Dynamic changes in C-Raf phosphorylation and 14-3-3 protein binding in response to growth factor stimulation: differential roles of 14-3-3 protein binding sites.

Phosphorylation events play a crucial role in Raf activation. Phosphorylation of serines 259 and 621 in C-Raf and serines 364 and 728 in B-Raf has been suggested to be critical for association with 14-3-3 proteins. To study the functional consequences of Raf phosphorylations at these positions, we developed and characterized phosphospecific antibodies directed against 14-3-3 binding epitopes: a monoclonal phosphospecific antibody (6B4) directed against pS621 and a polyclonal antibody specific for B-Raf-pS364 epitope. Although 6B4 detected both C- and B-Raf in Western blots, it specifically recognizes the native form of C-Raf but not B-Raf. Contrary to B-Raf, a kinase-dead mutant of C-Raf was found to be only poorly phosphorylated in the Ser-621 position. Moreover, serine 259 to alanine mutation prevented the Ser-621 phosphorylation suggesting an interdependence between these two 14-3-3 binding domains. Direct C-Raf.14-3-3 binding studies with purified proteins combined with competition assays revealed that the 14-3-3 binding domain surrounding pS621 represents the high affinity binding site, whereas the pS259 epitope mediates lower affinity binding. Raf isozymes differ in their 14-3-3 association rates. The time course of endogenous C-Raf activation in mammalian cells by nerve growth factor (NGF) has been examined using both phosphospecific antibodies directed against 14-3-3 binding sites (6B4 and anti-pS259) as well as phosphospecific antibodies directed against the activation domain (anti-pS338 and anti-pY340/pY341). Time course of Ser-621 phosphorylation, in contrast to Ser-259 phosphorylation, exhibited unexpected pattern reaching maximal phosphorylation within 30 s of NGF stimulation. Phosphorylation of tyrosine 340/341 reached maximal levels subsequent to Ser-621 phosphorylation and was coincident with emergence of kinase activity. Taken together, we found substantial differences between C-Raf.14-3-3 binding epitopes pS259 and pS621 and visualized for the first time the sequence of the essential C-Raf phosphorylation events in mammalian cells in response to growth factor stimulation.

Raf kinase was originally discovered as the oncogenic product of mouse sarcoma virus 3611 (1). Whereas invertebrates encode only a single Raf kinase, vertebrates express three isoforms, designated as A-, B-, and C-Raf. All of these isoforms share three highly conserved regions (CR1, CR2, and CR3). 1 CR1 and CR2 form the regulatory part of the kinases, and CR3 represents the catalytic domain (reviewed in Refs. [2][3][4]. The C-Raf gene encodes a protein of 648 amino acids, which is expressed as a 74-kDa polypeptide (5). A-Raf is a 68-kDa protein showing 60% homology to C-Raf (6). B-Raf is expressed as a full-length protein of 95 kDa or as smaller splice variants (7).
Despite intensive investigations, the mechanism of Raf activation is still not completely understood. A current model suggests that Raf association with plasma membrane lipids (or lipid microdomains called rafts) represents an important step in the Raf activation process (8). The binding to the activated small G protein Ras (Ras-GTP) occurs (in this model) in the plane of the inner leaflet of plasma membrane. Transition from the inactive to the active membrane-associated form of Raf involves Raf dimerization, several phosphorylation events, and complex formation between Raf and 14-3-3 proteins. Targeting of Raf to the membranes and binding to activated Ras reorients Raf molecules and induces conformational changes, which allow phosphorylation-dephosphorylation events. Phosphorylation events play a crucial role in Raf activation. Morrison et al. (9) identified three major phosphorylation sites in C-Raf isolated from platelet-derived growth factor-activated NIH3T3 cells: the serine residues at positions 43, 259, and 621 (9). Phosphorylation of serine 621 seems to be essential for Raf activation, because the mutation of serine 621 to alanine resulted in a Raf protein that could no longer be activated even in the presence of Ras and Src kinases (5). 2 In contrast, exchange of serine 259 by alanine or aspartic acid resulted in enhancement of kinase activity indicating that phosphorylation of serine 259 is inhibitory (10). The one or more kinases that phosphorylate this serine have not yet been fully elucidated. However, it was reported that Akt-as well as protein kinase A-mediated phosphorylation of serine 259 inhibits Raf kinase activation (10 -12). Exchange of serine 43 to alanine had no effect on C-Raf activity, and its phosphorylation is thought to be mediated by protein kinase A (13). Additional phosphorylation sites for C-Raf have been reported for threonine 268 (autophosphorylation site (9)), serine 338, which may be a target of p21-activated kinase kinase (14,15) and serine 497/499 in the activation loop (16,17). Particular importance has been assigned to the Src-dependent phosphorylation of C-Raf tyrosine residues at positions 340/341 (18) that may be required for maximal Raf activation. Substitution of these residues with the negatively charged aspartic acid, mimicking the phosphorylated status increased the basal activity of C-Raf (18). Members of 14-3-3 protein family have been found to participate in numerous signal transduction pathways. Since their discovery at least seven mammalian isoforms of 14-3-3 proteins have been identified and cloned (19 -21). All of the 14-3-3 proteins form homodimers and/or heterodimers that interact with signaling proteins, including protein kinase C, Raf kinases, kinase suppressor of Ras (KSR), Cdc25 phosphatases, and BAD protein (21). A common outcome of 14-3-3 protein binding may be translocation of target proteins into the cytosol. With respect to the Raf activation cycle, 14-3-3 proteins have been initially reported to support Raf activation (22)(23)(24). On the other hand, it was also described that 14-3-3 proteins are not essential for Raf function (25,26). A possible explanation for these contradictory observations might be the fact that Raf contains at least three potential 14-3-3 binding domains. It has been well documented that 14-3-3 proteins bind to epitopes surrounding phosphoserines 621 and 259 in C-Raf (20,27). A third 14-3-3 binding domain has been proposed for the cysteine-rich domain, because mutations of residues 143-145 in C-Raf caused a loss of 14-3-3 binding (28). Recently, serine 233 has been reported to play a role in 14-3-3 binding as well (12). However, 14-3-3 binding affinities for these putative C-Raf binding sites have not been determined for full-length proteins. B-Raf kinase purified from Sf9 cells has also been found to bind 14-3-3 proteins. The putative 14-3-3 binding sites are presumed to be analogous to those in C-Raf, i.e. domains surrounding phosphoserines 364 and 728. Whereas 14-3-3␤ and 14-3-3 were identified (22,29) in several yeast two-hybrid screens using C-Raf as bait, B-Raf has been reported to interact with 14-3-3, -, and - (30). Because the C-Raf phosphoserine 621 binding site is highly conserved in all three Raf isoforms, it has been assumed that A-Raf interacts with 14-3-3 as well. Indeed, using A-Raf as a bait in a two-hybrid screen we isolated 14-3-3 isoforms , ␤, ⑀, , and as possible interaction partners (3). 3 To study the dynamic changes of Raf phosphorylation, particularly with respect to 14-3-3 binding sites, we generated antibodies directed against these binding epitopes. Here we describe development, characterization, and experimental use of monoclonal antibody 6B4, which specifically recognizes C-Raf-serine 621 (and serine 728 in B-Raf) only in the phosphorylated state. Additionally, for analysis of 14-3-3 binding sites in B-Raf we used a novel phosphospecific antibody directed against B-Raf-phosphoserine 364. The high affinity clone 6B4 is suitable for immunoprecipitation of native C-Raf kinase but not of B-Raf. We show here that the phosphorylations of serines 621 and 259 are dependent on each other and that the endogenous kinase activity of C-Raf, contrary to B-Raf, influences strongly the degree of serine phosphorylation in position 621. Moreover, we examined the kinetics of C-Raf association with 14-3-3 proteins by surface plasmon resonance (SPR) technology and found that the 14-3-3 binding domain surrounding phosphoserine 621 reveals higher affinity for 14-3-3 proteins than pS259 domain and represents probably the major binding site.
