INTRODUCTION

Phosphorylation on tyrosine, serine, or threonine residues is a key regulatory mechanism in mammalian cells used to regulate the mitogenic or oncogenic potential of proteins by augmenting their enzymatic activity or modifying their association with other signal transducers (1). In this manner, Raf-1 activity also appears to be regulated by phosphorylation of key serine/threonine and tyrosine residues in that kinase activity correlates with the phosphorylation state of the Raf-1 protein (2). Raf-1 is rapidly phosphorylated and activated following stimulation with growth factors and mitogens (3). Kovacina et al. (4) demonstrated that treatment with phosphatase remarkably decreased the catalytic activity of Raf-1 in insulin-stimulated cells. While Raf-1 is exclusively phosphorylated on serine residue in resting and mitogen-stimulated cells, it has been demonstrated that phosphorylation of Raf-1 increased on threonine and tyrosine residues after IL-2 or TCR stimulation (5). However, the stoichiometry of tyrosine phosphorylation is low compared to the extent of serine/threonine phosphorylation. Ser⁴³, Ser²⁵⁹, and Ser⁶²¹ are known to be major in vivo phosphorylation sites in mammalian and Sf9 insect cells expressing human Raf-1 proteins. Phosphorylation of both Ser²⁵⁹ and Ser⁶²¹ modify the catalytic activity of Raf-1 (6).

Although tyrosine phosphorylation of Raf-1 has been demonstrated in several systems, including IL-3-¹ and granulocyte/macrophage colony-stimulating factor-stimulated murine myeloid cells (7), IL-2-treated murine T cells (8), platelet-derived growth factor-stimulated or v-src-transformed murine fibroblasts (2, 3), the biological relevance of this phosphorylation has not yet been clarified. Fabien et al. (9) recently suggested an importance for the tyrosine phosphorylation in regulating the biological activity of Raf-1 and identified Tyr³⁴⁰ and Tyr³⁴¹ as major tyrosine phosphorylation sites of Raf-1. They demonstrated the phosphorylation of Tyr³⁴⁰ and Tyr³⁴¹ by coexpressing Raf-1 with the activated tyrosine kinase, pp60^v-src, in baculovirus Sf9 cells (9). More recently the data of Pulmiglia et al. (10) has defined in more detail the NH₂ terminus of Raf-1, demonstrating that in certain systems the tyrosine phosphorylation of Raf-1 alone is sufficient to activate Raf-1 kinase activity. In these experiments, mutation of residues 53-156 required for Ras-Raf-1 binding abrogated activation by Ras but had no effect on activation of Raf-1 by activated Src in Sf9 cell system. These and other data suggest that several independent mechanisms may exist for the regulation of Raf-1. This is a paradigm which may play itself out more than once in mammalian signal relay.

Raf-1, the proto-oncogene product of the c-raf-1 gene which is the cellular homologue of the murine transforming gene v-raf, is a 72-76-kDa phosphoprotein with intrinsic kinase activity for serine and threonine residues (2). Raf-1 is an effector of Ras and is one of the activators of mitogen-activated protein kinase kinase (MEK) (11). Sequence analysis suggests that Raf family proteins have three unique conserved domains, named conserved region 1 (CR1), CR2, and CR3 (12). CR1 is a cysteine-rich residue having a putative zinc binding region. CR2 is a serine/threonine-rich region, and CR3 contains the protein kinase domain. Both CR1 and CR2 are located in the amino-terminal half of the Raf-1, which appears to regulate the catalytic activity of carboxyl-terminal kinase domain. The v-Raf protein of murine sarcoma virus 3611 is observed to have a deletion of the amino-terminal half of the protein. Deletion or mutation of the amino terminus activates the oncogenic transforming potential of Raf-1 (13, 14). In protein kinase cascades, Raf-1 appears to be a central intermediate in the transmission of proliferative, developmental, and oncogenic signals by mediating signals from receptor or nonreceptor tyrosine kinases, from p21^ras to serine/threonine kinases, including MAP kinase kinase, MEK, MAP kinase, or ribosomal S6 kinase (RSK) ultimately leading to activation of transcriptional factor, such as NF-B/Rel, in the nucleus (15, 16, 17). The role of Raf-1 in post-mitotic cells is less clear.

