the rafts.

With regard to the first issue, the composition and physical properties of the plasma membrane are critical for microdomain stability. Depletion of cellular cholesterol destabilizes the detergent-insoluble complexes and inhibits signal transduction by GPI-anchored proteins in T cells (62). Incubation of T cells with polyunsaturated fatty acids causes inhibition of the Ca²⁺ response to stimulation via CD3 and CD59 and displacement of the Src family tyrosine kinase Lck from the detergent-insoluble complexes (63). Mixtures of cholesterol and sphingomyelin or dipalmitoylphosphatidylcholine form membranes that enhance the detergent insolubility of GPI-anchored proteins (64). The current hypothesis for the physical basis of these phenomena invokes the formation of cholesterol and sphingolipid-rich domains where the bilayer exists in the liquid-ordered (l_o) phase (65, 66). In this phase, the lipid chains are highly ordered, as in the gel phase, yet the lipids remain free to diffuse. These unusual properties are thought to facilitate partitioning of proteins with suitable lipid anchors. This idea was tested in vitro using placental alkaline phosphatase (67), whose GPI anchor contains saturated palmitic and stearic acyl chains (68). However, the GPI-anchored acetylcholinesterase and CD59 from human RBCs are also detergent insoluble (69, 70) despite containing highly unsaturated 2-acyl chains (71, 72). Thus, unsaturated lipid anchors appear to be compatible with incorporation into rafts. Although the structure of the GPI anchor of FcRIIIB is unknown, it is unlikely to exhibit drastically different properties. Indeed, preliminary experiments indicate that FcRIIIB, but not FcRIIA, in resting PMNs segregates into detergent-insoluble membrane fractions (our unpublished observations).

Regarding the second issue, we admit that evidence for inclusion of transmembrane proteins in rafts is still weak. However, the following observations indicate that single-pass transmembrane proteins may indeed associate with rafts. Influenza virus hemagglutinin (HA) and CD44, both type I proteins like FcRIIA, and influenza virus neuraminidase, a type II protein, partition in detergent-insoluble complexes (73, 74, 75). Glycophorin, a type I glycoprotein abundant in RBCs, promotes formation of cholesterol-rich domains in model membranes (76).

Two properties of transmembrane domains have been suggested to determine the affinity of membrane proteins for rafts: their length and the presence of specific amino acids. The first property was invoked in the mechanism of protein sorting within the Golgi apparatus. The transmembrane domains of type II proteins retained in this organelle were found to be shorter (15 aa), on the average, than those of analogous proteins delivered to the plasma membrane (20 aa). This effect was attributed to the thicker hydrophobic core of trans-Golgi and plasma membranes, a result of their enrichment in cholesterol and saturated sphingolipids (77). For thermodynamic reasons, proteins tend to partition where the hydrophobic thickness of the bilayer matches the length of their hydrophobic transmembrane domain (78). This mechanism of protein sorting has received experimental and theoretical support (79, 80).

A role for the transmembrane domain sequence in determining the affinity of proteins for rafts has been inferred from mutagenesis studies on HA and neuraminidase (73, 75). Although no targeting motif was identified, the work on HA unveiled a requirement for isoleucine and leucine residues to be in contact with the outer membrane leaflet. A preference for large hydrophobic residues in the N-terminal half of the transmembrane helix of type I proteins had been previously noted (81).

Although these requirements are rather loose, the FcRIIA transmembrane domain satisfies both of them. It is predicted to be 24 aa long, exceeding the value cited by Bretscher and Munro, and its N-terminal sequence is rich in ß-branched hydrophobic residues (I²¹⁸IVAVVI). Moreover, a cysteine, Cys²⁴¹, resides at the cytoplasmic membrane-water interface predicted by the stop transfer sequence of basic residues, R²⁴²KKR (82). Cysteines at this location are targets for palmitoylation (83). Although it is unknown whether such modification occurs in vivo, acylation could increase the affinity of FcRIIA for rafts.

Although further investigation is necessary to validate our model, our data provide new insight into the signaling mechanism of FcRIIIB and the functional architecture of the PMN plasma membrane. Although activation of many receptors relies on intramolecular conformational changes or oligomerization, the signaling capacity of GPI-anchored proteins may derive from their ability to induce the formation of microdomains of defined composition within the plasma membrane. Thus, aggregation of FcRIIIB may lead to formation of supramolecular assemblies comprising all the components, including FcRIIA, necessary for generating cellular responses.

	Acknowledgments

 
We thank Heikki Väänänen for expert assistance with microscopy and Josef Eisinger for valuable discussions.

