Concentration of free cytoplasmic Ca2+
Glycosphingolipid-enriched membrane microdomains
Linker for activation of T cells
Rat basophilic leukemia
Previous studies using cytochalasins and latrunculin B, inhibitors of actin polymerization, showed that filamentous (F)-actin had a negative regulatory role in Fcϵ receptor I (FcϵRI) signaling. How F-actin is involved in regulating the activation of mast cells is unknown. In this study we investigated the role of F-actin in mast cell activation induced by aggregation of the glycosylphosphatidylinositol (GPI)-anchored proteins Thy-1 and TEC-21, and compared it to activation via FcϵRI. Pretreatment of rat basophilic leukemia cells with latrunculin B inhibited the Thy-1-induced actin polymerization and elevated the Thy-1-mediated secretory and calcium responses. Inhibition of actin polymerization followed by Thy-1 aggregation resulted in an increased tyrosine phosphorylation of Syk, phospholipase Cγ (PLCγ), Gab2 and linker for activation of T cells (LAT) adapters, and some other signaling molecules. Enzymatic activities of phosphatidylinositol 3-kinase, PLCγ, and phosphatase SHP-2 were also up-regulated, but tyrosine phosphorylation of ezrin was inhibited. Similar changes were observed in FcϵRI-activated cells. Significant changes in intracellular distribution, tyrosine phosphorylation, and/or enzymatic activities of signaling molecules occurred in latrunculin-pretreated cells before cell triggering. The combined data suggest that actin polymerization is critical for setting the thresholds for mast cell signaling via aggregation of both FcϵRI and GPI-anchored proteins.
Aggregation of the Fcϵ receptor I (FcϵRI) on mast cells and basophils, as well as other immunoreceptors on other cell types, by natural ligands or Ab initiates tyrosine phosphorylationof the receptor subunits, recruitment of signaling molecules, and their translocation to various compartments of the cell 1. In addition to enzymes, such as kinases and phosphatases, other molecules play a critical role in this process, including adapters and components of the cytoskeleton.
Previous studies have shown that pretreatment of rat basophilic leukemia (RBL) cells, which have been extensively used as a model for analysis of mast cell activation, with inhibitors of actinpolymerization, such as cytochalasin D and latrunculin B, enhances the FcϵRI-mediated Ca2+ mobilization and degranulation 2, 3. Cytochalasin D, which inhibits actin polymerization by capping the barbed end of actin filaments and preventing their elongation 4, enhanced both the early and the late events in FcϵRI-induced mast cell signaling 2. In contrast, latrunculin, which inhibits actin polymerization by sequestering the monomeric actin 5, had no effect on degranulation inducedby activation of the cells with A23187 and PMA or pervanadate, suggesting that it acts mainly on early stages of FcϵRI signaling 2. Further experiments showed an increased tyrosine phosphorylation of FcϵRI and Syk in latrunculin-pretreated and Ag-stimulated cells, and suggest that filamentous (F)-actin negatively regulates the interaction of aggregated FcϵRI with Lyn kinase and thus the initiation of signaling pathways 2.
Mast cells and RBL cells can also be activated by aggregation of the glycosylphosphatidylinositol (GPI)-anchored proteins Thy-1 and TEC-21 6–8. This activation is independent of the surface expression of FcϵRI 9 and tyrosine phosphorylation of FcϵRI subunits 8, and seems to reflect indirect interactions of these proteins with Lyn kinase within the glycosphingolipid-enriched membrane microdomains (GEM) 8, 10. It has been suggested that these microdomains are also involved in FcϵRI signaling 11.
To gain more insight in the role of actin filaments in mast cell signaling, we investigated early activation events initiated by aggregation of Thy-1 or TEC-21 in latrunculin-pretreated cells, and compared them to those induced by aggregation of FcϵRI. Since increased calcium and secretory responses in latrunculin-pretreated cells occurred within seconds after Ag exposure, we also analyzed the changes in subcellular distribution and other properties of the signaling molecules immediately preceding the mast cell triggering and sensitization with Ab.