Finally, we examined the time course of endogenous C-Raf activation in mammalian cells using phosphospecific antibodies directed against the major phosphorylation sites. We found that maximal phosphorylation of serine 621 occurs early in the activation cycle. Phosphorylation in positions tyrosine 340/341 peaked delayed to serine 621 phosphorylation and was characterized by appearance of discrete dots at the plasma membrane. Thus, using phosphospecific C-Raf antibodies we resolved for the first time the sequence of the C-Raf phosphorylation events at the early stage of stimulation in mammalian cells.

EXPERIMENTAL PROCEDURES
Materials-Benzamidine, leupeptin, aprotinin, phenylmethylsulfonyl fluoride, and Nonidet P-40 were obtained from Sigma. Glutathione-Sepharose was purchased from Amersham Biosciences, and Ni-NTAagarose was from Qiagen. Monoclonal anti-phospho-ERK antibodies were from New England Biolabs. Polyclonal anti-B-Raf-pS364 phosphospecific antibody was produced by Quality Controlled Biochemicals (Hopkinton, MA). Phosphospecific anti-C-Raf-pS259 antibody was from New England Biolabs, and polyclonal anti-14-3-3 antibodies (K-19 and H-8) were from Santa Cruz Biotechnology. Phosphospecific anti-C-Raf-pY340/pY341 was from BioSource, International. Antibody directed against C-Raf-phosphoserine 338 was purchased from Upstate Biotechnology. Horseradish peroxidase-conjugated polyclonal anti-rabbit and anti-mouse IgG were from Amersham Biosciences. NGF was from Cell-Systems. Phosphopeptide pS259 and pS621 containing 13 amino acids were purified by high-pressure liquid chromatography, and the molecular weight was verified by mass spectroscopy.
Generation of Monoclonal Antibodies Recognizing C-Raf Phosphoserine 621--Monoclonal antibodies were generated as described previously (31). Briefly, phosphopeptide pS621 (KINRSApSEPSLHRA), conjugated to thiolated keyhole limpet hemocyanin, was injected subcutaneously into BALB/c mice (10 g/animal and injection) four times at 14-day intervals. Two weeks later, on three consecutive days animals received booster injections. Spleenocytes from the immunized mice were fused with non-producer myeloma cells. Hybridoma cell lines secreting antibodies directed against the phosphoserine-Ser-621 C-Raf epitope were identified by screening using phosphorylated and nonphosphorylated peptides bound to enzyme-linked immunosorbent assay plates. Monoclonal antibodies were purified from serum-free cell culture supernatants by thiophilic adsorption chromatography.
Cloning of His 6 -and GST-tagged Raf Genes-Human C-Raf wt and C-Raf mutants such as C-Raf-S259A, C-Raf-S621A, and C-Raf-S259A/ S621A have been modified to introduce the recognition sequence for NheI immediately upstream of the ATG codon. In addition, adenine of ATG was converted into cytidine, which changes this codon from methionine into leucine. Modified gene was cleaved with NheI, and sticky ends were filled-in with Klenow enzyme plus dNTPs. After heat inactivation of the polymerase the C-Raf-containing fragment was released by further digestion with XbaI and ligated into BamHI (filled-in, see above)/XbaI-cleaved pFastBac-Hta (Invitrogen). C-Raf protein expressed by this system was N-terminally extended by 28 amino acids, including a 6ϫ histidine tag for affinity purification on an Ni-NTA matrix. This extension did not influence biological properties of Raf kinases in vivo or in vitro. 4 Human A-and B-Raf wt were also Nterminally his-tagged and cloned in a way similar to C-Raf. pFastBac-GST-C-Raf expression vector has been constructed by ligating BamHIcleaved GST gene with BamHI-cleaved pFastBac-C-Raf.
Infection of Sf9 Insect Cells, Purification of Raf Kinases, SDS-PAGE, and Western Blot Analysis-For the production of recombinant Raf kinases the Sf9 cells were infected with the desired baculoviruses at a multiplicity of infection of 5 and incubated for 48 h at 30°C. The cells were then washed with PBS buffer and pelleted at 1100 rpm (Megafuge 1.0R, Heraeus). The Sf9 cell pellets (2 ϫ 10 8 cells) were lysed in 10 ml of Nonidet P-40 lysis buffer containing 25 mM Tris-HCl, pH 7.6, 150 mM NaCl, 10 mM sodium pyrophosphate, 25 mM ␤-glycerophosphate, 25 mM NaF, 10% glycerol, 1 mM sodium vanadate, 0.75% Nonidet P-40, and a mixture of standard proteinase inhibitors (1 mM benzamidine, 1 mM phenylmethylsulfonyl fluoride, 10 g/ml leupeptin, and 10 g/ml aprotinin) for 45 min with gentle rotation at 4°C. The lysate was centrifuged at 27,000 ϫ g (SS34 rotor, Sorvall centrifuge) for 30 min at 4°C. The supernatants (10 ml) containing GST-tagged Raf kinases were incubated with 0.5 ml of glutathione-Sepharose beads for 2 h at 4°C with rotation. After incubation the glutathione-Sepharose beads were washed three times with Nonidet P-40 buffer, whereby the third wash contained only 0.2% Nonidet P-40 instead of 0.75%. Raf kinases bound to the beads were eluted 3 times with 0.5 ml of 25 mM Tris-HCl, pH 7.6, 150 mM NaCl, 25 mM ␤-glycerophosphate, 25 mM NaF, 10% glycerol, 0.1% Nonidet P-40, and 20 mM glutathione. The purification procedure for His-tagged Raf kinases (B-and C-Raf) was similar as described above, with the exception that the Sf9 cell lysates (10 ml) were incubated with 0.5 ml of Ni-NTA-agarose. The bound proteins were then eluted with imidazole using a step gradient. 14-3-3 protein was expressed in Escherichia coli as glutathione S-transferase (GST) fusion proteins using pGEX2T vector (Amersham Biosciences) and purified by glutathione-Sepharose affinity chromatography. Purified 14-3-3 was dialyzed against 20 mM Tris-HCl (pH 7.6) and 150 mM NaCl and concentrated to 6 mg/ml. The purity of Raf kinases and 14-3-3 was documented by SDS-polyacrylamide gel electrophoresis (10% gels) and staining with Coomassie Blue. For Western blot analysis the gels were transferred to nitrocellulose membranes (Schleicher & Schuell) and probed either with phosphospecific antibodies or with antibodies specific for A-, B-, and C-Raf and phospho-ERK as indicated in the legends to the figures. After washing procedure the membranes were incubated with specific secondary horseradish peroxidase-conjugated antibodies and detected by enhanced chemiluminescence (ECL, Amersham Biosciences).