Proteins of the FcR family have a number of conserved biological characteristics of multisubunit Ig supergene family (18, 19). FcRs, receptors for the Fc portion of IgG, are composed of three groups including FcRI (CD64), FcRII (CD32), or FcRIII (CD16) according to their binding affinity for the ligand. FcRI, found in monocytes and macrophages, is a 74-kDa glycoprotein that binds monomeric IgG with high affinity (20, 21). The FcRI receptor signaling via a conserved sequence of amino acids termed the immunoreceptor tyrosine-based motif (ITAM) (22). Signaling through the ITAM shares a number of conserved features among the Ig gene superfamily of multisubunit receptors (23). We and others have reported that the FcRI receptor stimulation results in the sequential activation of FcRI, Hck, Syk and MAP kinase (24). The FcRI receptor is also linked to the cytoskeleton and is involved in a number of well characterized cell biologic signals (activation of respiratory burst, phagocytosis, or cell motility, etc.). Importantly, the myeloid respiratory burst is a well characterized response known to be regulated by small GTPases, Rac1, Rac2, and Rap1a (25). It is tempting to speculate that the respiratory burst may be also regulated by the known effectors of these GTPases (i.e. Raf-1 or PAK65, etc.) (26).

Cross-linking of FcR induces activation events, including tyrosine phosphorylation of  subunit (27) and activation of phospholipase C-1 and 2 (28), increases phosphatidylinositol hydrolysis and calcium mobilization (29), production of cytokines (30), and generation of superoxide anions (31). Our laboratory uses the myeloid cell line U937 differentiated in IFN (termed U937IF cells) to study FcRI signal transduction. We are interested in the mechanism by which signals are transmitted from FcRI receptor to the respiratory burst. We previously reported that the stimulation of FcRI receptor in U937IF cells results in tyrosine phosphorylation and activation of Syk (32). More recently we demonstrated that FcRI cross-linking activated Src family kinase, Hck and a mobility shift of MAP kinase (33). These results lead us to hypothesize a role for Raf-1 in FcRI signal relay. Herein, we demonstrate that Raf-1 is tyrosine-phosphorylated after FcRI stimulation (10-fold increase). Both Hck and Syk are activated following FcRI stimulation, making them good candidate kinases for the tyrosine phosphorylation of Raf-1. We also observe the tyrosine phosphorylation of Shc and the association of Shc and Grb2 in U937IF cells activated by FcRI stimulation (not FcRII or FcRIII activation). Our results suggest that Raf-1 is major substrate for protein tyrosine kinases following FcRI cross-linking, which results the sequential activation of Shc, Grb2, Ras, Raf-1, and MAP kinases transmitting FcRI signals that result in the assembly of an active respiratory burst complex.

EXPERIMENTAL PROCEDURES

The FcR-specific antibodies were obtained from Medarex Inc. (West Lebanon, NH). The mAb 197 and mAb 32.2 are specific for the FcRI subunit, mAb 32.2 is a F(ab)₂ fragment of IgG, mAb IV.3 is specific for FcRII subunit, and mAb 3G8 is specific for FcRIII subunit of Fc receptor for IgG. The cross-linking antibody was a rabbit anti-mouse F(ab)₂ fragment (RM) purchased from Organon Teknika Corp. (West Chester, PA). Anti-Raf-1 antibody was purchased from Santa Cruz Biotechnology Inc. (Santa Cruz, CA). Anti-phosphotyrosine (anti-Tyr(P)) and anti-Shc antibodies were purchased from Upstate Biotechnology, Inc. (Lake Placid, NY), and anti-Grb2 antibody was obtained from Transduction Laboratories (Lexington, KY). The anti-MAP kinase antiserum (polyclonal antibody 1913.2) against the peptide KELIFEETARFQPGY, corresponding to the extreme COOH terminus of the Xenopus Erk2, was provided by Jonathan A. Cooper, Fred Hutchinson Cancer Center (Seattle, WA). This region of the Xenopus MAP kinase is 100% conserved with human Erk2 and 85% conserved in human Erk1 (34).

