Structural Requirements of SLP-76 in Signaling via the High-Affinity Immunoglobulin E Receptor (Fc{varepsilon}RI) in Mast Cells

The adapter SLP-76 plays an essential role in Fc ${varepsilon}$ RI signaling,since SLP-76^-/- bone marrow-derived mast cells (BMMC) fail todegranulate and release interleukin-6 (IL-6) following Fc ${varepsilon}$ RIligation. To define the role of SLP-76 domains and motifs inFc ${varepsilon}$ RI signaling, SLP-76^-/- BMMC were retrovirally transducedwith SLP-76 and SLP-76 mutants. The SLP-76 N-terminal and Gadsbinding domains, but not the SH2 domain, were critical for Fc ${varepsilon}$ RI-mediateddegranulation and IL-6 secretion, whereas all three domainsare essential for T-cell proliferation following T-cell receptor(TCR) ligation. Unexpectedly, the three tyrosine residues inSLP-76 critical for TCR signaling, Y112, Y128, and Y145, werenot essential for IL-6 secretion, but were required for degranulationand mitogen-activated protein kinase activation. Furthermore,a Y112/128F SLP-76 mutant, but not a Y145F mutant, stronglyreconstituted mast cell degranulation, suggesting a criticalrole for Y145 in Fc ${varepsilon}$ RI-mediated exocytosis. These results pointto important differences in the function of SLP-76 between Tcells and mast cells.

INTRODUCTION

Top

Introduction

The high-affinity receptor for immunoglobulin E (IgE) (Fc ${varepsilon}$ RI)is a multimolecular complex of the IgE-binding ${alpha}$ subunit, twosignal-transducing ${gamma}$ subunits, and a ß subunit thatpromotes assembly of the receptor and amplifies signal transduction(3, 32). Both ${gamma}$ and ß chains contain immunoreceptortyrosine-based activation motifs (ITAMs) within their intracellulardomains. Upon Fc ${varepsilon}$ RI cross-linking, the ITAMs of the ${gamma}$ and ßsubunits become phosphorylated by the Src family tyrosine kinaselyn and recruit the protein tyrosine kinase Syk, which in turnphosphorylates intracellular proteins such as LAT, phospholipaseC- ${gamma}$ (PLC- ${gamma}$ ), Vav, and the adapter protein SLP-76 (9, 21, 28, 35).

SLP-76 is predominantly expressed in hematopoietic cells andhas three major protein-interacting domains (7, 25, 38, 46).Three tyrosine residues (Y113, Y128, and Y145) in the N-terminaldomain become phosphorylated by Syk family protein tyrosinekinases following T-cell receptor (TCR) engagement and providebinding sites for the SH2 domains of Vav, Nck, and Itk. Thebinding of Vav and Nck to phosphotyrosine residues Y113 andY128 may link SLP-76 to the JNK (Jun amino-terminal kinase)pathway and to the actin cytoskeleton (5, 10, 54-56). Y145 hasbeen implicated in the binding of SLP-76 to Itk (6, 53). Directinteraction of PLC- ${gamma}$ with SLP-76 as well as formation of a complexinvolving LAT and Itk, which, respectively, bind and phosphorylatePLC- ${gamma}$ , may be required for PLC- ${gamma}$ activation (49, 57, 59). SLP-76associates constitutively via its central proline-rich domainwith the SH3 domain of Gads, which recruits it to LAT followingTCR stimulation (1, 31, 33). This allows the translocation ofSLP-76 to glycolipid-enriched microdomains (GEMs) (24) and mayalso link it via Sos to the Ras/mitogen-activated protein kinase(MAPK)/extracellular signal-regulated protein kinase (ERK) pathway(29, 36). Proteins that directly interact with the SLP-76 SH2domain include ADAP (formerly known as SLAP-130/FYB), the Ser/Thrkinase HPK1, and a 62-kDa phosphoprotein (11, 36, 37, 48).

SLP-76^-/- mice lack T cells, indicating that signals integratedby SLP-76 are critical for T-cell development (8, 43). SLP-76also plays an important role in TCR signal transduction andT-cell activation. SLP-76-deficient Jurkat cells exhibit severelyimpaired signaling after stimulation through the TCR-CD3 complex.PLC- ${gamma}$ 1 activation, calcium mobilization, ERK1/2 phosphorylation,and interleukin-2 (IL-2) production are all severely compromised(59).

SLP-76-deficient mice have normal numbers of mast cells in theirskin and bronchi, and their bone marrow cells differentiatenormally in vitro into mast cells upon culture in IL-3-containingmedium (44). However, SLP-76^-/- bone marrow-derived mast cells(BMMC) fail to release the granular enzyme ß-hexosaminidaseand to secrete IL-6 after Fc ${varepsilon}$ RI cross-linking. These findingsindicate that SLP-76 plays an essential role in Fc ${varepsilon}$ RI signaling.We took advantage of the availability of SLP-76^-/- BMMC andtransduced them retrovirally with SLP-76 mutants to addressthe role of SLP-76 domains and residues for its adapter functionin signaling via Fc ${varepsilon}$ RI.