2.1 Enhanced Thy-1-mediated secretory and calcium responses in latrunculin-pretreated cells
Previous studies showed that inhibition of F-actin polymerization correlated with enhanced calcium and secretory responses in Ag-activated RBL-2H3 cells, and suggested that GEM could be involved in this process 2, 3. Because the Thy-1 glycoprotein seems to be one of the major protein components of the exoplasmic leaflet of GEM in RBL cells and since its aggregation also induces cell activation, we decided to find out whether the Thy-1-mediated secretory response is also promoted by latrunculin. First we determined the changes in F-actin polymerization in Thy-1-activated cells. For these experiments the biotinylated anti-Thy-1.1 mAb-sensitized RBL cells were pre-incubated without or with 0.5 μM latrunculin and then exposed to streptavidin to induce Thy-1 aggregation. Data presented in Fig. 1A show that aggregation of Thy-1 caused a rapid increase in the level of detergent-insoluble F-actin with a maximum attained at 1 min (1.8-fold increase), followed by a slow decline to basal levels. There was an approximately 25% inhibition of detergent-insoluble F-actin basal levels in latrunculin-pretreated cells, and Thy-1 aggregation induced only a moderate increase in the amount of F-actin. Interestingly, the Thy-1 aggregation induced a prompter and more extensive formation of detergent-insoluble F-actin than aggregation of FcϵRI-IgE complexes with TNP-BSA (Fig. 1B).
In accordance with published data 2, 0.5 μM latrunculin alone did not cause any degranulation in RBL cells. Only weak and variable secretory responses were observed in OX7-sensitized and latrunculin-pretreated cells. However, an exposure to latrunculin followed by Thy-1 aggregation resulted in a dramatic degranulation response. Thus, although in control cells the Thy-1 activation for 5 min released approximately 14% of β-glucuronidase, more than 50% was released in latrunculin-pretreated and Thy-1-activated cells. Significant differences between control and latrunculin-pretreated cells were already noticeable at the first time interval analyzed (1 min), indicating that very early activation events are affected (Fig. 2A). Although the extent of degranulation in control cells was lower in Thy-1-activated than in FcϵRI-activated cells, the secretory responses in latrunculin-pretreated cells were comparable (Fig. 2A, B).
The differences between control and latrunculin-pretreated cells were confirmed by measurements of free cytoplasmic Ca2+ ([Ca2+]i). Again, aggregated Thy-1 induced a quicker and more extensive calcium response in latrunculin-pretreated cells than in control cells (Fig. 2C). However, in contrast to the secretory response, the calcium response in latrunculin-pretreated cells was lower in Thy-1- than in Ag-stimulated cells (Fig. 2D). The enhanced secretory and calcium responses in latrunculin-pretreated cells were not specific for Thy-1- or FcϵRI-induced activation, because they were also observed in RBL cells stimulated via TEC-21 glycoprotein (not shown). These data suggest that pretreatment with latrunculin generally enhances the activation induced via FcϵRI and GPI-anchored proteins.
2.2 Is FcϵRI involved in latrunculin-elevated Thy-1 signaling?
The similarity between secretory and calcium responses in latrunculin-pretreated cells triggered by means of Thy-1, TEC-21, or FcϵRI suggested that in the absence of stimulated actin polymerization it might be FcϵRI that is involved in Thy-1-mediated signaling. Physical interaction between FcϵRI and GPI-anchored proteins was analyzed by equilibrium ultracentrifugation in sucrose gradients. Most of the Thy-1 glycoprotein from cells solubilized with 1% Triton X-100 (Fig. 3A) or 0.06% Triton X-100 (not shown) was found in low-density fractions of sucrose gradient (fractions 3–10, corresponding to 15–30% sucrose). A significant amount of Thy-1 was also detected in fraction 23 containing cytoskeleton/ nuclear remnants. After pretreatment of the cells with latrunculin, the proportion of Thy-1 in low-density fractions increased and the peak showed a slight shift to the high-density fractions; a corresponding decrease of Thy-1 content in high-density fractions was observed.
In contrast to Thy-1, the majority of FcϵRI in unstimulated cells either untreated (Fig. 3B) or pretreated with latrunculin (not shown) was found in high-density fractions. After FcϵRI aggregation most of the receptors were associated with low-density fractions, and latrunculin had no significant effect on this distribution except that there was a reduced amount of FcϵRI associated with fraction 23. As described before 3, 8, 11, association of aggregated FcϵRI with low-density fractions was only observed in cells solubilized with 0.06% Triton X-100. Interestingly, there was no increased association of FcϵRI with low-density fractions in cells pretreated with latrunculin, activated via Thy-1, and solubilized in 0.06% Triton X-100 (Fig. 3B, open triangles). These data suggested that F-actin polymerization had no effect on physical association of aggregated Thy-1 with FcϵRI. Thus the enhanced calcium/secretory responses observed in latrunculin-pretreated and Thy-1-activated cells cannot be simply explained by an increased physical association of FcϵRI with GEM.