U937 cells were maintained in RPMI 1640 with 10% fetal calf serum and differentiated with 250 units/ml human recombinant IFN (kindly provided by Genentech Corp., San Francisco, CA) for 4 days. U937IF cells were cultured at a concentration of 5 × 10⁵ cells/ml, and the medium was replenished with fresh IFN (250 units/ml) every 2 days, as described (35). Flow cytometric analysis of U937IF cells demonstrated equal expression of FcRI, FcRII, and FcRIII (data not shown). For stimulation of FcR receptors of U937IF cells, cells were washed twice in cold HBSS and adjusted to a concentration of 4 × 10⁷ cells/ml. Cells in a 0.5-ml volume were incubated on ice for 30 min with anti-FcR antibodies (0.25 µg/sample). We then added 10 µg/ml RM (F(ab)₂ fragment) antibody at 37 °C for different times. Stimulated cells were cooled rapidly with cold HBSS and centrifuged at 500 × g for 5 min in a cold centrifuge. A cell pellet was lysed with 800 µl of Triton X-100 extraction buffer (EB buffer) on ice for 30 min or resuspended in 25 µl of 1 × sample buffer/2 × 10⁶ cells for whole cell lysates.

Cell lysates were prepared in a lysis buffer (EB buffer) containing 1% Triton X-100, 10 mM Tris, pH 7.6, 50 mM NaCl, 0.1% bovine serum albumin, 1 mM phenylmethylsulfonyl fluoride, 1% aprotinin, 5 mM EDTA, 50 mM NaF, 0.1% 2-mercaptoethanol, 5 µM phenylarsine oxide, and 100 µM sodium orthovanadate. Lysates were cleared by centrifugation at 15,000 × g for 45 min at 4 °C. For precipitation of Raf-1 protein, we added 10 µl of the polyclonal anti-Raf-1 antibody to cleared cell lysates. After incubation on ice for 2 h, 100 µl of a 10% solution of formalin-fixed Staphylococcus aureus was added to the anti-Raf-1 immunoprecipitates and incubated on ice for 1 h. The absorbed immune complexes were washed three times in EB buffer and resuspended with 25 µl of 1 × sample buffer. After boiling at 98 °C for 5 min, samples were resolved by SDS-PAGE.

Immunoprecipitates were resolved on 10% acrylamide, 0.193% bisacrylamide gels by SDS-PAGE. Proteins were transferred onto nitrocellulose membranes (1 mAh/cm²) using a dry transfer system (Alert Inc., Seattle, WA), as described (36). The blot was incubated with blocking solution (10 mM Tris-HCl, pH 7.5, 150 mM NaCl, 5% powdered milk) for 1 h at room temperature and then incubated with specific anti-phosphotyrosine (anti-Tyr(P)), anti-Raf-1, anti-Shc, anti-Grb2, or anti-MAP kinase antibodies for 2 h at room temperature with continuous agitation. After three washes in rinse solution (10 mM Tris-HCl, pH 7.5, 150 mM NaCl), the membranes were incubated at room temperature for 1 h with secondary antibody conjugated with horseradish peroxidase for enhanced chemiluminescence (ECL, Amersham Corp.) or conjugated with alkaline phosphatase for colorimetric development. To reprobe the membrane, we stripped membrane with 0.1 M glycine, pH 2.5, at room temperature for 30 min and then reblotted with primary antibody.

The generation of superoxide anions by U937IF cells was measured as the superoxide dismutase-inhibitable reduction of ferricytochrome c at 550 nm in a microtiter plate reader (Molecular Devices Inc., Menlo Park, CA), using air-oxidized and dithionite-reduced cytochrome c as standards. Cells were preincubated for 10 min at 37 °C in HBSS in the wells of a 96-well microtiter plate. The final reaction mixture contained 2 × 10⁶ U937IF cells and 80 µM ferricytochrome c in 250 µl of HBSS. One-half of the wells received superoxide dismutase (25 µg/ml). After the addition of antibodies or various agonists, the plates were incubated at 37 °C and agitated. Serial spectrometric determination was recorded to construct a kinetic curve for the production of superoxide anion. Maximum reduction of ferricytochrome c (25 nm) was achieved by adding 5 µl of freshly prepared sodium dithionite. Production of superoxide anion is expressed as nanomoles of superoxide-dismutase-inhibitable cytochrome c reduction/2 × 10⁶ cells.