MATERIALS AND METHODS

Top

Materials and Methods

Cells and cell culture. Bone marrow cells were cultured in WEHI-3-conditioned medium(WCM) as a source of IL-3 (44). After 3 to 5 weeks of culture,90% or more of the cells derived from wild-type (WT) and SLP-76^-/-bone marrow are mast cells, as evidenced by fluorescence-activatedcell sorting (FACS) analysis for Fc ${varepsilon}$ RI expression. To assessFc ${varepsilon}$ RI expression, the cells were successively incubated withmouse IgE, biotinylated rat anti-mouse IgE, and streptavidin-CyChrome(all from PharMingen). Cells were analyzed on a FACScaliburflow cytometer (Becton Dickinson Immunocytometry Systems).

cDNA constructs and viral constructs. SLP-76 mutants were generated from mouse SLP-76 cDNA by PCRand then cloned into the Moloney murine leukemia virus (MLV)-basedretroviral pMMP vector. SLP-76 cDNA was cloned upstream of aninternal ribosomal entry site that precedes the gene encodinggreen fluorescent protein (GFP). This vector construct onceintegrated into the host genome directs expression of a bicistronicmRNA encoding both SLP-76 and GFP.

Virus production and viral transduction of BMMC. High-titer vesicular stomatitis virus G protein (VSV-G)-pseudotypedretrovirus-containing supernatants were obtained after calciumphosphate transfection of the 293T packaging cell line (50)with individual retroviral vectors as well as expression plasmidsfor MLV Gag-Pol and VSV-G (41). Viral particle-containing supernatantwas added to SLP-76-deficient BMMC cultures prestimulated withstem cell factor (Biosource International) and 8 µg ofPolybrene per ml (Sigma, St. Louis, Mo.) in fibronectin (Sigma)-coatedplates. After 24 h, the medium was changed to WCM. GFP expressionwas assessed after 2 to 3 days of culture, and GFP-expressingcells were sorted (FACSVantage SE flow cytometer; Becton Dickinson).

ß-Hexosaminidase release assay. BMMC (10⁶) were incubated in WCM containing 2.5 µg ofrat IgE per ml for 1 h on ice. After washing, pellets were resuspendedon ice in WCM containing 0.1, 1, and 10 µg of F(ab')₂fragments per ml of mouse anti-rat immunoglobulins (Igs) (JacksonImmunoResearch) and incubated at 37°C for 20 min. The reactionwas stopped by centrifugation. The pellets were resuspendedin their original volume and lysed with WCM containing 0.5%Triton X-100. Aliquots of supernatants and cell lysates wereincubated in duplicates with substrate solution (1.3 mg of p-nitrophenyl-ß-D-2-acetamido-2-deoxyglucopyranozideper ml in 0.1 M citrate [pH 4.5]) as described previously (44).The percent release values for each experimental condition werecalculated by the formula [S/(S + P)] x 100, where S and P,respectively, are the ß-hexosaminidase contents ofthe supernatant and pellet from each sample. The net releasevalues were obtained by subtracting the optical density formedium alone from those for the supernatant and pellet.

Measurement of intracellular calcium. BMMC (10 x 10⁶/ml) were preloaded for 1 h on ice with 5 µgof rat IgE per ml and the calcium-sensitive dye indo 1-AM (5µM) in calcium buffer (phosphate-buffered saline [pH 7.4],supplemented with 5 mM glucose, 1 mM CaCl₂, 0.5 mM MgCl₂, 10mM HEPES, and 1% bovine serum albumin). Cells were washed twicewith calcium buffer and resuspended to the original concentration.Three hundred microliters of cells was analyzed for calciummobilization in a spectrophotometer (LS50; Perkin-Elmer CetusInstruments). Excitation and emission wavelengths were 331 and410 nm, respectively. Receptor-mediated calcium release wasmonitored after addition of F(ab')₂ anti-rat Ig (25 µg/ml),and maximal release was determined after addition of ionomycin(10 µM final concentration). Values were plotted as apercentage of calcium release triggered by ionomycin.

Measurement of IL-6 production. BMMC (10⁶) preloaded for 1 h on ice with 2.5 µg of ratIgE per ml were incubated with 1 µg of F(ab')₂ anti-ratIg per ml for 24 h in WCM at 37°C. Supernatants were assayedfor IL-6 with an enzyme-linked immunosorbent assay (ELISA) kit(R&D Systems).

Western blotting and immunoprecipitation. SLP-76 proteins were detected after lysis in Laemmli bufferand Western blotting with a rabbit anti-SLP-76 antiserum (30).For immunoprecipitation experiments, BMMC were preloaded for1 h on ice with 2.5 µg of rat IgE per ml and then stimulatedfor 2 min (SLP-76 immunoprecipitation) or 7 min (PLC- ${gamma}$ immunoprecipitation)at 37°C with 10 µg of F(ab')₂ anti-rat Ig per ml inWCM. After centrifugation, cells were lysed for 15 min on icein lysis buffer (50 mM Tris [pH 7.4], 150 mM NaCl, 1 mM EDTA,1% NP-40) supplemented with the protease inhibitors diisopropylfluorophosphate (3 mM; Sigma), phenylmethylsulfonyl fluoride(PMSF; 1 mM; Sigma), antipain (50 µg/ml), (4-amidinophenyl)-methane-sulfonylfluoride (APMSF; 20 µg/ml), and complete protease inhibitorcocktail (all from Roche Molecular Biochemicals), as well asphosphatase inhibitors (0.5 mM Na₃VO₄, 5 mM NaF). Clarifiedcell lysates were subjected to immunoprecipitation with antibodiesto SLP-76 (30), PLC- ${gamma}$ 1 (Upstate Biotechnology), or PLC- ${gamma}$ 2 (SantaCruz Biotechnology). Immunoblotting was performed with antiphosphotyrosinemonoclonal antibody 4G10 (Upstate Biotechnology), anti-pTyr(Sigma), anti-btk (BD Biosciences), and anti-Vav (Upstate Biotechnology).

