Shp2 Activates Fyn and Ras to Regulate RBL-2H3 Mast Cell Activation following FcεRI Aggregation

The protein-tyrosine phosphatase (PTP) Shp2 has been implicated in many immunoreceptor signaling pathways, but its role in immunoreceptor FcεRI signaling, which leads to the activation of mast cells and blood basophils, is still largely undefined. Using Shp2 knockdown RBL-2H3 (RBL) mast cells, we here reported that Shp2 is required for the activation of RBL cells induced by FcεRI. FcεRΙ-evoked degranulation, calcium mobilization, and synthesis of cytokine transcripts (IL-1β, IL-10, and monocyte chemoattractant protein 1 (MCP-1)) were reduced in Shp2 knockdown RBL cells. Signaling regulatory mechanism investigation using immunoblotting, immunoprecipitation, and GST pull-down assay reveals that the down-regulation of Shp2 expression in RBL cells leads to decreased activities of Fyn, PLCγ, JNK, p38MAPK, and Ras/Erk1/2 after FcεRΙ aggregation. Further studies suggest that Paxillin phosphoryaltion was also impaired, but PAG phosphorylation was normal after FcεRΙ stimulation as a consequence of the inhibition of Shp2 expression in RBL cells. Collectively, our data strongly indicate that Shp2 is essential for the activation of RBL cells in response to FcεRΙ aggregation. Shp2 regulates this process through Fyn and Ras with no involvement of PAG. In addition, we identify Paxillin as an indirect substrate of Shp2 in FcεRΙ-initiated signaling of RBL cells.


Introduction
The activation of mast cells via aggregation of the receptor FceRI expressed on their surfaces is the central event in allergic diseases. Activated FceRI with high-affinity for immunoglobulin E (IgE) contributes to inflammatory symptomology by inducing the release of proinflammatory mediators (degranulation), and the production of various cytokines and chemokines. Like many other immunoreceptors, FceRI is made up of multiple chains, an IgEbinding section represented by the a subunit, and two signal transduction sections represented by the b and c subunits, both of which contain ITAMs in tails [1,2]. Upon the cross-linking of IgEbearing FceRI by polyvalent antigen, the b and c subunits are phosphorylated at ITAMs by the receptor-proximal Src family protein tyrosine kinases (SFKs) [3,4]. The Syk kinase is recruited to tyrosine-phosphorylated ITAMs and becomes activated to regulate its downstream targets in conjunction with the activated SFKs, such as Lyn, Fyn, and Hck [5][6][7][8][9]. A large number of downstream signaling relay molecules are involved in the following progress, including the linker for activated T cells (LAT) [10], phospholipase Cc (PLCc) [11], Grb2-associated binder 2 (Gab2) [12,13], Phosphatidylinositol 3 ' -kinase (PI3K) [12,14], Akt kinase and mitogen-activated protein kinases (MAPKs) [7,12,15]. Such tight assembly of these molecules together with the signal transduction among each other by means of phosphorylation and dephosphorylation eventually results in the activation of mast cells. This activation is characterized by enhanced calcium flux, prominent degranulation and increased gene expression activity [16,17]. Two acknowledged signaling pathways initiated by SFKs, including Lyn/Syk/LAT [5,6,10,18] and Fyn/PI3K [7,12,13], have been established to explain this process. However, the precise molecular mechanism involved in this progress is still not completely understood.