RESULTS

Tyrosine Phosphorylation of Raf-1 upon FcRI Activation

To evaluate the involvement of Raf-1 in FcRI signaling, we examined whether Raf-1 is tyrosine-phosphorylated after FcRI stimulation in U937IF cells. Lysates of 2 × 10⁷ U937IF cells were immunoblotted for Raf-1 after differentiation in IFN. IFN increased the expression of p74 Raf-1 (3-fold increase). Raf-1 expression directly correlated with the length of exposure to IFN and peaked at 4 days after differentiation (data not shown). We subsequently used U937IF cells differentiated for 4 days with IFN for all experiments. To determine whether FcRI cross-linking can induce the tyrosine phosphorylation of Raf-1 in myeloid cells, 2 × 10⁷ U937IF cells were first incubated with mAb 197, followed by stimulation with RM. We immunoprecipitated Raf-1 from resting or FcRI-stimulated U937IF cells with a rabbit anti-Raf-1 antibody and performed anti-Tyr(P) immunoblots (mAb 4G10) (Fig. 1). Marked tyrosine phosphorylation of Raf-1 was detected after cross-linking of FcRI receptor (10-fold increase) (Fig. 1A, lanes 6-8). The tyrosine phosphorylation of Raf-1 is very rapid and reaches its maximum response 30 s to 1 min after stimulation (Fig. 1A, lanes 6 and 7). Raf-1 was not tyrosine-phosphorylated in resting cells (Fig. 1A, lane 3) or in U937IF cells stimulated with FcRI alone or RM alone (Fig. 1A, lanes 4 and 5). To confirm the identity of the 74-kDa protein, the membrane was stripped and reprobed for Raf-1 (Fig. 1B). All lanes except sham and preimmune immunoprecipitates brought down an equivalent amount of Raf-1 protein (Fig. 1B, lanes 3-8). The p74 immunoreactive bands of anti-Raf-1 immunoblot is superimposed on tyrosine-phosphorylated bands in anti-Raf-1 immunoprecipitates. The results show that Raf-1 immunoprecipitated from FcRI activated cells is tyrosine-phosphorylated. In other experiments, separate immunoblots of rabbit anti-Raf-1 immunoprecipitates were probed with mouse anti-Tyr(P) or mouse anti-Raf-1 antibodies. The results were exactly the same as our findings previously shown in Fig. 1. Parallel blots of anti-Raf-1 immunoprecipitates, probed with secondary antibody alone, showed no Raf-1 band. Finally, the whole cell lysate confirmed the integrity of our Raf-1 immunoblots (Fig. 1, lane 9). We suggest that the observed difference in intensity of the p74 band in the whole cell lysate versus the anti-Raf-1 immunoprecipitates is due to the presence of more than one p72-74 phosphoprotein with similar electrophoretic mobility. Evidence to support this conclusion include the identification of Raf-1 and Syk kinases as components of the p72-74 phosphoprotein bands (32).

Tyrosine phosphorylation of Raf-1 upon FcRI activation in U937IF cells.

To confirm these results, we performed anti-Tyr(P) immunoprecipitation using agarose-conjugated anti-Tyr(P) antibodies on resting and FcRI-stimulated U937IF cells. We probed these immunoprecipitates with rabbit anti-Raf-1 antibody (Fig. 2A). Only anti-Tyr(P) immunoprecipitate from FcRI activated reacted specifically with anti-Raf-1 antibody (Fig. 2A, lane 3). FcRIII activation did not induce tyrosine phosphorylation of Raf-1 (Fig. 2A, lane 4). When the membrane was stripped and reprobed for phosphotyrosine, the tyrosine-phosphorylated Raf-1 band was superimposed on the upper band of anti-Tyr(P) immunoblot indicated (Fig. 2B, lane 3). The whole cell lysate confirmed the integrity of anti-Tyr(P) and anti-Raf-1 immunoblots (Fig. 2B, lane 5). We observe a diminished number of tyrosine-phosphorylated proteins in our anti-Tyr(P) immunoprecipitates as compared to the whole cell lysates. This is likely due to decreased efficiency of immunoprecipitation by the agarose-conjugated anti-Tyr(P) antibody. The loss of membrane-bound proteins during stripping and washing may decrease the resolution of tyrosine-phosphorylated proteins in our anti-Tyr(P) immunoblot. These data confirm the results shown in Fig. 1 and reveal that Raf-1 is tyrosine-phosphorylated upon FcRI activation.