Immunoprecipitations-Sf9 insect cell pellets (1 ϫ 10 7 cells) infected with proteins of interest or PC12 cells (2 ϫ 10 7 cells) were lysed in 800 l of Nonidet P-40 lysis buffer and proteinase inhibitors (see above) for 45 min at 4°C. After centrifugation at 27,000 ϫ g for 20 min the lysates were incubated (1 h, 4°C) with phosphospecific 6B4 antibody or other antibodies as indicated. After adding either Protein-A or Protein-G-agarose (40-l bead volume) the incubation was continued at 4°C for 2 h. The agarose beads were washed twice with Nonidet P-40 buffer and once with kinase buffer. Kinase assay has been carried out directly with immunoprecipitated proteins as described below. The kinase assay mixture was then supplemented with Laemmli buffer, boiled for 5 min at 100°C, and applied to SDS-PAGE. After Western blotting the precipitated proteins and ERK were analyzed by enhanced chemiluminescence.
Kinase Activity Measurements-Kinase assays with Raf samples were performed using recombinant MEK and ERK-2 as substrates in 25 mM Hepes, pH 7.6, 150 mM NaCl, 25 mM ␤-glycerophosphate, 10 mM MgCl 2 , 1 mM dithiothreitol, and 1 mM sodium vanadate buffer (50-l final volume). Following additions of Raf-containing samples, the reaction mixtures were incubated for 30 min at 26°C. The incubation was terminated by addition of Laemmli sample buffer, and the proteins were separated by 10% SDS-PAGE and transferred to nitrocellulose membranes. The extent of ERK phosphorylation was determined by anti-phospho-ERK antibodies.
Immunostaining and Fluorescence Microscopy-PC12 cells were cultured in Dulbecco's modified Eagle's medium (Invitrogen), 10% horse serum, and fetal calf serum. For the experimental procedure the cells were cultured for 24 h under low serum conditions and treated with NGF (50 ng/ml) for the times indicated (0, 10, 20, 30, 60, 120, and 600 s). The cells were then fixed with 4% paraformaldehyde in PBS for 30 min at room temperature or overnight at 4°C. The cells were washed with Tris-buffered saline (TBS, 25 mM Tris-HCl, pH 7.4, 0.8% NaCl, 0.2% KCl), 10% goat serum, and 0.1% Tween 20. Immunostaining with the indicated phosphospecific C-Raf antibodies or rabbit anti-C-Raf occurred for 24 h at 4°C. Control samples were treated in the absence of the first antibody or after preadsorption with the corresponding recombinant Raf isoforms. Cells were washed three times with TBS containing 10% goat serum and incubated for 30 min with 2 ng/ml Cy-3-coupled goat anti-rabbit antibody (BIOMOL) or 2 ng/ml Cy-2-coupled antimouse antibody (BIOMOL). Cells were washed again three times with TBS and 10% goat serum, then covered with Mowiol in 50% glycerol/ PBS (v/v) and observed under a Leica confocal microscope (TCS, Leica, Heidelberg, Germany).
Calculation and Statistics-Immunohistochemical signals were analyzed using the National Institutes of Health Image software. Single estimations from independent experiments were pooled, and the results were expressed as mean and standard error of the mean. Statistical significance of differences was assessed by analysis of variance followed by BonferroniЈs test using Prism (GraphPad, San Diego, CA).
Biosensor Measurements-To determine quantitatively the interactions between the purified 6B4 antibody and different Raf preparations the surface plasmon resonance (SPR) technique was applied. All biosensor measurements were carried out on a BIAcore-X system (Biacore AB, Uppsala, Sweden) at 25°C. For that purpose the biosensor chip CM5 was first loaded with anti-GST antibody using covalent derivat-ization. Purified and GST-tagged C-Raf was injected in biosensor buffer (10 mM Hepes, pH 7.4, 150 mM NaCl, and 0.01% Nonidet P-40) at a flow rate of 10 l/min, which resulted in a deposition of ϳ600 -650 response units (RU). Next the purified 6B4 antibody was injected at increasing concentrations (25-250 nM) as indicated in Fig. 1B. The values for unspecific binding measured in the reference cell were subtracted. To measure the association between 14-3-3 proteins and Raf, GST-tagged 14-3-3 was first immobilized by anti-GST antibody. In the next step Raf kinases and Raf mutants were injected as indicated in Fig. 6. In this assay some unspecific binding of C-Raf to the GST group was observed (ϳ10% of the maximal Raf binding). However, addition of bovine serum albumin (0.2 mg/ml) reduced significantly the unspecific binding. The evaluation of the kinetic parameters for Raf binding to 6B4 was performed by non-linear fitting of binding data using the analysis program BiaEvaluation 2.1. The apparent association (k a ) and dissociation rates constant (k d ) were evaluated from the differential binding curves (Fc2-Fc1) as shown in Fig. 1B assuming an A ϩ B ϭ AB association type for the protein-protein interaction. The affinity constant K D was calculated from the equation To measure the competition between Raf and peptides for the phospho-specific antibody, the 6B4 antibody was covalently immobilized to the sensor chip CM5 and the competition curves between purified His-tagged C-Raf and peptides of interest were monitored.

Generation and Characterization of Monoclonal C-Raf-Ser-621 Phosphospecific Antibodies
To facilitate detection of C-Raf phosphorylation at serine 621 and to study the dynamic changes of phosphorylation in this position in response to growth factor stimulation, monoclonal antibodies directed against the phosphorylated C-Raf-pS621 peptide (pS621 peptide) were generated as outlined under "Experimental Procedures." Testing the supernatants from fused spleen cells by enzyme-linked immunosorbent assay using phosphorylated and non-phosphorylated Ser-621 peptides, four clones producing monoclonal antibodies specific to phosphoserine 621 were obtained (designated as 6B4, 12H2, 9G7, and 9C2 belonging to IgG2a or IgG1 Ig subclasses). Supernatants were assayed first by immunoblotting (Western blot analysis) using Raf proteins expressed in Sf9 insect cells as antigens. Proteins carrying the serine to alanine point mutations (C-Raf-S621A and S728A in B-Raf) and Raf kinases expressed in E. coli (not phosphorylated at these positions) served as negative controls.