SFKs are important signaling relay molecules and the inhibitory C-terminal tyrosine (Tyr527 in mammalian Src) is their key regulatory site. When Tyr527 is phosphorylated by C-terminal Src kinase (Csk), the activities of SFKs are suppressed in resting-state. Csk is a cytoplasmic protein lacking a membrane targeting sequence. It can be recruited to the membrane where SFKs reside by binding of its SH2 domain to tyrosyl phosphorylated docking proteins, including transmembrane glycoprotein PAG/Cbp (PAG) and focal adhesion protein Paxillin, to exert its inhibitory effect on SFKs [19,20]. This inhibitory state of SFKs can be rescued by the phosphorylation of Tyr416 (in mammalian Src), the activating regulatory site in SFKs [20]. The activation of SFKs is essential for mast cell activation in response to FceRI engagement, but the signaling regulatory mechanism of SFK activation in FceRI signaling remains elusive. Shp2 is an intracellular PTP and expressed ubiquitously. Increasing numbers of reports suggest that a wide range of biochemical functions are Shp2-dependent, such as embryonic development for mice and Xenopus [21,22], Noonan syndrome, and Juvenile myelomonocytic leukemia [23][24][25]. Previous investigations have indicated that Shp2 is required for SFK activation in receptor-tyrosine kinase (RTK) signaling and integrin signaling [20,26]. One widely accepted signaling mechanism involved in this regulation suggests that Shp2 dephosphorylates PAG and/or Paxillin to resist the recruitment of Csk to the membrane, leading to the activation of SFKs [20]. Another proposed explanation for this regulation supports the notion that Shp2 might directly dephosphorylate Tyr527, which results in Tyr416 activation via its loop autophosphorylation and therefore leads to SFK activation [27]. Recently, physical interaction of Fyn kinase with Shp2 in Madin-Darby canine kidney cells is reported, which may offer the possibility for such direct regulation [28]. In addition, Shp2 has been documented to be essential for Ras-MAPK/Erk1/2 activation in response to various growth factors and cytokines through RTK signaling [27]. Nevertheless, whether Shp2 has participations in FceRI signaling to regulate the activation of SFK and Ras/Erk1/2 pathway remains to further be clarified.
Previous studies suggest a critical role of Shp2 in c-Kit signaling controlling proliferation and differentiation in mast cells [29,30]. Furthermore, the involvement of Shp2 in FceRI signaling has recently been reported [31]. However, little is known about how Shp2 regulates mast cell activation after FceRI aggregation. In this study, we inhibited the expression of Shp2 in the rat mucosal mast cell analog, RBL cell line, with Shp2-specific shRNAs (Shp2 knockdown) or scramble shRNA (control) expressed by retroviruses to clarify the role of Shp2 in mast cell signaling through FceRI.