Anti-Raf-1 immunoblot of anti-Tyr(P) immunoprecipitates from U937IF cells stimulated by FcRI and FcRIII cross-linking.

We determined the kinetics of Raf-1 tyrosine phosphorylation upon FcRI activation and its relation to respiratory burst, a signaling pathway in myeloid cells known to be modulated by GTPases Rac and Rap1a (25, 26). Tyrosine phosphorylation of Raf-1 occurs 20 s after FcRI stimulation (Fig. 3A, lane 5) and reached a peak around 1-2 min (Fig. 3A, lanes 7 and 8) and disappeared 10 min after stimulation (Fig. 3A, lane 10). The respiratory burst begins 3-5 min after FcRI activation (Fig. 3B). FcRI cross-linking activated respiratory burst to produce 2.2 nM superoxide from 2 × 10⁶ cells 10 min after stimulation and showed peak response 30 min after stimulation. PMA also stimulated the respiratory response to produce superoxide anion, but its maximum response was delayed compared to activation of respiratory burst through FcRI stimulation. The respiratory burst is preceded by the tyrosine phosphorylation of Syk, Hck, MAP kinases, and Raf-1, suggesting that Raf-1 could function upstream in the FcRI signal pathway leading to the activation of the respiratory burst response (32, 33).

Kinetics of tyrosine phosphorylation of Raf-1 and respiratory burst upon FcRI activation.

Tyrosine Phosphorylation of Raf-1 Is Specific in FcRI Activation Pathway

To determine whether the tyrosine phosphorylation of Raf-1 we observed is specific for FcRI activation, we determined the effect of cross-linking of other Fc receptors, such as FcRII and FcRIII, as well as stimulation with other agonists, such as PMA (1 µg/ml), FMLP (1 µM), calcium ionophore (1 µM), or thrombin (0.05 unit/ml), on tyrosine phosphorylation of Raf-1 (Fig. 4). In these experiments, we immunoprecipitated an equivalent amount of Raf-1 (Fig. 4B, lanes 3-14). Only FcRI stimulation of U937IF cells resulted in the tyrosine phosphorylation of Raf-1 (Fig. 4A, lanes 11 and 12). PMA, FMLP, calcium ionophore, and thrombin did not induce the tyrosine phosphorylation of Raf-1 (Fig. 4A, lanes 4-7). Similarly, FcRII or FcRIII cross-linking did not induce the tyrosine phosphorylation of Raf-1 (Fig. 4A, lanes 8, 9, 13, and 14). These results indicate that FcRI stimulation specifically induces the tyrosine phosphorylation of Raf-1. Interestingly, FcRIII stimulation is not observed to induce the respiratory burst in U937IF cells, nor does it induce a mobility shift in MAP kinase, or the activation of Hck or Syk kinases (Fig. 6, lanes 11 and 12) (32, 33). Lane 15, a whole cell lysate of U937IF cells stimulated with FcRI, shows the prominent 74-kDa tyrosine-phosphorylated Raf-1 specific band.

Tyrosine phosphorylation of Raf-1 is specific for FcRI activation.

A mobility shift of MAP kinase upon FcRI activation.

Shc Is Tyrosine-phosphorylated after FcRI Activation

One mechanism for the activation of Raf-1 is the direct physical interaction with Ras-GTP, which is positively regulated by Grb2-SOS complex (5, 11). We performed experiments to determine if Shc becomes tyrosine-phosphorylated after FcRI stimulation. Shc was immunoprecipitated with rabbit anti-Shc antibody from U937IF lysates prepared from resting or FcRI-stimulated cells and immunoblotted with mouse anti-Tyr(P) antibody (Fig. 5A, lanes 3-11). We observed that the p52 isoform of Shc is tyrosine-phosphorylated after FcRI activation (Fig. 5A, lane 8). Anti-Shc immunoblot confirmed that the anti-Shc immunoprecipitation brought down the same as amount of p46 and p52 Shc proteins (Fig. 5C, lanes 3-11). The tyrosine phosphorylation of Shc is required for recruitment of a Shc-Grb2 physical interaction. The tyrosine phosphorylation of Shc and coprecipitation of Shc and Grb2 were observed only under condition of FcRI activation (Fig. 5B, lanes 8 and 9). Lane 12 represents a whole cell lysate of U937IF cells and confirms the integrity of anti-Grb2 and anti-Shc immunoblots. These results suggest that Shc is involved in FcRI signaling pathway and Grb2 associates with Shc in a tyrosine phosphorylation-dependent manner during FcRI activation. In other experiments, we have observed a physical interaction between the Shc-SH2 domain and the FcRI subunit.²

Shc is phosphorylated and associates with Grb2 upon FcRI activation.