Activated MAPKs were detected by immunoblotting cell lysateswith phosphoprotein-specific antibodies (phospho-ERK1/2; SantaCruz Biotechnology), phospho-SAPK/JNK and phospho-p38 (bothfrom NewEngland Biolabs). Immunoreactive bands were detectedby enhanced chemiluminescence (Amersham Pharmacia Biotech).

Statistical analysis and densitometry. Statistical analyses of the data were performed with the Prismsoftware (version 3.0a). Blots were scanned and analyzed withNIH Image software (version 1.62). Results were expressed asfold induction over baseline and were normalized to signal forloading.

RESULTS

Top

Results

Functional reconstitution of SLP-76^-/- BMMC by retroviral transfer of WT SLP-76. BMMC from SLP76^-/- mice were reconstituted by retroviral transferof a bicistronic construct encoding WT SLP-76 and GFP or ofa control vector encoding only GFP. GFP-expressing cells weresorted and analyzed. All cultures contained >90% GFP-positivecells at the time of analysis. The levels of Fc ${varepsilon}$ RI expressionwere equivalent in BMMC from WT mice and SLP-76^-/- mice, aswell as virally transduced BMMC from SLP76^-/- mice (Fig. 1A).Expression of WT SLP-76 protein was confirmed by Western blotting(Fig. 1B).

View larger version (35K):
[in this window]
[in a new window]
FIG. 1. Retroviral transduction with SLP-76 reconstitutes Fc ${varepsilon}$ RI-mediated signaling in SLP-76^-/- mast cells. (A) IgE receptor expression on BMMC from WT, SLP-76^-/- (KO) mice, and SLP-76^-/- BMMC reconstituted (reconst.) with control vector or WT SLP-76. Cells were treated with mouse IgE and then incubated with biotinylated anti-IgE and streptavidin-CyChrome (solid line). Control staining was with biotinylated anti-IgE and streptavidin-CyChrome alone (dashed line). (B) SLP-76 protein expression was assessed by immunoblotting after separation of proteins by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. (C) Calcium mobilization was detected in BMMC loaded with indo 1-AM. BMMC (10 x 10⁶/ml) were sensitized with IgE (5 µg/ml) for 1 h at room temperature and then stimulated with 25 µg of F(ab')₂ anti-rat Ig per ml followed later by ionomycin (10 µM) at the indicated time points. Results are expressed as percentage of ionomycin-induced calcium mobilization. (D) Release of ß-hexosaminidase. BMMC (10⁶) were incubated with 2.5 µg of IgE per ml for 1 h on ice and then stimulated with F(ab')₂ anti-rat Ig at the indicated concentrations. The results represent the mean ± standard deviation of three experiments, each performed in duplicate. **, P < 0.01; *, P < 0.05. ns, not significant as determined by Student's t test compared to WT. (E) IL-6 release by BMMC. BMMC (10⁶) preloaded for 1 h on ice with 2.5 µg of rat IgE per ml were incubated with 1 µg of F(ab')₂ anti-rat Ig per ml for 24 h. IL-6 release was determined by ELISA. The results represent the mean ± standard deviation of four experiments. **, P < 0.01. ns, not significant compared to WT, as determined by Student's t test.

SLP-76^-/- BMMC are severely deficient in their ability to mobilizecalcium, degranulate, and secrete IL-6 following Fc ${varepsilon}$ RI ligation(44). To validate the retroviral transduction system as a toolto examine the ability of SLP-76 mutants to restore SLP-76 function,BMMC transduced with WT SLP-76 were loaded with rat IgE andchallenged with F(ab')₂ anti-rat Ig. Retroviral transductionof SLP-76^-/- BMMC with WT SLP-76-encoding vector, but not controlvector, restored to normal their capacity to mobilize calcium(Fig. 1C) and to release the granule enzyme ß-hexosaminidaseat all three concentrations of cross-linking reagent tested(Fig. 1D). Retroviral transduction of WT SLP-76 also restoredto normal the capacity of SLP-76^-/- BMMC to secrete IL-6 followingFc ${varepsilon}$ RI ligation (Fig. 1E).

Reconstitution of SLP-76^-/- BMMC with SLP-76 mutants. Six SLP-76 mutants were generated in order to delineate therole of individual SLP-76 domains and tyrosine residues in signalingthrough Fc ${varepsilon}$ RI. Three deletion mutants included an N-terminaldeletion mutant (amino acids [aa] 2 to 156; SLP-76 ${Delta}$ 2-156), aGads binding site deletion mutant (aa 224 to 244; SLP-76 ${Delta}$ 224-244)and an SH2-domain deletion mutant (aa 422 to 533; SLP-76 ${Delta}$ SH2).In three other mutants, tyrosine (Y) residues Y112, Y128, andY145 in the N-terminal domain of murine SLP-76, which correspondto Y113, Y128, and Y145 in human SLP-76, were substituted forwith phenylalanine (F): Y112/128/145F in SLP-76 FFF, Y112/128F,in SLP-76 FFY, and Y145F in SLP-76 YYF (Fig. 2A). The FFY mutantwas specifically chosen because it might be uncoupled from Vav(14, 45) yet retain the ability to bind Itk via Y145 (6). SurfaceFc ${varepsilon}$ RI expression in BMMC reconstituted with each of the constructsand sorted for GFP expression was comparable to that of WT BMMC(Fig. 2B). The level of expression of SLP-76 mutant proteinswas verified by Western blotting and was similar to or higherthan SLP-76 expression in WT BMMC (Fig. 2C).