Clone 6B4 displayed the highest affinity and selectivity for Raf kinases phosphorylated in position S621. We purified the phosphospecific 6B4 antibody (belonging to the IgG2a subclass) from the hybridoma supernatant and concentrated it to 2 mg/ml. As shown in Fig. 1A the purified antibody 6B4 detected wild type C-Raf expressed in Sf9 cells with high intensity. Interestingly, we did not observe any significant differences between C-Raf wt and highly activated C-Raf expressed in the presence of Ras12V and Lck. The 6B4 antibody recognized neither C-Raf possessing serine to alanine mutation at position 621 (C-Raf-S621A) nor C-Raf expressed in bacteria. As depicted in Fig. 1A (assays done with crude cell lysates) we observed almost no cross-reactivity with other proteins demonstrating a high specificity for Ser-621-phosphorylated C-Raf. On the other hand, all of the C-Raf samples tested in Fig. 1 (C-Raf wt, C-Raf-89L/375W, and C-RafS621A from Sf9 lysates and C-Raf wt from E. coli lysate) were detected by a monoclonal antibody PBB-1, which is directed against the non-phosphospecific region of the catalytic domain (32) or by C-Raf polyclonal antibody (SP63) directed against the C-terminal region (33,34). To determine the sensitivity of the 6B4 antibody toward the C-Raf samples we used C-Raf purified from Sf9 insect cells. The phosphospecific antibody 6B4 was able to detect 5 ng of C-Raf at the 1:30,000 dilution (data not shown).
To determine the specificity and the affinity constants of monoclonal antibody 6B4 for C-Raf we used BIAcore technol-ogy, a method which allows monitoring of biomolecular interactions in real time. For that purpose, we applied two different approaches. First, we immobilized 6B4 antibodies to the CM5 biosensor chip (as described under "Experimental Procedures") and measured the association of C-Raf wt in the presence and absence of phosphopeptides pS621 and pS259. The phosphopeptide pS621 competed effectively with 6B4, whereas phosphopeptide pS259 showed practically no effect. The nonphosphorylated Ser-621 peptide interfered only marginally at higher concentrations (20 M) with the C-Raf/6B4 association (data not shown). Finally, to determine the affinity constant and kinetic parameters for association between C-Raf and pS621 antibody (6B4), another biosensor approach has been applied. For that purpose GST-tagged C-Raf was immobilized to the CM5 chip coated with anti-GST antibodies (see also "Experimental Procedures"). In the next step the associationdissociation curves with purified 6B4 in the concentration range of 25-250 nM were monitored (Fig. 1B). The apparent association rate constant (k a ) was 1.42 ϫ 10 5 M Ϫ1 s Ϫ1 , and the dissociation rate constant (k d ) was 7.39 ϫ 10 Ϫ4 s Ϫ1 , thus leading to affinity constant (K D ) of 5.20 nM. The extent of C-Raf association with 6B4 reached the value of 70% correlating well with mass spectroscopy (MS) data obtained independently. 5 These results reflect a high affinity binding between C-Raf kinase and Ser-621 phosphospecific antibody.

Assessment of B-Raf Phosphorylation at Positions Serine 728 and 364
Because B-Raf kinase possesses almost identical amino acid sequence surrounding serine 728, we tested 6B4 antibody for its ability to recognize B-Raf in Western blot experiment as well. Similar to C-Raf, 6B4 antibody detected B-Raf wt but not the B-Raf-S728A mutant. B-Raf expressed in E. coli was also not recognized by 6B4 ( Fig. 2A). In analogy to C-Raf, domains surrounding serine 364 (corresponding to Ser-259 of C-Raf) and 728 in B-Raf are supposed to represent binding sites for 14-3-3 proteins. Exact analysis of B-Raf phosphorylation degree in these positions and association studies with 14-3-3 proteins have not been reported. Therefore, we performed analysis of B-Raf serine 364 and 728 phosphorylation, using monoclonal 6B4 antibody (for phosphoserine 728) and the novel polyclonal antibody specific for B-Raf phosphoserine 364. As shown in Fig.  2A using equivalent amounts of purified B-and C-Raf proteins the degree of phosphorylation in positions serine 621 and 728 determined by 6B4 antibody was comparable in both kinases: ϳ70% were phosphorylated (based on results in Hekman, et al. 5 ). On the other hand, the phosphospecific antibodies directed against pS364 and pS259 for B-and C-Raf, respectively, differ in their specificity. Whereas anti-pS259 antibody recognized also B-Raf to lower degree, the novel anti-pS364 was specific for B-Raf ( Fig. 2A).
In contrast to the Western blot experiments, the 6B4 antibody immunoprecipitated C-Raf very effectively, whereas the degree of B-Raf precipitation was negligible (compare Figs. 1 and 2B). We tested in these experiments also the activated forms of B-and C-Raf (co-expression with Ras and Lck, depicted in Figs. 1 and 2 as R/L). Despite elevated kinase activities, no significant differences in the precipitation efficiency were observed. Surprisingly, association of phosphospecific antibodies 6B4 with GST-C-Raf⅐14-3-3 complex did not displace the bound 14-3-3 proteins, suggesting a ternary complex between Raf, 14-3-3, and 6B4 antibody. However, 14-3-3 in this ternary complex has been found to be attached to the pS259 binding site, because interactions of 6B4 (under immunoprecipitation conditions) with GST-C-Raf-BXB⅐14-3-3 complex (where C-Raf-BXB represents the CR3 region possessing only pS621 binding site) displaced completely the bound 14-3-3 proteins (data not shown). In contrast, association of GST-C-Raf FIG. 1. Analysis of C-Raf phosphorylation in position serine 621 and immunoprecipitation of C-Raf and C-Raf mutants by phosphospecific antibody 6B4. A, crude lysates from Sf9 insect cells and E. coli (24 g) expressing Raf kinases were subjected to SDS-PAGE (10% gels), blotted onto nitrocellulose, and probed with phosphospecific monoclonal antibody 6B4 or polyclonal anti-C-Raf antibody SP63. Crude lysates from Sf9 insect cells or E. coli expressing C-Raf wt or C-Raf mutants as indicated were immunoprecipitated by purified 6B4 and polyclonal anti-C-Raf SP63 antibodies as described under "Experimental Procedures." C-Raf kinases were co-expressed with or without Ras12V and Lck kinase (depicted as R/L in the figure). The precipitated proteins were supplemented by Laemmli buffer, boiled, subjected to SDS-PAGE (10% gels), and blotted onto nitrocellulose. Kinase activity of the precipitated material was tested with purified MEK and ERK-2 and detected by anti-phospho-ERK antibodies. B, quantitative Biosensor analysis of C-Raf association with phospho-specific 6B4 antibody. CM5 sensor chip was first loaded with anti-GST antibody using covalent derivatization. Purified and GST-tagged C-Raf was injected resulting in a deposition of ϳ600 -650 response units. Next the purified 6B4 antibody was injected at indicated concentrations (25-250 nM). The diagram shows differential binding curves (Fc2-Fc1 values), thus representing the specific association of 6B4 antibody to C-Raf. with lipid vesicles resulted in a complete removal of 14-3-3 proteins accompanied by reduction of kinase activity (8).