	Footnotes

 
¹ This work was supported by National Institutes of Health Grant AI24322. Confocal laser scanning microscopy was performed at the Mount Sinai School of Medicine-Confocal Laser Scanning Microscopy core facility, supported with funding from National Institutes of Health Grant 1S10RR09145-01 and National Science Foundation Grant DBI-9724504.

² Address correspondence and reprint requests to Dr. Jay Unkeless, Immunobiology Center, Box 1630, Mount Sinai School of Medicine, 1425 Madison Avenue, New York, NY 10029. E-mail address:

³ Address correspondence and reprint requests to Dr. Massimo Sassaroli, Department of Physiology and Biophysics, Box 1218, Mount Sinai School of Medicine, One Gustave L. Levy Place, New York, NY 10029. E-mail address:

⁴ Abbreviations used in this paper: PMN, polymorphonuclear leukocyte or neutrophil; [Ca²⁺]_i, intracellular free calcium concentration; DMS, N,N-dimethylsphingosine; GaM, F(ab')₂ of goat anti-mouse IgG; HA, hemagglutinin; IP₃, inositol 1,4,5-trisphosphate; ITAM, immunoreceptor tyrosine-based activation motif; LC-biotin, long chain biotin or sulfosuccinimidyl-6-(biotinamido) hexanoate; PC-PLD, phosphatidylcholine-specific phospholipase D; PI3K, phosphatidylinositol 3-kinase; PI(3,4,5)P₃, phosphatidylinositol 3,4,5-trisphosphate; RaG, F(ab')₂ of rabbit anti-goat IgG; SH2, Src homology 2; SK, sphingosine kinase; SP1, sphingosine 1-phosphate; wt, wild type; BAPTA, bis(2-aminophenoxy)ethane-N,N,N',N'-tetraacetate; NA, normal aperture.