View larger version (34K):
[in this window]
[in a new window]
FIG. 2. Reconstitution of SLP-76^-/- BMMC with SLP-76 and SLP-76 mutants. (A) Schematic representation of SLP-76 mutants. (B) IgE receptor expression on BMMC reconstituted with SLP-76 constructs. Cells were labeled as described in the legend to Fig. 1. Levels of Fc ${varepsilon}$ RI expression on WT, KO, and control vector- and WT SLP-76-transduced cells, are shown in Fig. 1A. (C) SLP-76 protein expression in WT, KO, and transduced BMMC as assessed by immunoblotting (Western blotting [WB]) with rabbit anti-SLP-76 antiserum. The membrane was reprobed with anti-PLC- ${gamma}$ 2 to control for loading.

Fc ${varepsilon}$ RI-mediated ß-hexosaminidase release. Reconstitution with either SLP-76 ${Delta}$ 2-156 or SLP-76 ${Delta}$ 224-244, completelyfailed to restore the capacity of SLP-76^-/- BMMC to releaseß-hexosaminidase after Fc ${varepsilon}$ RI ligation (Fig. 3A). Incontrast, ß-hexosaminidase release was substantiallyrestored by SLP-76 ${Delta}$ SH2 (to 77% of WT reconstituted BMMC at 10µg of anti-IgE per ml; n = 3). These results suggest thatboth the N-terminal domain and the Gads binding site of SLP-76,but not its SH2 domain, are essential for Fc ${varepsilon}$ RI-mediated mastcell granule exocytosis.

View larger version (37K):
[in this window]
[in a new window]
FIG. 3. IgE-mediated in vitro degranulation and IL-6 release of BMMC reconstituted (reconst.) with SLP-76 constructs. (A) ß-Hexosaminidase release of BMMC with mutant construct is shown in comparison with BMMC transduced with WT SLP-76. The results represent the mean ± standard deviation of three experiments. Cells were sensitized and stimulated as described in the legend to Fig. 1C. *, P < 0.05; **, P < 0.01; ***, P < 0.001, as determined by analysis of variance. (B) BMMC were sensitized and stimulated as described in the legend to Fig. 1E, and secreted IL-6 was quantitated 24 h later. The results shown represent the mean ± standard deviation of four experiments. **, P < 0.01; ***, P < 0.001. ns, not significant as determined by Student's t test compared to SLP-76.

SLP-76 ${Delta}$ 2-156 lacks the tyrosine residues that have been shownto play an important role in signaling via the TCR. We assessedthe role of these residues in Fc ${varepsilon}$ RI signaling by examining thecapacity of SLP-76 FFF, SLP-76 FFY, and SLP-76 YYF mutants torestore degranulation in SLP-76-deficient BMMC. SLP-76 FFF andSLP-76 YYF only partially restored the capacity of SLP-76^-/-BMMC to release ß-hexosaminidase following Fc ${varepsilon}$ RI ligation(to 43 and 32% of WT reconstituted BMMC, respectively, at 10µg of anti-IgE per ml; n = 3). In contrast, SLP-76 FFYsubstantially restored ß-hexosaminidase release (to78% of WT reconstituted BMMC, at 10 µg of anti-IgE perml; n = 3) (Fig. 3A). These results suggest that Y145 playsan important role in Fc ${varepsilon}$ RI-mediated granule exocytosis.

IL-6 production. IL-6 secretion remained severely impaired following reconstitutionwith SLP-76 ${Delta}$ 2-156 and SLP-76 ${Delta}$ 224-244, but was partially restoredfollowing reconstitution with SLP-76 ${Delta}$ SH2 (to 42% of WT reconstitutedBMMC; n = 4) (Fig. 3B). These results suggest that both theN-terminal domain and the Gads binding site of SLP-76 are essentialfor Fc ${varepsilon}$ RI-mediated IL-6 secretion by mast cells. However, theSH2 domain is required for optimal IL-6 secretion.

IL-6 secretion was restored to a large extent by all three Y-to-Fmutants: SLP-76 FFF, SLP-76 FFY, and SLP-76 YYF (to 65, 72,and 83%, respectively, of IL-6 secretion by BMMC reconstitutedwith WT SLP-76; n = 4) (Fig. 3B). This suggests that phosphorylationof the SLP-76 N-terminal tyrosine residues is not critical forFc ${varepsilon}$ RI-mediated IL-6 secretion.

Calcium mobilization. The rapid rise in intracellular calcium concentration that followsFc ${varepsilon}$ RI engagement is impaired in SLP-76^-/- BMMC (44). Calciummobilization in SLP-76 ${Delta}$ 2-156- and in SLP-76 ${Delta}$ 224-244 BMMC remainedindistinguishable from that of SLP-76^-/--transduced BMMC (Fig.4). In contrast, BMMC transduced with SLP-76 ${Delta}$ SH2 responded similarlyto WT BMMC (Fig. 4). These results suggest that the N-terminaldomain and the Gads binding site of SLP-76, but not its SH2domain, are essential for Fc ${varepsilon}$ RI-mediated calcium mobilizationin mast cells.