Characterization of Shp2 Knockdown RBL Cells
To evaluate the role of Shp2 in FceRI-induced activation of RBL cells, we established Shp2 knockdown RBL cell lines. WT RBL cells were transduced with retroviral vectors expressing shRNAs directed against Shp2 for knockdown effect (sh-Shp2-#1 and sh-Shp2-#2) or scramble shRNA for control (sh-Ctrl). Three populations of puromycin-resistant RBL cells (sh-Ctrl, sh-Shp2-#1, and sh-Shp2-#2) were collected and assayed for Shp2 expression by immunoblotting ( Figure 1A). Expression of sh-Shp2-#1 and sh-Shp2-#2 cells led to reductions in Shp2 expression of 66.3% 63.76%, and 88.2% 60.35%, respectively, in RBL cells relative to scramble shRNA. On the contrary, expression of Shp1, homologous preotein of Shp2, was unchanged in Shp2 knockdown RBL cells ( Figure S1). This indicates an effective and specific inhibition of Shp2 expression in RBL cells. Flow cytometry analysis showed that RBL cells from these three populations have similar surface expression of IgE receptor ( Figure 1B), which suggests that the knockdown of Shp2 expression did not impair FceRI expression.

Shp2 is Required for FceRI-Mediated RBL Cell Activation
To assess the functional role of Shp2 in RBL cell degranulation and cytokine production after FceRI aggregation, we investigated the consequences of the knockdown of Shp2 expression. Three populations of RBL cells (sh-Ctrl, sh-Shp2-#1, and sh-Shp2-#2) were stimulated with different concentrations of the antigen DNP-BSA (DNP) and degranulation was determined by the release of bhexosaminidase ( Figure 2A). We observed that degranulation levels in Shp2 knockdown RBL cells (3.54% 60.44% for sh-Shp2-#1, and 2.83% 60.50% for sh-Shp2-#2) were significantly lower than in controls (6.39% 60.31%; values correspond to 10 ng/ml DNP stimulation, the concentration of optical stimulation). In contrast, degranulation induced by 12-O-tetradecanoylphorbol-13-acetate (PMA) plus ionomycin remained at a parallel level among three populations of RBL cells ( Figure 2B). This implies that the secretion machinery is normal in Shp2 knockdown RBL cells and the defective degranulation is attributable to the inhibition of Shp2 expression. Besides, reductions in the degranulation in sh-Shp2-#2 population were more pronounced than in sh-Shp2-#1 population, which corresponded to the extent of the inhibition of Shp2 expression revealed by immunoblotting in Figure 1A. Transcripts of cytokines were subsequently determined by real-time RT-PCR. A significant decrease of the generation of IL-1b ( Figure 2C), IL-10 ( Figure 2D), and MCP-1 ( Figure 2E) transcripts after FceRI stimulation was observed in Shp2 knockdown RBL cells compared with control cells.
Given that mast cell degranulation is calcium-dependent, we examined the contribution of Shp2 to FceRI-evoked calcium mobilization. The calcium flux of RBL cells from the three populations was monitored over time following DNP stimulation ( Figure 3). Similar to degranulation, calcium response was markedly inhibited in Shp2 knockdown RBL cells compared with control cells. Collectively, our data strongly indicate that Shp2 is required for FceRI-mediated RBL cell activation.

SFK Activation is Impaired in Shp2 Knockdown RBL Cells
We next explored the FceRI-evoked signaling events caused by the inhibition of Shp2 expression. To evaluate the global effect of Shp2 knockdown on FceRI signaling, we determined the overall level of intracellular tyrosyl phosphorylation by immunoblotting ( Figure 4A). A marked reduction in the amount of tyrosinephosphorylated proteins was observed in the lysates from FceRIstimulated Shp2 knockdown RBL cells compared with control cells, which accounts for the positive role of Shp2 in FceRI-evoked RBL cell activation. For a more specified molecular mechanism, lysates from the three activated populations of RBL cells were subjected to immunoblotting with antibodies against the phospho (p)-Src (activating site Try416 and inhibitory site Tyr527). As showed in Figure 4, the phosphorylation of Tyr416 required for mast cell activation was largely inhibited in Shp2 knockdown RBL cells compared with control cells. In contrast, Tyr527 phosphorylation was even slightly enhanced in Shp2 knockdown RBL cells. The data suggest that Shp2 has a role in regulating FceRI-evoked SFK activation in RBL cells and the inhibitory C-terminal tyrosine in SFKs seems to be one of potential target sites for Shp2.

Fyn Kinase Activation is Defective in Shp2 Knockdown RBL Cells
Among SFKs, Fyn and Lyn kinases are two important members for mast cell activation. To determine whether the activation of Lyn and Fyn is Shp2-dependent, we analyzed the kinetics of their phosphorylation after FceRI aggregation in the three populations of RBL cells by immunoprecipitation with Lyn ( Figure 5A) and Fyn ( Figure 5B-C). Similar levels of Lyn tyrosyl phosphorylation (Tyr396, the activating site) were detected in the three populations. On the contrary, Fyn activation was remarkably decreased in Shp2 knockdown populations compared with control population. The control population showed robust tyrosyl phosphorylation (Tyr417, the activating site) of Fyn kinase, which peaked at 5 min and decreased at 15 min after stimulation. However, only weak phosphorylation signals were detected after FceRI aggregation in Shp2 knockdown populations, even at the time of peak stimulation. Furthermore, the phosphorylation of both PLCc1 ( Figure S2) and PLCc2 ( Figure 5D), two signaling relay molecules downstream of Fyn, was also defective in Shp2 knockdown populations. Of note, knockdown of Shp2 expression did not affect the expression of Lyn, Fyn, or PLCc proteins. Therefore, our results demonstrate that Shp2 regulates FceRI-evoked RBL cell activation via Fyn kinase, but not Lyn kinase.
Additionally, Gab2 tyrosyl phosphorylation was hardly affected in Shp2 knockdown populations compared with control population ( Figure 5E). This indicates that Gab2 phosphorylation is almost Fyn-independent in response to FceRI stimulation in RBL cells.