Mobility Shift of Erk2 upon FcRI Activation

We examined the activation of MAP kinase in U937IF cells stimulated with FcRI cross-linking (Fig. 6). Two bands of MAP kinases are detected in resting U937IF cells corresponding to Erk1 and Erk2, based on the molecular mass values of 42 and 44 kDa, respectively. We observed a mobility shift in MAP kinase in U937IF cells upon FcRI stimulation with mAb 197 (Fig. 6, lanes 8 and 9) or mAb 32.2 (Fig. 6, lane 11) of FcRI receptor. The kinetics of a mobility shift of MAP kinase is different. Stimulation of FcRI with mAb 197 induces MAP kinase mobility shift, reaching its peak around 1 min after FcRI stimulation (Fig. 6, lanes 7 and 8). In contrast, stimulation with mAb 32.2 shows a strong band of mobility shift 5 min after stimulation (Fig. 6, lanes 9 and 10). PMA also induced a mobility shift of MAP kinase (Fig. 6, lane 3). We confirmed that each lane was loaded with an equivalent amount of total protein by Coomassie Blue staining of gel. The mobility shift of MAP kinase induced by FcRI stimulation likely represents a phosphorylation of Erk2, resulting in a retarded migration on SDS-PAGE. No appreciable mobility shift was observed in U937IF cells stimulated with primary antibodies alone (Fig. 6, lanes 3-5), secondary antibody alone (Fig. 6, lane 6), and FcRIII cross-linking (Fig. 6, lanes 11 and 12).

DISCUSSION

The activity of Raf-1 appears to be regulated by multiple mechanisms in mammalian signaling. The presence of a cysteine-rich motif in CR1 suggests that certain modulatory lipids may function on the allosteric regulation of Raf-1 activity (37). The  isoform of the 14-3-3 family of proteins was also identified as a Raf-1 activator in NIH 3T3 cells (38). Several lines of evidence suggest that phosphorylation and/or alteration of the amino-terminal regulatory domain may be a mechanism for the regulation of Raf-1 activity. NH₂-terminal truncation of the Raf-1 cDNA modifies the catalytic activity of Raf-1. Many growth factors stimulate Raf-1 phosphorylation on serine, predominantly, and tyrosine residues in the NH₂ terminus (9, 39, 40, 41, 42). Phosphorylation of Ser²⁵⁹ and/or Ser⁶²¹ regulates Raf-1 activity and correlates with the enhancement of Raf-1 activity in response to various mitogens. In addition, mutation of two in vivo serine phosphorylation sites alters the activity of Raf-1 (15). Morrison et al. have recently reported that tyrosine phosphorylation regulates the activity of Raf-1, since a mutant containing a tyrosine to phenylalanine mutation at Tyr³⁴⁰ and Tyr³⁴¹ sites is not activated and since a truncated Raf-1 lacking tyrosine residues between positions 26 and 303 of amino terminus modifies the function of Raf-1 (43). However, the molecular basis by which tyrosine phosphorylation alters Raf-1 activity is unknown. In this study, we demonstrated that Raf-1 is tyrosine-phosphorylated upon FcRI stimulation, suggesting that Raf-1 is involved in FcRI signal transduction. Recent studies have showed that Raf-1 is tyrosine-phosphorylated in response to growth factors including platelet-derived growth factor, IL-2, IL-3, granulocyte/macrophage colony-stimulating factor, or insulin (4, 5, 8, 44). Platelet-derived growth factor treatment of NIH 3T3 or Chinese hamster ovary cells induces the tyrosine phosphorylation of Raf-1, which activates its serine/threonine enzymatic activity (3). Herein, we showed that FcRI cross-linking of U937IF cells with mAb 197 and RM induces tyrosine phosphorylation of Raf-1 (Fig. 1, lanes 6-8). Ongoing experiments will determine if tyrosine phosphorylation of Raf-1 upon FcRI stimulation coincide with the increase of catalytic activity. The kinetics of tyrosine phosphorylation of Raf-1 upon FcRI is similar to the effect of many growth factors (Fig. 3A). It occurs very rapidly and reaches a peak 1-2 min after FcRI stimulation and quickly returns to base line.