View larger version (39K):
[in this window]
[in a new window]
FIG. 4. Calcium mobilization in reconstituted BMMC in response to Fc ${varepsilon}$ RI ligation. Change of fluorescence of the calcium-sensitive dye indo 1-AM was monitored for the indicated time. IgE-sensitized cells were stimulated with F(ab')₂ anti-rat Ig and ionomycin at the indicated time points ( ${blacktriangleup}$ ). BMMC reconstituted (reconst.) with a mutant construct are shown in comparison with SLP-76^-/- (KO) or WT BMMC analyzed in parallel. Results are expressed as percentage of ionomycin-induced calcium mobilization. Similar results were obtained in at least five experiments for each of the mutants.

BMMC reconstituted with SLP76 FFF exhibited decreased calciumflux following Fc ${varepsilon}$ RI ligation. The calcium response was restoredby SLP-76 YYF and SLP-76 FFY. These results suggest that tyrosinephosphorylation of tyrosine residues in the N-terminal domainof SLP-76 is required for Fc ${varepsilon}$ RI-mediated calcium fluxes, butthat the function of Y112 and Y128 is redundant with that ofY145.

Tyrosine phosphorylation of SLP-76 and coimmunoprecipitation with btk and Vav. Fc ${varepsilon}$ RI ligation induces the rapid tyrosine phosphorylation SLP-76(44). Figure 5A shows that Fc ${varepsilon}$ RI ligation, resulted in tyrosinephosphorylation of WT SLP-76 and, to a lesser extent, the threeSLP-76 Y-to-F mutants (WT>YYF>FFY>FFF), indicatingthat Y112, Y128, and Y145 are the major sites of tyrosine phosphorylationfollowing Fc ${varepsilon}$ RI ligation.

View larger version (44K):
[in this window]
[in a new window]
FIG. 5. Tyrosine phosphorylation of PLC- ${gamma}$ in response to Fc ${varepsilon}$ RI ligation. (A) Tyrosine phosphorylation of SLP-76 and coimmunoprecipitation with btk and Vav. BMMC were sensitized with rat IgE (2.5 µg of rat IgE per ml) followed by cross-linking with F(ab')₂ anti-rat Ig (10 µg/ml) and incubated for 2 min at 37°C. Cell lysates were immunoprecipitated (IP) with rabbit anti-SLP-76 antiserum. Membrane was successively probed with anti-pTyr, anti-btk, anti-SLP-76, and anti-Vav antibodies. The degree of association of SLP-76 with btk and Vav after stimulation was normalized to the signal for SLP-76 as determined by densitometry. Similar results were obtained in two experiments. (B) Tyrosine phosphorylation of PLC- ${gamma}$ 1 and PLC- ${gamma}$ 2 in WT and SLP-76^-/- BMMC. BMMC were sensitized with IgE and stimulated for 7 min as described above. Cell lysates were immunoprecipitated with anti-PLC- ${gamma}$ 1 and anti-PLC- ${gamma}$ 2. PLC- ${gamma}$ 1/2 tyrosine phosphorylation was analyzed by immunoblotting with antiphosphotyrosine antibody (4G10). The membranes were reprobed with anti-PLC- ${gamma}$ 1 and anti-PLC- ${gamma}$ 2 to control for loading. (C) Tyrosine phosphorylation of PLC- ${gamma}$ 2 in SLP-76-reconstituted BMMC. The top two sets of lanes represent a single experiment with cells stimulated simultaneously and then processed in parallel. The bottom set of lanes represents a separate experiment for SLP-76 YYF. Membranes were reprobed with anti-PLC- ${gamma}$ 2 to control for loading. Similar results were found in three different experiments. Fold induction normalized to signal for loading was determined by densitometry.

Y112 and Y128 have been reported to be important for the associationof Vav with SLP-76, and Y145 has been shown to be implicatedin the association of the Tec family kinase Itk with SLP-76following TCR ligation (6, 40, 53, 55). There was a baselineassociation of Vav and btk with WT SLP-76 in mast cells, aspreviously described for Vav and Itk in Jurkat T cells (53,55). Fc ${varepsilon}$ RI ligation resulted in increased association of Vavwith WT SLP-76 and SLP-76 YYF, but not with SLP-76 FFY or SLP-76FFF. Fc ${varepsilon}$ RI ligation resulted in a modest increase in the associationof the Tec family kinase btk with WT SLP-76 and the SLP-76 FFYmutant and a weaker increase in the association of btk withSLP-76 YYF and SLP-76 FFF. These results suggest that Y112 andY128 are important for the increased association of Vav withSLP-76 and that Y145 is important for the increased associationof btk with SLP-76 after Fc ${varepsilon}$ RI ligation in mast cells. However,contributions from additional residues or domains of SLP-76to its association with Vav and btk cannot be ruled out.

Tyrosine phosphorylation of PLC- ${gamma}$ 2. Both PLC- ${gamma}$ 1 and PLC- ${gamma}$ 2 become rapidly tyrosine phosphorylatedafter ligation of Fc ${varepsilon}$ RI in mast cells (61). Phosphorylation ofPLC- ${gamma}$ 1 was reduced and phosphorylation of PLC- ${gamma}$ 2 was virtuallyabsent in SLP-76^-/- BMMC after Fc ${varepsilon}$ RI ligation (Fig. 5B). Thisdecrease in PLC- ${gamma}$ phosphorylation did not simply reflect a globaldecrease in protein tyrosine phosphorylation as total proteintyrosine phosphorylation following Fc ${varepsilon}$ RI ligation in SLP-76^-/-mast cells is indistinguishable from that in WT mast cells (44).