Shp2 Regulates Fyn Activation Independent of PAG
The inhibition of Shp2 expression resulted in markedly diminished Fyn activation after FceRI aggregation ( Figure 5B-C). This implicates that Shp2 can promote dephosphorylation of the inhibitory C-terminal of Fyn in some way. To determine whether Shp2 could dephosphorylate PAG to decrease the association of Csk and thus promote Fyn activation after FceRI engagement, we monitored the dynamic phosphorylation of PAG following antigen stimulation in three populations of RBL cells. The association of Csk with PAG in quiescent WT RBL cells was first confirmed ( Figure 6A), which is consistent with the previous report [32]. Next, immunoprecipitation was performed on lysates from the three activated populations of RBL cells. Immunoprecipitation of PAG followed by immunoblotting with the antibody against phosphotyrosine (4G10) showed that the levels of PAG tyrosyl phosphorylation ( Figure 6B, upper lane), and its association with Csk ( Figure 6B, third lane), in Shp2 knockdown populations were equivalent to that in control population. Similarly, Csk immunoprecipitates displayed comparable overall intensity of tyrosyl phosphorlation of both PAG and Csk ( Figure 6C). These data clearly demonstrate that PAG is not involved in the regulation of Shp2-dependent Fyn activation in FceRI signaling. When we analyzed the phosphorylation of Paxillin, another Csk-binding protein and potential substrate of Shp2, we found that the tyrosyl phosphorylation levels of Paxillin were elevated at 5 min of stimulation, and markedly enhanced at 15 min after stimulation in control population, whereas Shp2 knockdown populations showed significantly inhibited Paxillin tyrosyl phosphorylation after stimulation ( Figure 6D). This indicates that Shp2 promotes, rather than inhibits, Paxillin phosphorylation in FceRI signaling.

Ras and MAPKs Activities are Inhibited in Shp2 Knockdown RBL Cells
The activation of MAPKs is a remarkable event after FceRI aggregation responsible for production of many types of cytokines  [33,34]. To understand whether Shp2 has a role in FceRI-evoked MAPK activation, we determined the phosphorylation of MAPKs, including Erk1/2, JNK, and p38, in the three populations of RBL cells by immunoblotting ( Figure 7A). As indicated by a marked decrease in the phosphorylation levels of MAPKs in Shp2 knockdown populations compared with control population, our data suggest that Shp2 is required for optimal MAPK activation in response to FceRI aggregation. To investigate the relationship between Shp2 and Ras, an upstream molecule of Erk1/2, in FceRI signaling, GST-RBD (Ras binding domain in Raf) pull down assay was carried out to compare the amount of active Ras (in the form of GTP-bound) after FceRI stimulation in the three populations of RBL cells ( Figure 7B). The amount of Ras-GTP at 5 min after stimulation was reduced by 40-70% in Shp2 knockdown populations compared with that in control population. Therefore, our results suggest that Shp2 positively regulates Ras/ Erk1/2 signaling as well as the activation of JNK and p38MAPK in response to FceRI stimulation in RBL cells.