Serine 621 Phosphorylation in C-Raf, but Not in B-Raf, Requires Endogenous Kinase Activity
Functional consequences that are associated with serine 621 phosphorylation are of particular interest. For that purpose, we generated and purified C-Raf mutants possessing mutations at positions that are involved in binding of 14-3-3 proteins: i.e. C-Raf-S621A, C-Raf-S259A, and C-Raf-S621A/S259A. In addition, we investigated Raf kinases expressed in Sf9 insect cells, including fully activated C-Raf (expressed in the presence of Ras and Lck), the constitutively active mutant C-Raf-Y340D/ Y341D, the Ras binding-deficient mutant C-Raf-R89L, and the kinase-dead C-Raf mutant (C-Raf-K375W). Testing 6B4 antibody, we observed first that the degree of serine 621 phosphorylation was similar for C-Raf wt, for the constitutively active mutant C-Raf-Y340D/Y341D and for the Ras/Lck-activated C-Raf (see Fig. 1). These observations are in agreement with data obtained recently by MS analysis, 5 which demonstrated that the degree of Ser-621 phosphorylation in C-Raf was similar for activated and non-activated C-Raf. Furthermore, Morisson et al. found that the degree of Ser-621 phosphorylation of C-Raf in NIH 3T3 cells was not changed upon platelet-derived growth factor activation (9). On the other hand the kinase-inactive form of C-Raf (C-Raf-K375W) was very poorly phosphorylated at serine 621 ( Figs. 1 and 3); a fact that could not exclude the possibility of autophosphorylation or trans-phosphorylation at this position, although other mechanisms should also be considered. Contrary to C-Raf, the phosphorylation of serine 728 in B-Raf was not changed in the kinase-dead mutant form of B-Raf (Fig. 3). These results suggest that the mode of phosphorylation of serine 621 and 728 in C-and B-Raf differs from each other. In contrast, the degree of phosphorylation in positions serine 259 and 364 for C-and B-Raf, respectively, were not influenced by the K375W mutation that presumably blocks ATP binding.

Serine 259 Phosphorylation Is a Prerequisite For Effective Phosphorylation in Position Serine 621
To investigate the consequences of mutations in 14-3-3 binding domains (i.e. epitopes surrounding pS259A and pS621A) C-Raf wt and C-Raf mutants such as C-Raf-S621A, C-Raf-S259A, and C-Raf-S621A/S259A were tested for binding with 6B4 using both biosensor technique and Western blots. As expected, we observed effective binding of 6B4 to C-Raf wt and practically no binding to C-RafS621A or C-RafS621A/S259A (Fig. 4A). Surprisingly, the extent of association of 6B4 with C-RafS259A mutant, possessing intact serine 621, was only about 10% of that measured for C-Raf wt. The Western blots depicted in the inset to the Fig. 4A correlate well with biosensor data revealing that the degree of Ser-621 phosphorylation in S259A mutant was also significantly reduced (about 15-20% of the value determined for C-Raf wt). Nevertheless, the kinase activity of this preparation was still about 2-fold higher than that measured for C-Raf wt (Fig. 4B). Because the Ser-621 non-phosphorylated population of C-Raf in this sample is known to possess no kinase activity, these results demonstrate a strong stimulating effect of serine to alanine mutation in position 259 in terms of specific kinase activity. Consistent data were recently published (11,35).

Analysis of A-, B-and C-Raf Associations with 14-3-3 Proteins Using SPR Technique Putative 14-3-3 Binding Sites of C-Raf Differ in Their
Binding Affinities-Having investigated the phosphorylation status of C-Raf in positions serine 259 and serine 621 by phosphospecific antibodies, both representing 14-3-3 protein binding sites, we tried next to determine the binding strength of 14-3-3 protein for these two different (and isolated) binding sites. For that purpose we used first phosphospecific antibodies directed against phosphoserines 259 and 621 and the corresponding phosphopeptides as competitive inhibitors. Purified his-tagged C-Raf wt preparation possessing practically no 14-3-3 proteins bound has been used in these competition assays. The 14-3-3⅐C-Raf association was monitored by SPR technique. Because 14-3-3 isoform has been found to participate in the interac- tions with all three Raf isoforms, we used 14-3-3 for SPR studies presented here. For that purpose GST-14-3-3 protein was captured, and C-Raf association was monitored first in the presence or absence of 6B4 or anti-pS259 antibodies. As depicted in Fig. 5A a significant association between 14-3-3 and C-Raf was detectable. In the presence of anti-pS621 antibody the C-Raf⅐14-3-3 interaction was almost completely inhibited. In contrast, the anti-pS259 antibody exhibited considerably lower inhibitory effects. These results indicate that pS621 domain plays a predominant role in 14-3-3⅐C-Raf binding, because both phosphospecific antibodies used exhibit similar affinities for native C-Raf proteins, as determined by SPR technique (data not shown) and the degree of phosphorylation of serines 259 and 621 are comparable (60 and 70%, respectively).