We specifically focused on the capacity of SLP-76 mutants torestore PLC- ${gamma}$ 2 phosphorylation. SLP-76 ${Delta}$ 2-156 and SLP-76 ${Delta}$ 224-244failed to restore PLC- ${gamma}$ 2 phosphorylation, whereas a modest increasein PLC- ${gamma}$ 2 phosphorylation was detected in several independentexperiments in BMMC reconstituted with SLP-76 ${Delta}$ SH2 (Fig. 5C).These results suggest that the N-terminal domain and the Gadsbinding site of SLP-76 are essential for Fc ${varepsilon}$ RI-mediated PLC- ${gamma}$ 2activation.

Reconstitution with SLP76 FFF, SLP-76 FFY, and SLP-76 YYF resultedin normal phosphorylation of PLC- ${gamma}$ 2 following Fc ${varepsilon}$ RI cross-linking(Fig. 5C) suggesting that phosphorylation of Y112, Y128, andY145 may not be critical for Fc ${varepsilon}$ RI-mediated PLC- ${gamma}$ 2 activationin mast cells.

MAPK activation. Phosphorylation of ERK1/2 and p38 after Fc ${varepsilon}$ RI cross-linking wasmodestly reduced and less sustained in SLP-76^-/- BMMC than inWT BMMC (Fig. 6A). In contrast, phosphorylation of p54 JNK afterFc ${varepsilon}$ RI cross-linking was severely reduced in SLP-76^-/- BMMC (Fig.6B). This suggests that SLP-76 is essential for JNK activation,but not for ERK1/2 and p38 activation.

View larger version (54K):
[in this window]
[in a new window]
FIG. 6. Activation of MAPKs in response to Fc ${varepsilon}$ RI ligation in BMMC. BMMC were stimulated for the indicated times and lysed in sodium dodecyl sulfate-polyacrylamide gel electrophoresis sample buffer. Aliquots of the lysates were analyzed in parallel for phosphorylation of ERK1/2 and p38 (A) and SAPK/JNK (B) by Western blotting with the corresponding phosphospecific antibodies. The two sets of lanes in panels A and B represent the same experiment with cells stimulated simultaneously and then processed in parallel. Membranes were reprobed with kinase-specific antibodies to control for loading. Fold induction normalized to signal for loading was determined by densitometry. n.d., not determined.

Reconstitution with SLP-76 ${Delta}$ 2-156, SLP-76 ${Delta}$ 224-244, and SLP-76 ${Delta}$ SH2 failed to rescue the defect in MAPK phosphorylation afterFc ${varepsilon}$ RI ligation. However, SLP-76 YYF, but not SLP-76 FFF or SLP-76FFY, restored the phosphorylation of ERK, p38, and JNK to normal.This suggests that phosphorylation of the tyrosine residuesY112 and Y128 in the N-terminal domain of SLP-76 is criticalfor MAPK activation in mast cells.

DISCUSSION

Top

Discussion

In this study, we used SLP-76^-/- BMMC retrovirally transducedwith SLP-76 mutants to address the structural requirements forSLP-76 in signaling via Fc ${varepsilon}$ RI. The results obtained are summarizedin Table 1. They demonstrate a differential role for SLP-76domains in mast cell function and a critical role of residueY145 in IgE-mediated degranulation.

View this table:
[in this window]
[in a new window]
TABLE 1. Summary of IgE-mediated mast cell activation events in SLP-76^-/- BMMC reconstituted with SLP-76 mutants

SLP-76^-/- mast cells transduced with a SLP-76 deletion mutantlacking the N-terminal domain (SLP-76 ${Delta}$ 2-156) were indistinguishablefrom SLP-76^-/- mast cells in all Fc ${varepsilon}$ RI-mediated activation eventswe tested. The critical role of the N-terminal SLP-76 domainin mast cell activation is consistent with its requirement forT-cell development and activation. Transgenic expression ofSLP-76 ${Delta}$ 2-156 in SLP-76^-/- mice failed to rescue the block inthymocyte development from the DP to DN stage (30). Expressionof the same mutant in SLP-76-deficient Jurkat J14 cells didnot reconstitute NFAT activation following stimulation withsoluble anti-CD3 monoclonal antibody (57).

A major contribution to the function of the SLP-76 N-terminaldomain in T cells is made by tyrosine residues Y112, Y128, andY145, which become phosphorylated after TCR stimulation (15).Fc ${varepsilon}$ RI ligation resulted in tyrosine phosphorylation of WT SLP-76and, to a lesser extent, the three SLP-76 Y-to-F mutants (WT>YYF>FFY>FFF),indicating that Y112, Y128, and Y145 are the major sites oftyrosine phosphorylation of SLP-76 following Fc ${varepsilon}$ RI ligation inmast cells. The SLP-76 FFF mutant in which all three tyrosineresidues are substituted for with phenylalanine partially reconstitutedß-hexosaminidase release and calcium fluxes and substantiallyreconstituted IL-6 release, whereas JNK kinase phosphorylationremained impaired. The fact that calcium fluxes were reducedin the face of normal PLC- ${gamma}$ 2 phosphorylation is not entirelysurprising, because the degree of global phosphorylation ofPLC- ${gamma}$ may not always correlate with its activity (16). Takentogether with the complete failure of SLP-76 ${Delta}$ 2-156 to reconstituteFc ${varepsilon}$ RI signaling in SLP-76^-/- BMMC, the findings with SLP-76 FFFsuggest that residues or motifs in addition to tyrosines inthe N-terminal domain of SLP-76 may contribute to its adapterfunction in mast cells. This is supported by recent reportsthat SLP-76 FFF partially restores NF- ${kappa}$ B and NFAT signaling inSLP-76-deficient Jurkat T cells (22, 57). Furthermore, reconstitutionof SLP-76^-/- mice with SLP-76 FFF mutant partially rescues T-celldevelopment (26, 39); however, the function of the T cells inthese mice is severely impaired. In contrast to our findingsin mast cells, reconstitution with SLP-76 FFF completely failsto correct the defect in calcium mobilization and PLC- ${gamma}$ phosphorylationin response to TCR ligation. Furthermore, T cells from SLP-76^-/-mice reconstituted with SLP-76 FFF fail to cluster their TCRand proliferate in response to anti-CD3 (39).