Discussion
In this work, we studied the role of Shp2 in FceRI signaling utilizing Shp2-specific shRNAs expressed by retroviruses to inhibit Shp2 expression in RBL cells. Although previous reports have demonstrated a crucial contribution of Shp2 in development [35], proliferation [29,[36][37][38], and tumor suppression [39][40][41], little is known about the role of Shp2 in mast cell-dependent allergic reactions. Our results demonstrate an essential role of Shp2 in FceRI-mediated RBL cell activation. This conclusion is based on observation of inhibited degranulation (Figure 2A), synthesis of cytokine mRNAs (IL-1b, IL-10, and MCP-1; Figure 2C-E), and calcium flux (Figure 3) in Shp2 knockdown RBL cells after FceRI aggregation. Furthermore, our examination reveals that Shp2 could activate Fyn and Ras to promote FceRI-evoked activation of RBL cells. These results contribute to our understanding of the signaling regulatory mechanism involved in the activation of RBL cells after FceRI aggregation and raise the possibility of targeting Shp2 as a new therapeutic strategy for inflammatory diseases.  In mast cells, FceRI aggregation activates a complicate signal transduction cascade, the major pathway of which is Lyn/Syk/ LAT signaling. In the absence of either Syk or LAT, mast cells show impaired degranulation, cytokine generation, and PLCc phoshprylation in response to FceRI stimulation [5,10]. These defects are similar to those in the inhibition of Shp2 expression (Figure 2, 5D). This conformity prompted us to examine the activity of Lyn kinase in Shp2 knockdown RBL cells after FceRI cross-linking. Unexpectedly, Lyn kept showed continued high phosphorylation levels after antigen stimulation in both control and Shp2 knockdown RBL cells, without significant difference between the two ( Figure 5A). Furthermore, the synthesis of MCP-1 mRNA induced by FceRI aggregation was markedly inhibited in both Shp2 knockdown RBL cells ( Figure 2E) and Fyn-deficient BMMCs [42], but was unaffected in LAT-deficient BMMCs [10]. These data indicate that Shp2 is probably involved in Fyndependent signaling cascade, which is independent of Lyn but has some cross-talk with Syk/LAT signaling. Taking that Fyn is essential for the optimal activation of Syk after FceRI engagement into consideration [13], Fyn/PI3K signaling, the pathway initiated by Fyn kinase is an attractive candidate for such regulation [12,13]. In fact, FceRI-induced Fyn activation was largely diminished on account of the inhibition of Shp2 expression in RBL cells ( Figure 5B-C). Moreover, FceRI-evoked degranulation and phosphorylation of PLCc, JNK, and p38MAPK were all reduced in Shp2 knockdown RBL cells (Figure 2A, 5D, and 7A), similar to that reported in Fyn-deficient BMMCs [7,13,42]. We thus propose that Shp2 promotes FceRI-evoked RBL cell activation via Fyn/PI3K/PLCc signaling, in which Syk may function as a mediator in the transmission of signals from FceRI/ Shp2/Fyn to PI3K. PI3K is important for the phosphorylation of PLCc, which is able to cleave phosphatidylinositol-4,5-bisphosphate (PIP2) into inositol 1,4,5-trisphosphate (IP3) and diacylglycerol (DAG), leading to the elevation of intracellular calcium concentration with a contribution to degranulation [12,43]. In this way, our proposed regulatory model of Shp2 explains the defective degranulation ( Figure 2A) and calcium mobilization (Figure 3).
The fact that Gab2 phosphorylation was largely unaffected in Shp2 knockdown RBL cells ( Figure 5E) indicates a loose interaction of Gab2 with Fyn in FceRI signaling. This is consistent with our previous observation that FceRI-mediated phosphorylation of Gab2 was severely impaired in Lyn-deficient BMMCs, but almost normal in Fyn-deficient BMMCs [13], which suggests a requirement of Lyn to a greater extent than that of Fyn for Gab2 phosphorylation in FceRI signaling. In agreement with this suggestion, FceRI-evoked phosphorylation of both Lyn and Gab2 was hardly affected in Shp2 knockdown RBL cells ( Figure 5A and 5E). Nevertheless, more experiments are necessary to be carried out to delineate the regulatory signaling mechanism of Gab2 phosphorylaiton in response to FceRI activation. The most important function for a PTP is to dephosphorylate its specific targets leading to inactivation of its targets [44]. Given that Shp2, as a PTP, functions positively in FceRI-evoked RBL cell activation, we assumed that, as reported in RTK signaling [20], the phosphorylation levels of Shp2 substrates: PAG and Paxillin would elevate after FceRI stimulation in Shp2 knockdown RBL cells compared with control cells. However, as our results show, although