Tyrosine phosphorylation of the β and γ subunits of FcϵRI is the earliest known event occurring after FcϵRI engagement, and is enhanced by inhibitors of actin polymerization 2, 3. To determine the extent of tyrosine phosphorylation of β and γ subunits under different conditions, FcϵRI was immunoprecipitated from control or latrunculin-pretreated cells which were either nonsensitized and non-activated or activated via Thy-1 or FcϵRI, and then solubilized with 0.2% Triton X-100, which preserves the association of FcϵRI subunits. Immunoblotting analyses showed that latrunculin alone induced a weak increase in tyrosine phosphorylation of FcϵRI β subunit in non-activated or Thy-1-activated cells (Fig. 3C). In contrast, FcϵRI-activated cells exhibited strong tyrosine phosphorylation of both FcϵRI subunits, further elevated by latrunculin exposure. These data indicate that latrunculin causes an increase in tyrosine phosphorylation of FcϵRI subunits not only in activated but also in non-activated cells. The possibility that latrunculin might potentiate the secretory response independently of FcϵRI was also supported by data showing 2.2±0.3-fold increase (mean ± SD, n=4) of β-glucuronidase release from Thy-1-activated (30 min) RBL-γ–c.1 cells, defective in the expression of surface FcϵRI 9, after pretreatment with latrunculin.
2.3 Early activation pathways
The rapid increase in calcium and secretory responses in latrunculin-pretreated and Thy-1-, TEC-21- or FcϵRI-activated cells suggested that latrunculin affected the early biochemical events common for all these activation pathways. In an attempt to determine which pathways are primarily affected, we first analyzed tyrosine phosphorylation of Syk. This kinase becomes rapidly phosphorylated in Thy-1- 12, TEC-21- 8 or FcϵRI- 13 activated RBL cells, and has already been shown to exhibit an increased tyrosine phosphorylation in latrunculin-pretreated and FcϵRI-activated cells 2. Immunoblotting analysis showed that phosphorylation of Syk was slightly increased by latrunculin alone, i.e. in non-activated cells (Fig. 4A, B, 0 min). This increase was not attributable to the binding of anti-Thy-1 or IgE to the cells because a similar increase in Syk phosphorylation was also observed in latrunculin-treated but nonsensitized cells (not shown). Thy-1-mediated tyrosine phosphorylation of Syk was enhanced by latrunculin with a peak at 1 min after cell triggering. At later time intervals the difference between control and latrunculin-pretreated cells diminished but remained significant even after 30 min.
Next, we analyzed changes in phosphorylation of the Syk substrate, linker for activation of T cells (LAT). Tyrosine phosphorylation of this adapter molecule in non-stimulated cells was identical in both control and latrunculin-pretreated cells. After Thy-1 triggering the phosphorylation of LAT was enhanced by latrunculin; this stimulatory effect was transient (peak at 1 min) and disappeared after 15 min. In FcϵRI-activated cells (Fig. 4B), latrunculin had a similar effect on Syk and LAT, except that the enhanced Syk phosphorylation was more transient and the enhanced LAT phosphorylation extended for more than 15 min. We also analyzed the effect of latrunculin on Lyn kinase activity. However, in three independent immunocomplex kinase assays with enolase as a substrate we saw no change in Lyn kinase activity (not shown). These data suggested that rather than inducing changes in Lyn kinase activity, latrunculin might affect Lyn topography with respect to its substrates.
2.4 Changes in cellular distribution and properties of signaling molecules
In further experiments we investigated the spatial distribution and properties of signaling molecules in cells sequentially solubilized with saponin and Triton X-100. In this procedure, the cells are first permeabilized on ice with saponin, free cytoplasmic components are washed away, and the cellular ghosts are subsequently solubilized in Triton X-100. Our previous results showed that the procedure solubilized the plasma membrane components more efficiently than other methods routinely used for extraction of membrane proteins, and allowed a better assessment of the formation of large macromolecular complexes in the course of FcϵRI signaling 14, 15.