The respiratory burst is well described, and its molecular components have been cloned, including p47^phox, p67^phox, gp91 and p22, Rac1, Rac2, and Rap1a (45, 46). The respiratory burst response can be reconstituted in vitro using recombinant proteins or membrane preparations, making it an excellent model for study of mammalian signal relay (47) (Fig. 7). The respiratory burst in neutrophils, induced through stimulation with PMA or heterotrimeric G proteins, is regulated by serine phosphorylation and the conversion small GTPases Rac and Rap1a to their GTP-bound state (48). Gabig et al. (47) recently reported that the expression of dominant negative mutants of Rac and Rap1a blocks the FMLP-induced respiratory burst in HL-60 cells. Considerable similarity exists between effect of FMLP and Fc receptor signaling (49). Dusi et al. (50) have reported a potential role for MAP kinase in regulation of respiratory burst more recently. The respiratory burst of U937IF cells occurring after FcRI stimulation is less well described but likely is regulated similarly (Fig. 3B) (51). The respiratory burst begins around 5 min and reaches maximal response around 30 min after FcRI stimulation. Interestingly, the FcRI-induced respiratory burst occurs subsequent to tyrosine phosphorylation of Hck, Syk, and Raf-1. PMA also activates the respiratory burst response, but its kinetics differs from the FcRI stimulation. Ongoing experiments seek to determine if the respiratory burst is regulated by the tyrosine phosphorylation and activation of these nonreceptor protein tyrosine kinases and Raf-1. To answer the Raf-1 question, we will overexpress the dominant inhibitory Raf-1 mutant containing tyrosine to phenylalanine mutation at Tyr³⁴⁰ and/or Tyr³⁴¹ sites in U937IF cells followed by phenotypic analysis of FcRI-induced signals in these cells.

Schematic representation for FcRI signaling to the respiratory burst.

FcRI (CD64) is a 72-kDa integral membrane glycoprotein composed of three Ig-like extracellular domains, a single 21-amino acid transmembrane domain, and a 61-amino acid intracytoplasmic domain (19). Recent studies have demonstrated that FcRI signal transduction is mediated through multisubunit complex consisting of the ligand binding receptor molecule in association with a -chain homodimer (52), which contains a YXXL amino acid motif termed the ITAM (22). The  subunit of the FcRI and FcRI, as well as  subunit of TCR/CD3, contain ITAM sequences (53). The FcRI, FcRI, and FcRIII associate with the -chain for the stable transport and assembly of the receptors in the plasma membrane. The signaling pathway through the ITAM involves the activation of the Src family and the Syk/Zap70 family kinases (1, 33, 54). The -chain is rapidly phosphorylated upon FcRI activation on serine, threonine, and tyrosine residues and is associated with Syk (31, 55). In our study, the specificity of Raf-1 tyrosine phosphorylation upon FcRI activation in U937IF cells was investigated by activating different FcR receptors (i.e. FcRII or FcRIII) and by stimulation with other agonists such as PMA, FMLP, calcium ionophore, or thrombin. We found that tyrosine phosphorylation of Raf-1 is specifically induced upon FcRI cross-linking (Fig. 4). Importantly, we have generated similar data for Raf-1 tyrosine phosphorylation following FcRI cross-linking in human bone marrow derived primary macrophage (data not shown). Our results are novel in that they represent the first evidence implicating Raf-1 in Fc receptor signaling. The data are consistent with other results reported by Gupta et al. (56) showing the another ITAM-linked multisubunit receptors, TCR and BCR, signaling through the activation of Ras and Raf-1. Previously, Morrison et al. demonstrated that PMA enhances the serine/threonine activity, thereby increasing the catalytic activity of Raf-1 (57). It is interesting to speculate that the tyrosine phosphorylation of Raf-1 may be sequentially linked to signals mediated through phosphorylation of -chain, Hck, or Syk in this system. In this model, Raf-1 may be a substrate for Hck, Syk, or other tyrosine kinases. In support of such a model in preliminary experiments, we have observed a physical interaction between Hck and Raf-1.²