Shp2 expression was inhibited, the phosphorlation levels of PAG after FceRI stimulation were not increased compared with that of control cells ( Figure 6B, upper lane); nor was enhanced association of PAG with Csk observed (Figure 6B, third lane). This suggests that Shp2-mediated signaling downstream of the FceRI might require no involvement of PAG. Phosphorylation levels of Paxillin were decreased, but not elevated in Shp2 knockdown RBL cells ( Figure 6D), indicating that Paxillin lays downstream of Shp2 acting as an indirect, rather than a direct substrate of Shp2 in FceRI signaling. This aids us to exclude the possibility that Shp2 regulates FceRI-evoked Fyn activation by dephosphorylating Paxillin to reduce the binding of Csk. Based on these results, it is possible that Shp2 promotes FceRI-induced Fyn activation in RBL cells by directly dephosphorylating the inhibitory C-terminal of Fyn kinase. Further investigations are required to confirm this speculation. Besides, reduced Paxillin tyrosyl phosphorylation due to the inhibition of Shp2 expression in FceRI signaling is similar to that in integrin signaling. The marked response of Paxillin phosphorylation delayed at 15 min after DNP stimulation was also consistent with that in response to integrin stimulation [26]. These findings have implications for considering the potential role of Shp2 in contributing to RBL mast cell adhesion and migration, which are of great importance in allergic reactions.
It should be noted that both Lyn and PAG phosphorylation were not significantly enhanced in response to antigen sitmulaiton in our results. However, it does not mean that their phosphorylation cannot be induced in FceRI signaling. In fact, it has been documented that Lyn phosphorylation is elevated after FceRI aggregation. Once antigen stimulation, the aggregated FceRI is supposed to associate with more Lyn kinases and activate them, thereby amplifying signals. But as the most upstream molecule in FceRI signaling, Lyn completed this process rapidly, always less than 2 min [32]. We guess that our time for stimulation may be too long (5 min and 15 min) to capture the robust change of Lyn phosphorylation, leading no difference observed before and after stimulation in our results. Anyway, given that Lyn is constitutively associated with the receptor even in quiescent RBL cells, Lyn phosphorylation would be affected even in unstimulated mast cells if Shp2 has effects on Lyn activation, i.e. Shp2 knockdown RBL mast cells should show different phosphorylation levels from that of controls. Therefore, our stimulation times did not prevent us from assessing the role of Shp2 in Lyn activation. Similarly, it is also applies for PAG phosphorylation.
In addition, it seems that Shp2 is required for the activation of MAPKs after FceRI engagement in RBL cells ( Figure 7A). Defective gene expression in Shp2 knockdown RBL cells ( Figure 2C-E) further supports its essential role in FceRI-induced MAPK activation [33]. Impaired Fyn activation may account for the reduced activities of JNK and p38MAPK [42], whereas largely decreased Erk1/2 activity is due to the defective activation of Ras [27,45,46]. Importantly, reduced Ras activity in Shp2 knockdown RBL cells ( Figure 7B) facilitates us to place Shp2 upstream of Ras in FceRI signaling. This suggests a positive role of Shp2 in Ras/ Erk1/2 signaling downstream of the IgE receptor in RBL cells, which is consistent with previous studies conducted in other cells or receptor signaling (e.g. RTK signaling) [27,36]. However, when we focused on the question how Shp2 causes the activation of Ras, we found that there exists a matter of debate. The involved mechanisms are often different across cases, depending on cell types and/or stimulation factors [46]. As a common explanation in this community, Shp2 is supposed to dephosphorylate RasGAPbinding sites on receptors [47] or sprouty proteins [48,49], which negatively regulates Ras activation. As we did not detect sprouty2 expression in RBL cells, it may be an attractive field to further explore whether it has RasGAP-binding sites on FceRI and whether Shp2 could promote Ras activation by dephosphorylating these sites in mast cells.
We did not determine the relationship between Fyn and Ras, or Fyn and Paxillin. More efforts must be made to clarify the relationships among these signaling relay molecules. Actually, there have been no reports demonstrating a direct interaction between Fyn and Ras or Paxillin to date. Our study therefore suggests that Shp2 regulates FceRI-evoked RBL cell activation via Fyn kinase and Ras/Erk1/2 signaling (Figure 8).