Using this method we first analyzed the distribution and tyrosine phosphorylation of phospholipase Cγ (PLCγ), whose activity seems to be negatively regulated by F-actin, as suggested by an increased amount of inositol 1,4,5-triphosphate (IP3) produced in latrunculin-pretreated and Thy-1- or FcϵRI-activated cells (Fig. 5A, B). We found that both the tyrosine phosphorylation and the amount of PLCγ1 and PLCγ2 associated with saponin-permeabilized cells were increased in Thy-1-activated cells (Fig. 5C). Interestingly, latrunculin alone (0 min) induced an increased association of both enzymes with cellular ghosts. In Thy-1-activated cells, latrunculin further enhanced the phosphorylation of PLCγ1, particularly 30 s after Thy-1 triggering, but had no effect on PLCγ2. Increased association of PLCγ1 and PLCγ2 with macromolecular complexes was also observed in latrunculin-pretreated and IgE-sensitized cells (0 min).
These experiments were extended to several other signaling molecules [paxillin, ezrin, phosphatidylinositol 3-kinase (PI3K), Gab2 adapter and SHP-2 phosphatase] which are involved in mast cell signaling and could potentially be affected by F-actin. Paxillin is a 68-kDa cytoskeletal protein that accumulates at focal adhesion sites. It has already been shown that aggregation of FcϵRI in RBL cells results in increased tyrosine phosphorylation of paxillin and its cytoplasmic redistribution 16. In Thy-1- or FcϵRI-activated cells paxillin showed an increased association with large macromolecular complexes (Fig. 6). Latrunculin alone (0 min) also enhanced this association. In FcϵRI-activated cells pretreatment with latrunculin increased tyrosine phosphorylation of paxillin, suggesting an increased activity and/or changes in topography of the corresponding kinases. Interestingly, aggregation of Thy-1 induced more dramatic changes in the distribution of paxillin than did FcϵRI aggregation.
Tyrosine phosphorylation of ezrin, a compound which functions as a linker between the actin skeleton and integral plasma membrane proteins 17, was strongly inhibited by pretreatment with latrunculin in both Thy-1- and FcϵRI-activated cells. This inhibition was not attributable to a decreased amount of ezrin immunoprecipitated from saponin/Triton X-100-solubilized cells, as inferred from the results of immunoblotting with anti-ezrin Ab (Fig. 6).
Early activation events in mast cells are dependent on the activity of PI3K. PI3K catalyzes the synthesis of phosphatidylinositol 3,4,5-triphosphate (PIP3) and phosphatidylinositol 3,4-biphosphate, and functionally interacts with Gab2, SHP-2, and several other proteins 18. We therefore monitored in control and Thy-1-activated cells the subcellular distribution of PI3K, its enzymatic activity, and the tyrosine phosphorylation of both PI3K and associated molecules (Fig. 7). When PI3K was immunoprecipitated from control (nonsensitized) saponin/Triton X-100-solubilized cells, its amount significantly rose after an exposure of the cells to latrunculin (2.3±0.4-fold increases, mean ± SD, n=4). Furthermore, latrunculin alone reproducibly increased the amount of tyrosine-phosphorylated proteins associated with PI3K and the enzymatic activity of PI3K as detected by immunocomplex kinase assay. Enzymatic activity of PI3K in OX7-sensitized cells was increased ninefold after pretreatment of the cells with latrunculin (0 min). Aggregation of Thy-1-biotinylated OX7 complexes with streptavidin induced no further increase in PI3K activity; in fact it reduced it, especially at later time intervals (2 and 5 min). PI3K complexes isolated from IgE-sensitized cells exhibited similar properties, namely that PI3K activity remained low in IgE-sensitized cells and was dramatically increased by latrunculin pretreatment even in the absence of FcϵRI aggregation (not shown).
Previous studies showed that multiple signaling molecules could be assembled by Gab2 scaffolding protein 18. Tyrosine-phosphorylated Gab2 provides the binding sites for Src homology-2 (SH2) domain-containing proteins, including the p85 subunit of PI3K, PLCγ, and phosphatase SHP-2. Gab2 immunoprecipitation experiments showed a rapid association of Gab2 with macromolecular complexes in Thy-1-activated cells at all time intervals studied (Fig. 8). Latrunculin alone (in nonsensitized and non-activated cells) promoted association of Gab2 with saponin-permeabilized cells, but reduced its association with the complexes after Thy-1 triggering. The PI3K activity associated with immunoprecipitated Gab2 was increased by latrunculin in both non-activated and Thy-1-activated cells.