Similar effects have been observed using C-Raf phosphospecific peptides corresponding to the 14-3-3 binding domains (pS259 and pS621 peptides). Analogous to the experiment shown in Fig. 5A the C-Raf association with 14-3-3 was monitored in the presence and absence of pS259 and pS621 peptides. As demonstrated in Fig. 5B, the pS621 peptide revealed considerably higher inhibitory potential than peptide pS259 competing with an IC 50 value of ϳ100 nM. Thus, data presented in Fig. 5 suggest strongly that the 14-3-3 protein binding domain surrounding pS621 represents the high affinity and probably the major binding epitope. However, competition experiments shown in Fig. 5B) provide only relative affinities for the peptides used. Furthermore, 14-3-3 binding experiments performed with synthetic peptides do not reflect necessarily the real binding properties of the full-length proteins. In addition, the presence and exposure of two (or possibly three) 14-3-3 binding domains in C-Raf allows different interpretations of the data. Thus, to determine affinity constants in real times for Raf interactions with 14-3-3 proteins we used direct binding assays with purified Raf and 14-3-3. It was of particular interest to obtain binding data with respect to 14-3-3 binding for the isolated pS259 and pS621 binding epitopes. To determine affinity constants, GST-14-3-3 was captured again by anti-GST antibody and interactions were measured by injection of purified C-Raf wt, C-Raf-S621A, C-Raf-S259A, and C-Raf-S621A/S259A mutants. As demonstrated in Fig. 6A, intact C-Raf wt bound 14-3-3 very effectively. Apparent K D value for this interaction was estimated to be lower than 1 nM, because practically no dissociation was observed under the conditions used. Testing the C-Raf-S621A mutant, where the extent of serine 259 phosphorylation remained the same i.e. comparable to that value observed for the C-Raf wt protein (see Fig. 4) the 14-3-3 association was dramatically diminished resulting in a value of ϳ10% of that measured for C-Raf wt in the same time range, indicating lower affinity of pS259 domain for 14-3-3 proteins (Fig. 6B). The low degree of pS621 phosphorylation in C-Raf-S259A sample (15-20% of that value measured for pS259 phosphorylation, see also Fig. 4) was probably responsible for the lower degree of binding. To quantify the association rates of C-Raf-S259A and C-Raf-S621A mutants with 14-3-3, we used a higher concentration of NaCl (500 mM instead of 150 mM) in running buffer. Under these conditions a slow dissociation of bound C-Raf was measurable yielding apparent dissociation rate constants (k d ) of 4.94 ϫ 10 Ϫ7 s Ϫ1 and 6.07 ϫ 10 Ϫ7 s Ϫ1 for C-Raf-S259A and C-Raf-S621A mutants, respectively. The corresponding association rates were 9.16 ϫ 10 5 M Ϫ1 s Ϫ1 and 1.25 ϫ 10 5 M Ϫ1 s Ϫ1 thus exhibiting an association rate of ϳ7-fold higher for the C-Raf-S259A mutant. These results reflect higher affinity of pS621 domain for the 14-3-3 proteins. The cysteine-rich domain of C-Raf has been reported to represent the third 14-3-3 binding domain (28). However, the 14-3-3 binding affinity for this position seems to be very low, because we observed only little 14-3-3 binding to C-RafS621A/S259A (Fig. 6D). On the other hand, the possibility that the cysteinerich domain and the pS259 binding epitopes function effectively only in cooperation with the pS621 binding site can not be excluded. Taken together, data presented in Figs. 5 and 6 document that the 14-3-3 binding site surrounding pS621 possesses higher affinity for 14-3-3 proteins than the pS259 binding domain. However, a considerably higher degree of 14-3-3 binding by the C-Raf wild type (Fig. 6) is apparently due to the lower dissociation rates and to the difference between bivalent and monovalent binding of C-Raf to 14-3-3.

B-and C-Raf Possess Higher Binding Affinities for 14-3-3 Proteins Than A-Raf
To compare 14-3-3 binding properties of all three Raf isoforms, we also investigated associations of 14-3-3 proteins with purified His-tagged B-and A-Raf. We used first B-Raf wt preparation purified from Sf9 insect cells, which contains certain amounts of endogenous 14-3-3 proteins (according to protein staining, ϳ40 -50% of B-Raf was complexed with 14-3-3). Nev- ertheless, an effective association of B-Raf with 14-3-3 was monitored with an association rate similar to that obtained for C-Raf (Fig. 7A). To remove 14-3-3 proteins bound to B-Raf we treated B-Raf⅐14-3-3 complex with Empigen BB, a zwitterionic detergent that has been used to dissociate 14-3-3 from phosphorylated keratins (36). Although the endogenously associated 14-3-3 proteins were completely released from B-Raf using 1% Empigen BB in the washing buffer, we observed no association with 14-3-3 indicating that Empigen BB treatment either changed the B-Raf conformation or masked the 14-3-3 binding epitopes (data not shown). Dephosphorylation of B-Raf samples treated with Empigen BB can be excluded, as phos-phospecific antibodies documented presence of phosphate in these positions. Contrary to B-and C-Raf binding data, A-Raf revealed lower binding efficiency for 14-3-3 (Fig. 7A). This finding is in agreement with data reported by Yuryev et al. (37) that the N-terminal domain of A-Raf does not interact with 14-3-3 proteins. Moreover, A-Raf was shown to be selectively localized in purified rat liver mitochondria (37). Apparently, A-Raf transport into mitochondria is facilitated due to lower affinity for cytosolic 14-3-3 proteins. The specificity of B-Raf FIG. 4. Biosensor analysis of the 6B4 antibody associations with the C-Raf kinases. A, the phosphospecific 6B4 antibody was immobilized to the CM5 biosensor chip by covalent modification. Purified C-Raf wt (10 pmol) or the C-Raf Ser-259 and Ser-621 mutants (10 pmol) were injected, and the degree of association was monitored. The bars shown represent the associations reached at equilibrium. The inset shows Western blots carried out with the same C-Raf samples as used for biosensor measurements. Results shown in the bar diagram represent mean Ϯ S.D. from two independent measurements. B, kinase activity of C-Raf wt compared with C-Raf serine 259 and serine 621 mutants. The kinase assay was performed as described under "Experimental Procedures" and monitored by anti-phospho-ERK antibodies.

FIG. 5. Analysis of 14-3-3 protein association with C-Raf.
A, inhibition of C-Raf wt binding to 14-3-3 proteins in the presence of 6B4 or anti-pS259 antibodies. Purified and GST-tagged 14-3-3 protein was captured by immobilized anti-GST antibody. Approximately 1200 resonance units of GST-14-3-3 were bound for each measurement. In the next step his-tagged C-Raf wt (10 pmol) was injected at a flow rate of 5 l/min in the presence or absence of phosphospecific antibodies as indicated. In the absence of antibodies ϳ400 resonance units were bound. Phosphospecific antibodies 6B4 or anti-pS259 (1, 5, and 10 g per sample) were preincubated with 10 pmol of C-Raf for 30 min at 24°C before injection. B, inhibition of C-Raf wt binding to 14-3-3 proteins by pS621 and pS259 phosphopeptide. Binding of his-tagged C-Raf wt (10 pmol) to immobilized GST-14-3-3 was performed as described in A. Phosphopeptides pS621 and pS259 over a range of concentration between 10 Ϫ3 and 10 Ϫ9 M were mixed with C-Raf and injected without incubations. association with 14-3-3 shown in Fig. 7A was further investigated using the phosphopeptide pS621 as a competitive inhibitor, because the consensus sequence (RSXpSXP) of this 14-3-3 binding site has been shown to be identical for both B-and C-Raf. As shown in Fig. 7B the co-injection of B-Raf wt with phosphopeptide pS621 inhibited effectively the binding to 14-3-3. On the other hand, injection of 100 M phosphopeptide pS621 to the prebound 14-3-3⅐Raf complex (B-and C-Raf were investigated) caused no dissociation of Raf at all, demonstrating high affinity binding between 14-3-3 and Raf (data not shown). The association curves measured for A-, B-, and C-Raf association with 14-3-3 (Fig. 7A) determined by BIAcore technique represent only the initial part of binding process. To determine binding at equilibrium conditions, we incubated purified A-, B-, and C-Rafwt with GST-14-3-3 (and GST alone) for 4 h at 6°C and analyzed bound Raf after precipitation with glutathione-Sepharose beads. As depicted in Fig. 7C after prolonged incubation the amounts of 14-3-3 associated Raf kinases were comparable indicating that in the case of A-Raf only the initial association rate differs from values measured for Band C-Raf.