Antibodies and Reagents
Rabbit antibodies to Shp2, Lyn, Ras, Erk1/2, p38, JNK, Akt, Src, PLCc1, phospho-Src (Tyr416 and 527), phospho-Paxillin (Tyr118), phospho-Erk1/2, phospho-JNK, phospho-p38 and phospho-Akt (Ser473) were from Cell Signaling Technology (Danvers, MA, USA). Rabbit antibody to Erk2, PLCc2, Shp1, and HRP-conjugated antibodies to mouse and rabbit immuno- To control for loading, blots were stripped and reprobed with anti-total Erk1/2, JNK, and p38 antibodies. Shp2 expression was also determined by immunoblotting with anti-Shp2 antibody followed by reprobing with anti-GAPDH antibody. Representative blots of at least 3 experiments are shown. (B) Upper panels, RBL cells from the three populations were sensitized, starved, and stimulated with 10 ng/ml DNP for indicated times and WCL were prepared. Active Ras (Ras-GTP) amounts in WCL were determined by incubating the WCL with GST-RBD, followed by immunoblotting with anti-Ras antibody according to Materials and Methods. 5-10% of WCL of each sample was loaded as a control for the total Ras in RBL cells. Lower panel, activated Ras was measured by quantifying the Ras-GTP levels at 5 min after stimulation normalized to total Ras. Data are expressed as mean 6 SEM and representative results from at least three experiments are shown. **, P,0.01, Student t test. doi:10.1371/journal.pone.0040566.g007 Figure 8. Model of Shp2 action in FceRI signaling. Upon FceRI stimulation, Fyn kinase and Ras become activated in a Shp2-dependent manner. Activated Fyn transmits positive signals to PI3K, leading to PI3K activation that accounts for calcium mobilization as well as for Akt phosphoryaltion. MAPKs also become phosphorylated following Ras and Fyn activation that are responsible for production of most cytokines. Paxillin is phosphoryated indirectly by Shp2 upon FceRI stimulation. The direct substrates of Shp2 in FceRI signaling remain to be defined. Another acknowledged signaling pathway Lyn/Syk/LAT is also shown. doi:10.1371/journal.pone.0040566.g008 globulin were from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Mouse monoclonal antibodies to Fyn and Csk-binding protein (PAG) were from Pierce Biotechnology (Rockford, IL, USA). Mouse monoclonal antibody to Csk and Paxillin were from BD Biosciences (San Jose, CA, USA). Mouse monoclonal antibody to phosphotyrosine (4G10) was from Millipore (Bedford, Mass, USA). Monoclonal anti-dinitrophenyl (DNP)-IgE and DNPbovine serum albumin (BSA) were from Sigma-Aldrich (St. Louis, MO, USA). FITC-conjugated monoclonal rat anti-mouse IgE (IgG1) and rat monoclonal IgG1were from from BD Pharmingen (San Jose, CA, USA). Leupeptin, aprotinin, pepstatin, antipain, bglycerophosphate, sodium fluoride (NaF), Sodium orthovanadate (Na 3 VO 4 ), and phenylmethylsulfonylfluoride (PMSF) were from Sigma-Aldrich.