To determine whether there are any changes in phosphatases associated with Gab2, we probed the Gab2 immunocomplexes by immunoblotting and in-gel phosphatase assay. Data presented in Fig. 8 (bottom) show that latrunculin alone significantly enhanced the amount of SHP-2 immunoprecipitated with Gab2, and also increased the activity of a phosphatase with molecular weight corresponding to SHP-2. Immunodepletion experiments confirmed that this was indeed SHP-2. The phosphatase activity associated with Gab2, namely SHP-2, was elevated in OX7-sensitized cells and was further enhanced after Thy-1 aggregation. In OX7-sensitized and latrunculin-pretreated cells, SHP-2 phosphatase activity was further enhanced, but declined slightly after Thy-1 triggering (Fig. 8). Gab2 immunocomplexes isolated from IgE-sensitized but non-activated cells also exhibited an increased amount of Gab2 and enhanced activity of PI3K and SHP-2 after exposure to latrunculin (not shown).
Data presented in this study indicate that pretreatment of RBL cells with latrunculin caused a significant decrease in the basal level of F-actin, and comparably inhibited actin polymerization induced by aggregation of both Thy-1 and FcϵRI. Inhibition of actin polymerization correlated with increased calcium and secretory responses not only in FcϵRI-activated cells, as has already been described 2, but also in Thy-1-activated cells, supporting thus the notion that these two activation pathways employ similar signaling modules 7, 19. The relatively low secretory and calcium responses in Thy-1-activated cells could be an outcome of more stringent regulation of Thy-1 by F-actin; when actin polymerization is inhibited and the amount of F-actin reduced, the differences between Thy-1 and FcϵRI activation pathways disappear.
The similarity between Thy-1- and FcϵRI-mediated activation pathways in latrunculin-pretreated cells could imply that FcϵRI is involved in both these activation pathways. However, our finding that a pretreatment with latrunculin followed by Thy-1 aggregation did not enhance the association of the receptor with GEM suggests that Thy-1 is not physically associated with FcϵRI under these conditions and that Thy-1 could initiate mast cell signaling in the absence of FcϵRI. This conclusion is supported by the results of our previous studies in which Thy-1-mediated activation of RBL cells was observed even in the absence of FcϵRI expression 9, 12, and by the finding of an increased Thy-1-mediated activation of latrunculin-pretreated FcϵRI-defective cells (this study). Interestingly, latrunculin alone induced a weak tyrosine phosphorylation of FcϵRI β subunit (Fig. 3C, see also 2), suggesting that microfilaments are involved in this process. Our finding that latrunculin pretreatment did not change the activity of Lyn kinase as determined by immunocomplex kinase assays is consistent with a key role of Lyn topography in initiation of cell signaling.
Latrunculin dramatically affected, although in an opposite manner, the distribution and tyrosine phosphorylation of two cytoskeletal proteins, paxillin and ezrin. Paxillin was poorly immunoprecipitated from control saponin-permeabilized cells, but its amount significantly increased after FcϵRI triggering (Fig. 6). It is remarkable that paxillin redistribution was observed not only in activated cells (with elevated F-actin levels), but also in cells pretreated with latrunculin alone (with reduced F-actin levels), a clear indication of a complex regulatory role of F-actin in this process. Increased binding of paxillin to activated and permeabilized cells could reflect an interaction of Lyn-SH2 domain with tyrosine-phosphorylated paxillin 20. It is tempting to speculate that this trans interaction could locally increase the Lyn kinase activity by blocking the cis interaction of Lyn-SH2 domain with Lyn C-terminal phosphotyrosine 21, and could thus contribute to enhanced tyrosine phosphorylation of FcϵRI and subsequent signaling molecules.
In contrast to paxillin, ezrin failed to exhibit any substantial changes in its association with large complexes in the course of Thy-1 or FcϵRI signaling, and this association was not affected by F-actin levels. On the other hand, tyrosine phosphorylation of ezrin dramatically increased in Thy-1-/FcϵRI-activated cells and was almost completely blocked by latrunculin. Ezrin, as well as other members of the ezrin/radixin/moesin (ERM) family of proteins, can mediate the anchoring of some transmembrane proteins to the actin skeleton. ERM proteins interact with the plasma membrane proteins through their N-terminal domain, and with the actin cytoskeleton through their C-terminal domain. The decreased phosphorylation of ezrin in latrunculin-pretreated cells suggests that ezrin is a negative regulator of Thy-1/FcϵRI signaling in mast cells. This conclusion is strengthened by the results of studies on B cell receptor activation documenting an increased tyrosine phosphorylation of ezrin accompanied by an inhibition of Syk tyrosine phosphorylation and calcium mobilization response 22.