Additional lines of evidence in this report support a potential role for Raf-1 in FcRI signal transduction. We have observed that FcRI stimulation induced the tyrosine phosphorylation of Shc. A mobility shift occurs in MAP kinase after FcRI cross-linking in U937IF cells (Fig. 5, lanes 7-10). The Shc-Grb2 complex is known to activate Ras. Raf-1 is a downstream effector of Ras and is known to activate MEK and MAP kinase (1). Alternatively, Raf-1 can be activated through Ras-dependent or -independent mechanisms (2). Ras-GTP, which is positively regulated by Grb2-SOS complex, activates Raf-1 through direct physical interaction (58). The adaptor protein, Shc, is thought to be involved in signaling from cytoplasmic tyrosine kinases through Grb2 and SOS. Ravichandran et al. (59) have demonstrated that Shc is tyrosine-phosphorylated and associated with the -chain of the TCR upon T cell receptor activation. Other laboratories have reported that the heterologous expression of the Shc-SH2 domain in T cells blocks TCR signaling, suggesting that Shc plays an important role in TCR functions (60).

We found that Shc is phosphorylated and associated with Grb2 upon FcRI stimulation (Fig. 5A, lanes 8 and 9). Tyrosine phosphorylation of Shc increased the binding of Grb2 to Shc in our system (Fig. 5B, lane 8). It is well known that tyrosine phosphorylation of Shc is linked to the activation of Ras. Recent studies have shown that the tyrosine phosphorylation of Shc induces its interaction with Grb2, which is essential for binding of nucleotide exchange protein SOS and Ras activation (15, 61). In other experiments, we observed Shc is physically associated with -chain of FcRI receptor in U937IF cells.² Our data support the notion that tyrosine phosphorylation of Raf-1 connects the FcRI signaling pathway sequentially through -chain, Shc, Grb2-SOS, and possibly Ras. Recent data suggest a direct connection between Ras and Rac in several signaling pathways (62). These observations suggest a mechanism by which the conversion of GDP^ras to GTP^ras could lead to the formation of GTP^rac, known to be required for the assembly of the respiratory burst response in myeloid cells. Hence we hypothesize that GTP^rac needed to assemble the respiratory burst comes from the activation of Ras.

MAP kinases are located downstream of Raf-1 kinase. MAP kinases have been implicated in signaling pathway of many hematopoietic receptors, such as TCR, BCR, FcRI, and FcRI receptors (63, 64). We show that FcRI cross-linking induces a mobility shift of MAP kinase, suggesting that MAP kinase is phosphorylated upon FcRI activation (Fig. 6). Both FcRI specific antibodies, mAb 197 and mAb 32.2, induce a mobility shift in MAP kinase. The mobility shift induced by mAb 197 stimulation is more rapid than that of mAb 32.2 (Fig. 6, lanes 7-9). PMA also induces a MAP kinase mobility shift (Fig. 6, lane 2). Stimulation of FcRIII does not induce this response and failed to induce the tyrosine phosphorylation of Raf-1. Our data suggest that Erk2 is phosphorylated, as manifested by a retarded mobility on SDS-PAGE. Experiments are ongoing to determined if a direct relationship exists between the tyrosine phosphorylation of Raf-1 and activation of Raf-1 kinase and MAP kinase.

Our data also demonstrate that Raf-1 is tyrosine-phosphorylated upon FcRI stimulation. Tyrosine phosphorylation of Raf-1 is correlated with phosphorylation of Shc, which associates with Grb2 in a tyrosine phosphorylation-dependent manner. Based on these results, we propose a model for FcRI signal transduction that involves tyrosine phosphorylation of Raf-1. This model predicts the sequential activation of FcRI, FcRI, Hck, Syk, Shc/Grb2/Sos, Ras, and Raf-1 and the activation of MEK and MAP kinases (29, 31, 32, 33, 61) (see Scheme I above).

The further study of the role of Raf-1 tyrosine phosphorylation and its interaction with other components in FcRI signaling may clarify the molecular mechanisms that connect upstream cell surface receptors and their associated nonreceptor protein tyrosine kinases to the downstream activation of serine/threonine kinase cascades. The elucidation of these signaling pathways in macrophages will contribute to our understanding of the role of Raf-1 in post-mitotic cell functions, including macrophage activation leading to the activation of the respiratory burst.