Time Course of C-Raf Phosphorylation in Response to NGF Suggests a Multistep Process
In the previous experiments we analyzed the phosphorylation status of C-and B-Raf kinases expressed in Sf9 insect cells. These data reflected the steady-state phosphorylation events of Raf kinases. To study the dynamic changes as a function of physiological state we analyzed the phosphorylation pattern of (endogenous) C-Raf in mammalian cells as a function of time in response to nerve growth factor (NGF). For that purpose we used PC12 pheochromocytoma cells and antibodies directed against the essential phosphorylation sites of C-Raf, which have been proposed to be involved in the C-Raf activation-deactivation process. In addition to the monoclonal anti-C-Raf-pS621 (6B4) and anti-pS259 antibodies, we also employed anti-pS338-and anti-pY340/pY341-C-Raf-phosphospecific antibodies, which are directed against the activation domain. PC12 cells were cultured under low serum conditions and treated with medium alone or with NGF in the time range between 10 s and 10 min (Fig. 8). After the fixation procedures the cells were incubated either with a polyclonal C-Raf antibody or with phosphospecific antibodies as indicated in Fig. 8. The polyclonal C-Raf antibody showed homogeneous distribution of C-Raf. The labeling intensity and distribution were not affected by NGF treatment. Immunostaining of the cells with anti-pS621 antibody revealed unexpected pattern of labeling. Phosphorylation of serine 621 peaked clearly 30 s after cell stimulation with NGF. After 60 s a dramatic decrease of serine 621 phosphorylation (and migration of labeled C-Raf to the plasma membrane) was observed. In the time range between 2 and 10 min the extent of Ser-621 phosphorylation remained relatively constant (see also Fig. 9). However, after 10 min the distribution of the Ser-621 phosphorylation appeared more diffuse, and some patches in the nuclear regions became visible. In contrast to serine 621, the phosphorylation of serine 259 increased already after 10 s following NGF stimulation. In the time range bemonitored. For evaluation of binding data the zero point (black arrow) was adjusted at the beginning of the Raf association. The maximal association degrees (gray arrow) are presented separately as a bar diagram. The negative control in A represents association of Raf with the immobilized GST protein. Results shown in the bar diagram represent mean Ϯ S.D. values from three independent experiments. B, inhibition of B-Raf association with GST-14-3-3 by phosphopeptide pS621. B-Raf association with 14-3-3 was measured as described in A. Indicated concentrations of phosphopeptide pS621 were co-injected with B-Raf (20 nM) without preincubations, and the degree of B-Raf⅐14-3-3 binding inhibition was monitored. C, binding of purified A-, B-, and C-Raf wt to GST-14-3-3 under equilibrium conditions. His-tagged A-, B-, or C-Raf wt (10 pmol each) were incubated with purified GST-14-3-3 (50 pmol) or with purified GST protein (50 pmol) for 4 h at 6°C. Subsequently the GST-tagged proteins were precipitated by glutathione-Sepharose beads (2 h at 6°C), and the bound Raf kinases were analyzed by SDS-PAGE and Western blotting. The amounts of bound Raf were detected by anti-penta-his antibody (Qiagen). Similar results were obtained from two independent experiments. FIG. 7. Comparison of 14-3-3 protein associations rates with A-, B-, and C-Raf monitored by SPR technique. A, purified GST-14-3-3 was immobilized to the sensor chip as described in Fig. 9. A-, B-, and C-Raf wt (20 nM each) were injected immediately after GST-14-3-3 deposition (ϳ1200 resonance units for each run), and association was tween 20 s and 2 min Ser-259 phosphorylation decreased slightly but continuously. After 10 min, Ser-259 phosphorylation increased considerably. These results reflect again the different properties and functional roles of 14-3-3 binding domains surrounding phosphoserine 621 and 259.
With respect to C-Raf activation particular importance has been assigned to the Ras-induced phosphorylation of serine 338 and Src-dependent phosphorylation of C-Raf tyrosine residues at positions 340/341 (18,38). We have analyzed the phosphorylation pattern of these residues using the corresponding phosphospecific antibodies. Whereas serine 338 was only weakly phosphorylated at the 30-and 60-s time points, and quantification of the staining was not possible due to the low intensity of the signals, phosphorylation of tyrosine 340/341 was more pronounced reaching a maximum at 60 s thus correlating with the peak of kinase activity of C-Raf (see Figs. 8 -10). Interestingly, the distribution of phosphotyrosine 340/341 was not homogenous in the early stages of stimulation. Already 20 s after stimulation we observed microlocalization of tyrosine 340/ 341 phosphorylation visualized as punctate staining. At the 30and 60-s time points we observed additional staining, which localized closer to the plasma membrane. After 10 min only diffuse labeling pattern remained. A kinase activity profile of C-Raf following NGF treatment of PC12 cells is shown in Fig.  10. C-Raf was immunoprecipitated by a polyclonal anti-C-Raf antibody, and kinase activity was detected in the presence of purified MEK and ERK. As depicted in Fig. 10, NGF induced rapid and transient activation of C-Raf after 1-2 min. Notably, the phosphorylation of the endogenous ERK, detected directly from cell lysates using anti-pERK antibody, revealed very similar profile (Fig. 10), indicating that C-Raf activation does not interfere with activation of B-Raf in the short time range of PC12 cell activation (0 -120 s). These data are in agreement with our previous observations (39). Taken together, use of the monoclonal 6B4 and other C-Raf phosphospecific antibodies allowed to elucidate in more details the time course of endogenous C-Raf activation in mammalian cells.

DISCUSSION
For examination of dynamic changes in Raf phosphorylation and subcellular localization in the course of growth factorinduced activation, we developed a phosphospecific Raf monoclonal antibody (6B4) directed against C-Raf serine 621 possessing high affinity for native C-Raf but not for B-Raf. Kinetic parameters and affinity constants for 6B4 antibody binding to purified C-Raf have been determined using the biosensor technique. Use of 6B4 antibody allowed us to determine the phosphorylation pattern in C-and B-Raf preparations dependent on activation status and to elucidate the nature of 14-3-3 binding to different C-Raf interacting motifs. Moreover, time course of endogenous C-Raf activation in PC12 cells detected by 6B4 and other phosphospecific antibodies revealed new insights into C-Raf phosphorylation events during the activation cycle.