Several lines of evidence presented in this study show that actin filaments contribute to spatial distribution and/or functional properties of other signaling molecules, including PI3K, Gab2, and SHP-2. PI3K catalyzes the synthesis of PIP3, which is involved in recruiting the molecules with pleckstrin homology to the plasma membrane. Importantly, the activity of PI3K associated with macromolecular complexes was dramatically increased in cells pretreated with latrunculin alone, reaching levels observed in Thy-1-/FcϵRI-activated cells. Although the activity of PI3K in latrunculin-pretreated cells was not further enhanced by Thy-1/FcϵRI triggering, the high initial activity levels may be sufficient to promote strong and rapid activation responses. It has been shown that the p85 subunit of PI3K from epithelial cells interacted with tyrosine-phoshorylated ezrin, and that this interaction increased the PI3K enzymatic activity 23. Tyrosine phosphorylation of ezrin is inhibited in latrunculin-pretreated cells (see above), and PI3K could therefore preferentially bind to another substrate, e.g. Gab2, an event which would presumably result in an enhanced PI3K activity. In fact, we have found that inhibition of actin polymerization led to an increased association of Gab2 with large signaling complexes and that these complexes were enriched of PI3K and SHP-2 (Fig. 8).
An important question is how the actin filaments regulate the activation via GPI-anchored proteins and FcϵRI. A previous study 3 showed that inhibition of actin polymerization stimulated FcϵRI-mediated tyrosine phosphorylation of several substrates, sustained for many tens of minutes; this implied that actin polymerization regulated the kinetics of Ag-induced tyrosine phosphorylation by modulating the lifetime of FcϵRI-Lyn interactions. In contrast, our work shows that inhibition of actin polymerization with latrunculin induced a formation of signaling complexes even in non-activated cells. This leads us to postulate that latrunculin-induced changes in topography and/or activity of signaling molecules before cell triggering are responsible, at least in part, for prompter and more extensive activation observed after Thy-1/TEC-21 or FcϵRI engagement. It should be noted that formation of signaling assemblies was observed not only in OX7- or IgE-sensitized cells, but also in nonsensitized cells; these findings exclude the possibility that latrunculin potentiated the formation of signaling complexes initiated by Thy-1 dimers or FcϵRIoligomers produced by IgE aggregates present in IgE preparations.
Adhesion of mast cells to the extracellular matrix is essential for their development and function in allergic responses, as well as for innate immunity 24. These interactions provide bi-directional signals resulting on one hand in the induction of mast cell proliferation and activation and, on the other hand, in the release of mediators from activated mast cells affecting the outcomes of inflammatory reactions. Experimental findings which are presented here and are complementary to previously published results (for references see Sect. 1), namely that cell adherence modulates the activation of RBL cells 25, that the engagement of surface receptors enhances actin polymerization, and that inhibition of actin polymerization increases the readiness of the cells to become activated, support the concept that actin filaments contribute to the setting of dynamic threshold for mast cell signaling. This notion is also supported by our unpublished data indicating that latrunculin enhances FcϵRI-mediated secretory and calcium responses in mouse bone marrow-derived mast cells.
4 Materials and methods
4.1 Ab and reagents
The origins of mAb specific for Thy-1.1 (0X7), Syk kinase, Lyn kinase, LAT adaptor, TEC-21, FcϵRI β subunit, and TNP-specific IgE (IGEL b4 1) have been described previously 8, 15. Polyclonal Ab specific for Syk, Lyn, LAT and IgE were prepared by immunizing rabbits with the corresponding recombinant proteins. Rabbit anti-IgE Ab was affinity-purified on Sepharose 4B with immobilized IGEL b4 1. Ab were biotinylated with EZ-LinkTM Sulfo-NHC-LC-Biotin (Pierce, Rockford, IL). Polyclonal Ab specific for Gab2, ezrin, PLCγ1, PLCγ2, and SHP-2 were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Ab specific for paxillin and phospho-Tyr (PY-20) conjugated to horseradish peroxidase (HRP), as well as HRP-conjugated anti-mouse and anti-rabbit IgG were purchased from Transduction Laboratories (Lexington, KY). Rabbit Ab specific for the p85 subunit of PI3K was obtained from Upstate Biotechnology (Lake Placid, NY). Fura-2-AM and streptavidin were purchased, respectively, from Molecular Probes (Eugene, OR) and Serva (Heidelberg, Germany). All other chemicals were obtained from Sigma-Aldrich (St. Louis, MO).