In an attempt to further define the roles of individual N-terminaldomain tyrosine residues in Fc ${varepsilon}$ RI-mediated activation of mastcells, we reconstituted SLP-76^-/- BMMC with two mutants, SLP-76FFY and SLP-76 YYF. These two mutants were chosen on the basisof the observations in T cells that Y112 and Y128 (both withinpYESP motifs), which are mutated in SLP-76 FFY, upon phosphorylationmediate the binding of SLP-76 to Vav and Nck (5, 10, 54-56),whereas Y145 (pYEPP motif), which is mutated in SLP-76 YYF,is implicated in the association of SLP-76 and the Tec familykinase Itk (6, 53). Our results suggest that Y112 and Y128 areimportant for the increased association of Vav with SLP-76,and that Y145 is important for the increased association ofbtk with SLP-76 after Fc ${varepsilon}$ RI ligation in mast cells. Granule exocytosiswas restored by the SLP-76 FFY mutant, but not by the SLP-76YYF mutant. This suggests that Y145 is critical for degranulation,possibly by recruiting btk or other Tec family kinases to theSLP-76-Gads-LAT-PLC- ${gamma}$ complex. This hypothesis is supported bythe finding that degranulation is impaired in mast cells frombtk-deficient mice (20). Vav, which has been shown to associatewith SLP-76 YYF (14, 45), but not with SLP-76 FFY, is importantfor IgE-mediated exocytosis. Our results suggest that directassociation of Vav with SLP-76 may not be required, since SLP-76FFY restored exocytosis (34). It should be noted that Vav phosphorylationfollowing Fc ${varepsilon}$ RI cross-linking is normal in SLP-76^-/- mast cellsand that Vav can associate with Fc ${varepsilon}$ RI (44, 51).

Calcium mobilization and PLC- ${gamma}$ 2 phosphorylation following Fc ${varepsilon}$ RIligation were restored by both SLP-76 FFY and SLP-76 YYF. Thissuggests that Y112 and/or Y128 plays a role in calcium mobilizationthat is redundant with that of Y145. Y145-mediated recruitmentof Tec kinases to the SLP-76-Gads-LAT-PLC- ${gamma}$ complex may explainthe capacity of SLP-76 FFY to restore calcium mobilization inSLP-76^-/- mast cells. The finding that SLP-76 YYF restores calciummobilization suggests that interaction of Vav and SLP-76 maybe important for this process. The observation that calciumfluxes, but not exocytosis, was restored by SLP-76 YYF suggeststhat while calcium mobilization is essential for degranulation,other signaling pathways, which are dependent on phosphorylationof the SLP-76 residue Y145, are also involved.

Phosphorylation of JNK was severely reduced in SLP-76^-/- BMMC.Expression of SLP-76 YYF, but not SLP-76 FFY, restored JNK phosphorylation.Vav1-deficient BMMC fail to activate JNK after Fc ${varepsilon}$ RI stimulation(34, 52). Restoration of JNK activation by SLP-76 YYF supportsthe notion that a functional association between SLP-76 andVav is required for optimal activation of the JNK kinase pathway.In addition, the adaptor protein Nck, which, like Vav, bindsto phosphorylated Y113 and Y128 may contribute to JNK activationby recruiting PAK1 to the Vav-SLP-76-containing complex (5,56, 58). The finding that SLP-76 FFF and SLP-76 FFY restoredIL-6 secretion, but not JNK phosphorylation, may reflect thepresence of JNK-independent pathways for the induction of Fc ${varepsilon}$ RI-mediatedIL-6 production. The activation of such pathways would be dependenton motifs in the N-terminal domain of SLP-76 other than tyrosineresidues, since IL-6 secretion is abolished in mast cells reconstitutedwith SLP-76 ${Delta}$ 2-156.

SLP-76 YYF, but not SLP-76 FFY, restored ERK1/2 and p38 phosphorylationto WT levels, suggesting that optimal activation of the ERK1/2and p38 pathways may require the association of SLP-76 withVav and/or Nck. It should be noted that Fc ${varepsilon}$ RI-mediated ERK1/2and p38 phosphorylation was nevertheless substantial in theabsence of SLP-76. In contrast, ERK1/2 activation is undetectablein SLP-76-deficient Jurkat T cells following TCR ligation (59).These differences suggest that Fc ${varepsilon}$ RI, but not the TCR, may becoupled to ERK1/2 activation independently of SLP-76. This couldpossibly be via LAT/Grb2/Sos (29, 36, 47) and/or Vav, whichmay be linked to ERK via a Rac-PAK1-MEK cascade (2, 17). However,since LAT and Vav are both phosphorylated normally in SLP-76-deficientT cells and in SLP-76^-/- BMMC (44, 59), it is unlikely thatthese two pathways explain the differential phosphorylationof ERK1/2 in these two cell types. RasGRP family members, whichare expressed in both mast cells and T cells, can activate theRas/MAPK pathway through membrane recruitment by the PLC- ${gamma}$ productdiacylglycerol (DAG) (12, 13, 60). PLC- ${gamma}$ 1 phosphorylation isobservable after Fc ${varepsilon}$ RI ligation in SLP76^-/- BMMC, but is undetectablein SLP-76-deficient Jurkat T cells following TCR ligation (59).This suggests that activation of ERK1/2 in SLP76^-/- BMMC maybe mediated by RasGRP.