Retroviral Infection of RBL Cells
pSuper-puro retroviral plasmids were transfected into Plate-E cells using Lipofectamine 2000 (Invitrogen, Carlsbad, CA,USA). Viral particle-containing culture supernatants were collected after 24, 36, and 48 h, respectively. The supernatants were filtered through 0.45 mm filters and the viral supernatant was used to infect RBL cells after addition of 4 mg/ml polybrene (Sigma-Aldrich). Infected RBL cells were then incubated at 37uC for 12 h. After 12 h, the infected cells were selected in the presence of 3 mg/ ml puromycin for 5-7 days. Mass populations of puromycinresistant RBL cells (sh-Ctrl, sh-Shp2-#1, and sh-Shp-#2) were used for experiments.

Flow Cytometry Analysis (FACS)
Sensitized RBL cells were stained for FceRI expression with FITC-rat-anti-mouse IgE monoclonal antibody (IgG1) or FITCisotype control (rat monoclonal IgG1) antibody for 1 h on ice, followed by flow cytometry analysis using a FACScan (Becton Dickinson FACStar).

Immunoprecipitation and Immunoblotting
Sensitized RBL cells were starved in MEM containing 2% BSA (Amersco, Solon, Ohio, USA) overnight, and either stimulated with DNP (10 ng/ml) or left unstimulated for the indicated times. Cells were lysed for 40 min at 4uC with NP-40 lysis buffer (50 mM Tris-HCl [pH 7.5], 150 mM NaCl, 0.1% NP-40, 5 mM EDTA, 5 mM EGTA, 15 mM MgCl 2 , 60 mM b-glycerophosphate, 10 mM NaF, 2 mM Na3VO4, 200 mM PMSF, 4 mg/ml aprotinin, 20 mg/ml leupeptin, 2 mg/ml antipain, and 2 mg/ml pepstatin). The lysates (1 mg of total protein) were precleared by incubating together with 20 ml of protein G-Sepharose (Roche Applied Science, Mannheim, Germany) for 3 h at 4uC. Precleared lysates were centrifuged and transferred into new tubes followed by incubating with the indicated antibodies and 30 ml of protein G-sepharose overnight at 4uC. Immunoprecipitates were resolved by SDS-PAGE, transferred to polyvinylidene difluoride (PVDF) membranes (Roche Applied Science), and probed with the appropriate primary antibodies followed by secondary goat antimouse or anti-rabbit antibodies. For immunoblotting, whole cell lysates (WCL) were mixed with 5 6 sample buffer after determination of protein concentrations using a bicinchoninic acid (BCA) protein assay kit (Pierce Biotechnology). They were fractionated onto SDS-PAGE as described above.

GST-Pull Down Assay
Ras-GTP pull down assay was performed as described [52]. Briefly, the GST-RBD (the Ras binding domain in Raf) fusion protein was produced in Escherichia coli DH5a-competent bacteria and purified by glutathione agarose beads (Sigma-Aldrich) on the day of the assay. Cells were lysed with NP-40 buffer as described in the Immunoprecipitation and Immunoblotting section. The lysates were incubated with GST-RBD at 4uC for 1 h. The complex was washed, resolved using SDS-PAGE and immunoblotted with Ras antibody.

b-hexosaminidase Assay
Degranulation assay was performed as described [53] with the following modifications. RBL cells were stimulated with 100 ml of DNP or PMA plus ionomycin, or left unstimulated for 1 h at 37uC. 100 ml of specific substrate for b-hexosaminidase 4methylumbelliferyl-N-acetyl-b-D-glucosaminide (2.4 mM, Sigma-Aldrich) was added and incubated for another 30 min at 37uC. Aliquots (50-100 ml) of the supernatant were transferred into a new 96-well tissue culture plate and analyzed using a fluorescence microplate reader (Tecan Inc.) as described [53].

Measurement of Calcium Flux
Sensitized RBL cells were stained with 4 mM Fluo3-AM (Dojindo, Japan) in Tyrode's buffer (135 mM NaCl, 5 mM KCl, 1.8 mM CaCl 2 , 1.0 mM MgCl 2 , 5.6 mM glucose, 20 mM HEPES, and 1 mg/mL BSA at pH 7.4) for at least 45 min at 37uC. The stained cells were challenged with DNP (10 ng/ml) for the indicated lengths of time. Single wavelengths was collected at 520 nm (488 nm excitation) over time and monitored using a laser scanning confocal microscope (Leica Microsystems Heidelberg GmbH, Germany). Changes in [Ca 2+ ] were reflected by varied fluorescence intensity of Ca 2+ -bound Fluo-3. Calcium flux-time curves were captured and analyzed using the ''LAS AF Lite'' software package (Leica), and the values were measured after subtracting the background fluorescence.

Real-Time RT-PCR
Sensitized RBL cells were stimulated with DNP (10 ng/ml), or left unstimulated, in Tyrode's buffer for 1 h at 37uC. Total RNA was isolated using the TRI reagent (Invitrogen). First-strand complementary DNA (cDNA) was obtained as described [10]. The synthesized cDNA was subjected to real-time RT-PCR conducted by an ANTIBODYI 7300 real-time PCR instrument (Applied Biosystems) using a SYBR green-based real-time PCR kit (Takala, Japan) according to the manufacture's instruction. The primers used were as follows: MCP Indicated populations of RBL cells were sensitized, starved, and stimulated with 10 ng/ml DNP for indicated times. Lysates were immunoprecipitated with PLCc1 antibody followed by immunoblotting with 4G10 antibody. Representative blots of three independent experiments are shown. (TIF)