4.2 Cell stimulation, immunoprecipitation, and immunoblotting
RBL cells (subclone 2H3) and mutant cells (RBL-γ–c.1) defective in the surface expression of FcϵRI 9 were cultured as previously described 7. For activation, cells were harvested, resuspended in culture medium, and sensitized or not in suspension with IgE (IGEL b4 1; ascites ultracentrifuged at 100,000×g for 1 h; diluted 1:1,000), biotinylated anti-Thy-1.1 (3 μg/ml), or biotinylated anti-TEC-21 (3 μg/ml). After 30 min at 37°C, the cells were washed and resuspended to a concentration of 5×106/ml in buffered saline solution containing 20 mM Hepes pH 7.4, 135 mM NaCl, 5 mM KCl, 1.8 mM CaCl2, 5.6 mM glucose, 1 mM MgCl2, and 0.1% BSA, followed by incubation for 15 min in the presence or absence of latrunculin B (0.5 μM).
IgE-sensitized cells were then activated by exposure to TNP-BSA (1 μg/ml); the cells sensitized with biotinylated Ab were activated by exposure to streptavidin (10 μg/ml). After activation, the cells were solubilized and analyzed by SDS polyacrylamide gel electrophoresis (PAGE) 15. In some experiments the cells were incubated for 5 min on ice in PBS containing 0.1% saponin, 5 mM MgCl2, and 1 mM Na3VO4, and then extracted for 15 min on ice in lysis buffer supplemented with 1% Triton X-100. Immunoblots were quantified by Luminescent Image Analyzer LAS-3000 (Fuji Photo Film Co., Tokyo, Japan).
4.3 In-gel phosphatase assay
Anti-Gab2 precipitates were prepared from 107 cells. In-gel phosphatase assays were performed as described 26 with some modifications. Briefly, the substrate poly(Glu4Tyr)n was radiolabeled with [γ-32P]ATP (ICN, Irvine, CA) using recombinant human cSrc expressed as a glutathione S-transferase fusion protein in Escherichia coli, kindly provided by Dr. F. D. Böhmer (Jena, Germany). The substrate was incorporated into 10% SDS-PAGE (acrylamide/bisacrylamide 30:0.8). Following electrophoresis, the SDS was removed and proteins were renatured by incubation in 6 M guanidine hydrochloride followed by incubation in buffers containing 0.04% Tween-20 and 0.3% 2-mercaptoethanol. Final renaturation step was carried out in a buffer containing 50 mM Tris-HCl (pH 8.0), 0.3% 2-mercaptoethanol, 1 mM EDTA, and 0.04% Tween-20. Protein tyrosine phosphatase activities were detected in autoradiographs of dried gels as regions from which the 32P had been selectively removed. Quantification was performed by Fuji Bio-Imaging Analyzer Bas 5000 (Fuji Photo Film Co.).
4.4 Other methods
Details on sucrose density gradient ultracentrifugation and determination of PI3K activity and total amounts of IP3 and F-actin were previously described 15. [Ca2+]i were measured using Fura-2 fluorescence probe 27.
We thank H. Mrázová and R. Budovičová for excellent technical assistance. This work was supported by project LN00A026 (Center of Molecular and Cellular Immunology) from the Ministry of Education, Youth and Sports of the Czech Republic, grants 204/03/0594 and 301/03/0596 from the Grant Agency of the Czech Republic, grants A7052006 and A5052310 from the Grant Agency of the Academy of Sciences of the Czech Republic, and grant NB6758–3/01 from the Ministry of Health of the Czech Republic. The research of P. H. was supported in part by Research goal No. 002 from the Third Faculty of Medicine, Charles University, Prague, and the research of P. D. was supported by an International Research Scholar's Award from Howard Hughes Medical Institute.