SLP-76 ${Delta}$ 224-244, which lacks the Gads binding site, failed toreconstitute all aspects of Fc ${varepsilon}$ RI signaling examined. This suggestsan absolute requirement for Gads binding in targeting SLP-76to signaling complexes in mast cells. However, the same mutantwas able to partially restore T-cell development when introducedin the SLP-76^-/- background and to partially restore calciumfluxes in the T cells from these mice (30, 39) as well as inSLP-76-deficient Jurkat T cells (4, 38, 57). Thus, while SLP-76binding to Gads is obligatory for Fc ${varepsilon}$ RI signaling in mast cells,it is in part dispensable for TCR signaling in T cells. Residualsignaling in T cells reconstituted with SLP-76 ${Delta}$ 224-244 mightresult from a Gads-independent recruitment of SLP-76 to LATand/or GEMs in T cells, as suggested by the partially restoredTCR signaling in LAT-deficient T cells expressing a LAT mutantthat does not recruit Gads (62).

The SH2 domain of SLP-76 mediates its association with ADAP/FYB/SLAP130and with HPK1 (11, 37, 48). Our results demonstrate that theSH2 domain of SLP-76 plays a minimal role in Fc ${varepsilon}$ RI-mediated calciumfluxes and degranulation, although one of its ligands, ADAP,has been reported to promote mediator release upon overexpressionin RBL-2H3 cell line (18). On the other hand, the SH2 domainof SLP-76 is required for optimal IL-6 release and for JNK activationafter Fc ${varepsilon}$ RI ligation, possibly because it may be linked via HPK1to MAPK pathways (23, 27). In light of the results from theother mutants examined, JNK activation following Fc ${varepsilon}$ RI ligationmay require contributions from Y112, Y128, the Gads bindingdomain and the SH2 domain of SLP-76.

The restricted role of the SH2 domain in Fc ${varepsilon}$ RI-mediated activationis consistent with the observations that deletion or mutationof the SH2 domain has a modest effect on the capacity of SLP-76to reconstitute TCR signaling and suggest that the SH2 domainfunctions similarly in mast cells and T cells. SLP-76 ${Delta}$ SH2 andSLP-76 SH2 domain mutant (R448K) reconstitute T-cell developmentin SLP-76^-/- mice and the T cells of these mice mobilize calciumnormally following TCR ligation, although they fail to proliferate(39). Furthermore, the SLP-76 SH2 domain mutant reconstitutescalcium fluxes and NFAT activation in Jurkat T cells (4, 38,57). It is interesting to note that the phenotype of ADAP-deficientmice resembles that of SLP-76^-/- mice reconstituted with SLP-76 ${Delta}$ SH2, as proximal TCR signaling pathways are normal in T cellsfrom these mice, but proliferation and IL-2 secretion are impaired(19, 22). The role of ADAP and HPK1 in Fc ${varepsilon}$ RI-mediated signalinghas not been determined.

The SLP-76 family of adaptor includes SLP-76, Blnk, and Clnk.Clnk is expressed in mast cells and in cytokine-stimulated Tcells, but not in resting T cells (18a). Clnk is tyrosine phosphorylatedupon Fc ${varepsilon}$ RI cross-linking and associates with PLC- ${gamma}$ , Vav, Grb2,and LAT. More importantly, overexpression of a mutant form ofClnk inhibits Fc ${varepsilon}$ RI-mediated degranulation, calcium mobilization,NFAT activation, and phosphorylation of LAT. Signaling via Clnkmay underlie the residual calcium mobilization and PLC- ${gamma}$ 1 phosphorylationobserved in SLP-76^-/- mast cells and may account for some ofthe differences between the capacity of SLP-76 mutants to reconstituteFc ${varepsilon}$ RI signaling in mast cells and T-cell function in SLP-76^-/-transgenic mice.

IgE-mediated mast cell degranulation plays a very importantrole in allergic diseases. The critical role of the SLP-76 residueY145 in mast cell degranulation demonstrated in this study suggeststhat it may provide a good target for therapeutic interventionsaimed at inhibiting the allergic response.

ACKNOWLEDGMENTS

We thank A. Flint for sorting GFP-expressing cells. Furthermore,we thank H. Oettgen and M. de la Fuente for critical evaluationof the manuscript.

A.K. received a research fellowship from the Swiss NationalScience Foundation and an AAAAI Zeneca Research Award. Thiswork was supported by USPHS grants AI-35714, in part by researchgrant no. 1-FY00-673 from The March of Dimes Birth Defects Foundation,and by research grants from the Hood Foundation and the Baxter,Aventis, and Gentiva Corporations.

FOOTNOTES

* Corresponding author. Mailing address: Division of Immunology, 300 Longwood Ave., Boston, MA 02115. Phone: (617) 355-7603. Fax: (617) 739-3145. E-mail: raif.geha{at}tch.harvard.edu.

REFERENCES

Top