Received for publication July 30, 1999. Accepted for publication October 12, 1999.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Keegan, A. D., W. E. Paul. 1992. Multichain immune recognition receptors: similarities in structure and signaling pathways. Immunol. Today 13:63.[Medline]
2. Reth, M.. 1989. Antigen receptor tail clue. Nature 338:383.[Medline]
3. Isakov, N.. 1997. Immunoreceptor tyrosine-based activation motif (ITAM), a unique module linking antigen and Fc receptors to their signaling cascades. J. Leukocyte Biol. 61:6.[Abstract]
4. Ninomiya, N., K. Hazeki, Y. Fukui, T. Seya, T. Okada, O. Hazeki, M. Ui. 1994. Involvement of phosphatidylinositol 3-kinase in Fc receptor signaling. J. Biol. Chem. 269:22732.[Abstract/Free Full Text]
5. Fleit, H. B., S. D. Wright, J. C. Unkeless. 1982. Human neutrophil Fc receptor distribution and structure. Proc. Natl. Acad. Sci. USA 79:3275.[Abstract/Free Full Text]
6. Boros, P., J. A. Odin, T. Muryoi, S. K. Masur, C. Bona, J. C. Unkeless. 1991. IgM anti-FcR autoantibodies trigger neutrophil degranulation. J. Exp. Med. 173:1473.[Abstract/Free Full Text]
7. Hundt, M., R. E. Schmidt. 1992. The glycosylphosphatidylinositol-linked Fc receptor III represents the dominant receptor structure for immune complex activation of neutrophils. Eur. J. Immunol. 22:811.[Medline]
8. Edberg, J. C., R. P. Kimberly. 1994. Modulation of Fc and complement receptor function by the glycosyl-phosphatidylinositol-anchored form of FcRIII. J. Immunol. 152:5826.[Abstract]
9. Vossebeld, P. J., J. Kessler, A. E. von dem Borne, D. Roos, A. J. Verhoeven. 1995. Heterotypic FcR clusters evoke a synergistic Ca²⁺ response in human neutrophils. J. Biol. Chem. 270:10671.[Abstract/Free Full Text]
10. Watson, F., L. Gasmi, S. W. Edwards. 1997. Stimulation of intracellular Ca²⁺ levels in human neutrophils by soluble immune complexes: functional activation of FcRIIIb during priming. J. Biol. Chem. 272:17944.[Abstract/Free Full Text]
11. Rosales, C., E. J. Brown. 1992. Signal transduction by neutrophil immunoglobulin-G Fc receptors: dissociation of intracytoplasmic calcium concentration rise from inositol 1,4,5-trisphosphate. J. Biol. Chem. 267:5265.[Abstract/Free Full Text]
12. Della Bianca, V., M. Grzeskowiak, S. Dusi, F. Rossi. 1993. Formation of inositol (1,4,5) trisphosphate and increase of cytosolic Ca²⁺ mediated by Fc receptors in human neutrophils. Biochem. Biophys. Res. Commun. 196:1233.[Medline]
13. Dusi, S., M. Donini, V. Della Bianca, F. Rossi. 1994. Tyrosine phosphorylation of phospholipase C-2 is involved in the activation of phosphoinositide hydrolysis by Fc receptors in human neutrophils. Biochem. Biophys. Res. Commun. 201:1100.[Medline]
14. Edberg, J. C., J. J. Moon, D. J. Chang, R. P. Kimberly. 1998. Differential regulation of human neutrophil FcRIIa (CD32) and FcRIIIb (CD16)-induced Ca²⁺ transients. J. Biol. Chem. 273:8071.[Abstract/Free Full Text]
15. Choi, O. H., J. H. Kim, J. P. Kinet. 1996. Calcium mobilization via sphingosine kinase in signalling by the FcRI antigen receptor. Nature 380:634.[Medline]
16. Rock, K. L., H. Reiser, A. Bamezai, J. McGrew, B. Benacerraf. 1989. The LY-6 locus: a multigene family encoding phosphatidylinositol-anchored membrane proteins concerned with T-cell activation. Immunol. Rev. 111:195.[Medline]
17. Presky, D. H., M. G. Low, E. M. Shevach. 1990. Role of phosphatidylinositol-anchored proteins in T cell activation. J. Immunol. 144:860.[Abstract]
18. Zhou, M., R. F. d. Todd, J. G. van de Winkel, H. R. Petty. 1993. Cocapping of the leukoadhesin molecules complement receptor type 3 and lymphocyte function-associated antigen-1 with Fc receptor III on human neutrophils: possible role of lectin-like interactions. J. Immunol. 150:3030.[Abstract]
19. Simons, K., E. Ikonen. 1997. Functional rafts in cell membranes. Nature 387:569.[Medline]
20. Odin, J. A., J. C. Edberg, C. J. Painter, R. P. Kimberly, J. C. Unkeless. 1991. Regulation of phagocytosis and [Ca²⁺]_i flux by distinct regions of an Fc receptor. Science 254:1785.[Abstract/Free Full Text]
21. Coleman, D. T., X. Chen, M. Sassaroli, C. Bancroft. 1996. Pituitary adenylate cyclase-activating polypeptide regulates prolactin promoter activity via a protein kinase A-mediated pathway that is independent of the transcriptional pathway employed by thyrotropin- releasing hormone. Endocrinology 137:1276.[Abstract]
22. Werb, Z., P. M. Tremble, O. Behrendtsen, E. Crowley, C. H. Damsky. 1989. Signal transduction through the fibronectin receptor induces collagenase and stromelysin gene expression. J. Cell Biol. 109:877.[Abstract/Free Full Text]
23. Sassaroli, M., R. da Costa, H. Vaananen, J. Eisinger, M. B. Poh-Fitzpatrick. 1992. Distribution of erythrocyte free porphyrin content in erythropoietic protoporphyria. J. Lab. Clin. Med. 120:614.[Medline]
24. Bamezai, A., V. Goldmacher, H. Reiser, K. L. Rock. 1989. Internalization of phosphatidylinositol-anchored lymphocyte proteins. I. Documentation and potential significance for T cell stimulation. J. Immunol. 143:3107.[Abstract]
25. Okada, T., Y. Kawano, T. Sakakibara, O. Hazeki, M. Ui. 1994. Essential role of phosphatidylinositol 3-kinase in insulin-induced glucose transport and antilipolysis in rat adipocytes: studies with a selective inhibitor wortmannin. J. Biol. Chem. 269:3568.[Abstract/Free Full Text]
26. Edberg, J. C., C. T. Lin, D. Lau, J. C. Unkeless, R. P. Kimberly. 1995. The Ca²⁺ dependence of human Fc receptor-initiated phagocytosis. J. Biol. Chem. 270:22301.[Abstract/Free Full Text]
27. Krause, K. H., W. Schlegel, C. B. Wollheim, T. Andersson, F. A. Waldvogel, P. D. Lew. 1985. Chemotactic peptide activation of human neutrophils and HL-60 cells: pertussis toxin reveals correlation between inositol trisphosphate generation, calcium ion transients, and cellular activation. J. Clin. Invest. 76:1348.
28. Yatomi, Y., F. Ruan, T. Megidish, T. Toyokuni, S. Hakomori, Y. Igarashi. 1996. N,N-dimethylsphingosine inhibition of sphingosine kinase and sphingosine 1-phosphate activity in human p