Using the phosphospecific 6B4 antibody we investigated first the degree of serine 621 phosphorylation in C-Raf wt and diverse forms of activated C-Raf kinases as well as C-Raf mu- tants, such as C-Raf-S621A, C-Raf-S259A, Ras-decoupled C-Raf-R89L, and kinase-dead C-Raf-K375W. Our data show that both highly activated and unstimulated C-Raf possess the same degree of serine 621 phosphorylation (ϳ70%). In contrast, a kinase-dead mutant of C-Raf was not phosphorylated on serine 621 indicating a requirement for intrinsic kinase activity. These data confirm and extend results reported by Thorson et al. (40) who showed that truncated C-Raf lacking regulatory domain was not phosphorylated at serine 621 in the case of kinase-deficient mutation. Additionally, we demonstrate that B-Raf possesses other properties in this context, i.e. the kinasedead mutant of B-Raf was phosphorylated in position serine 728 to the same degree as the wild type form (see Fig. 3). Serine to alanine exchange in position 259 strongly inhibited phosphorylation of serine 621 in vivo demonstrating that the mutant protein is a poor substrate for Ser-621 kinase(s). Furthermore, contrary to serine 621, the phosphorylation of serine 259 does not depend on endogenous Raf kinase activity or exchange of serine 621.
Finally, we demonstrate here for the first time that 14-3-3 binding affinities for the C-Raf binding sites surrounding phosphoserine 259 and 621 differ strongly from each other. These data were obtained by both direct binding assays using purified A-, B-, and C-Raf (Figs. 6 and 7) and by competition assays (Fig.  5, A and B). To our knowledge, no kinetic data on 14-3-3 protein associations with purified and full-length A-, B-, and C-Raf kinases have previously been published. Only the interaction of 14-3-3 proteins with phosphopeptide pS259 has been investigated (27). Using SPR measurements Muslin et al. obtained for this interaction K D values of ϳ120 -140 nM. Our binding studies presented here carried out with full-length and functional A-, B-, and C-Raf kinases combined with competition assays demonstrate unambiguously that Raf proteins bind 14-3-3 with high affinity and that the binding kinetics for epitopes surrounding pS259 and pS621 in C-Raf differ significantly from each other (k a value for pS621 binding to 14-3-3 is more than 7-fold higher than that for pS259). Therefore, our data support findings showing that C-Raf dephosphorylation in position 259 may preferentially occur in the course of C-Raf activation (10,11). Due to lower binding affinity between 14-3-3 proteins and pS259 epitope the dephosphorylation at this position appears indeed more probable than dephosphorylation at pS621. Moreover, our finding that pS621 epitope represents the high affinity binding site is in agreement with our previous results, which showed that the intact and phosphorylated pS621 binding site (but not pS259) is necessary and sufficient for Raf heterodimerization and maintenance of Raf activity (41). Simultaneous binding of 14-3-3 dimers to multiple protein binding sites has been found not only for Raf kinases. Proteins such as Bad, Cdc25, Cbl, and KSR possess also two recognition motifs for 14-3-3 proteins (21).
Whereas C-Raf expressed in Sf9 insect cells possesses a were treated with 50 ng/ml NGF for the times indicated. C-Raf kinase was immunoprecipitated from the cell lysates, and in vitro kinase assay was performed in the presence of MEK and ERK (filled bars) as described under "Experimental Procedures." To determine the activity of endogenous ERK in response to NGF (open bars) cell lysates (ϳ20 g of protein) were applied to SDS-PAGE and immunoblotted to nitrocellulose membrane. The detection of both endogenous and purified ERK phosphorylation was performed by anti-pERK antibody and ECL. The signals were quantified by scanning laser densitometry. No timedependent changes in amount of immunoprecipitated C-Raf were observed (data not shown). relatively high degree of serine 621 and 259 phosphorylation (60 -70%), representing the steady-state situation, the immunostaining of endogenous C-Raf in mammalian cells revealed that the degree of phosphorylation in these positions clearly depends on hormonal stimulation of the cells. As demonstrated in Figs. 8 and 9 the phosphorylation degree of Ser-621 and Ser-259 in unstimulated cells appear to be very low. In contrast to Ser-259 phosphorylation that reached maximal value at the 10-min time point, Ser-621 phosphorylation peaked already after 30 s of cell stimulation. Interestingly, Ser-621 phosphorylation dropped down after 60 s, although the kinase activity increased during the same time (Fig. 10). We suggest that this observation reflects the formation (and rearrangements) of a large C-Raf complex, including 14-3-3, KSR, CNK, MEK, and ERK. The formation of the large C-Raf signalosome may reduce accessibility of phosphospecific antibody 6B4 for the pS621 epitope. A moderate increase of serine 259 phosphorylation was detectable already 10 s after cell stimulation. This observation confirms our in vitro data shown in Fig. 4, which demonstrate interdependence between serine 259 and serine 621 phosphorylation. On the other hand, the increase of serine 259 phosphorylation after prolonged stimulation (10 min) correlates with the decrease of C-Raf kinase activity and support the formation of the inactive and locked C-Raf⅐14-3-3 complex. Using phosphospecific antibody against phosphotyrosine 340/ 341 we observed very early (beginning at 20 s after NGF stimulation) the formation of characteristic patches. The phosphorylation of tyrosine 340/341 has been shown to occur by tyrosine kinases belonging to the Src family (18). The doubly acylated tyrosine kinases of the Src family have been found to associate with rafts and are incorporated into detergent-resistant membranes (42). Previously, we demonstrated that C-Raf exhibits high affinities for artificial vesicles resembling rafts (8). Based on these data we suggest that C-Raf co-localizes with the raft microdomains containing Src kinases and that C-Raf phosphorylation by Src occurs primarily in raft microdomains as a very early event of C-Raf activation. In summary, we present here time-and growth factor-dependent resolution of C-Raf phosphorylation events at very early time of cell stimulation using a cell culture system, in which we previously demonstrated that C-Raf activation is rapid and transient (39).
Based on data presented here we conclude that (i) the serine 621 phosphorylation in C-Raf, in contrast to B-Raf, requires both endogenous kinase activity and phosphorylation in position 259 (cooperativity of these two regulatory sites), (ii) the 14-3-3 protein binding domain surrounding C-Raf-pS621 represents the high affinity binding site, whereas the pS259 epitope mediates lower affinity binding, (iii) the rate of 14-3-3 association with B-Raf does not differ substantially from that with C-Raf, however, A-Raf association with 14-3-3 reveals lower affinity compared with B-and C-Raf, (iv) phosphorylation of serine 259 represents an early event in the course of To transfer the initial hormone signal further the active C-Raf signalosome couples with MEK and ERK in the presence of KSR (e). The C-Raf inactivation pathway involves rephosphorylation of serine 259 and reorganization to the locked C-Raf (f). A fraction of locked C-Raf may associate with specific membrane lipids (g). In this context we did not discuss the multicomplex formation with additional structure, scaffold, and adapter proteins such as HSP50, HSP90, p50/Cdc37, CNK, SUR-8, BAG/Bcl2, and G␤␥ subunits.