protein kinase c-mediated serine phosphorylation directly

THE JOURNAL OF BIOUGICAL CHEMISTRY 0 1994 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 269, No. 2, Issue of January 14, pp. 1249-1256, 1994 Prinfed in U.S.A.

Protein Kinase C-mediated Serine Phosphorylation Directly Activates Raf-1 in Murine Hematopoietic Cells*

(Received for publication, July 2, 1993, and in revised form, September 21, 1993)

Michael P. Carroll$ and W. Stratford May$ From The Johns Hopkins Oncology Center, Baltimore, Maryland 21287

We have previously found that Raf-1, which is activated by hematopoietic growth factors in association with phosphorylation, is required for hematopoietic cell proliferation. Recently, 12-O-tetradecanoylphorbol 13- acetate has been found to mediate Raf-1 phosphorylation, suggesting that protein kinase C (PKC) may be involved in the Raf-1 activation mechanism(&. Since PKC can be activated by hematopoietic growth factors, it was investigated as a potential “Raf-1 kinase-kinase.“ Re- sults demonstrate that bryostatin 1, a pharmacologic activator of PKC, induces activation of Raf-1 in FDC-P1 cells. PKC inhibitors H7 and staurosporine block both bryostatin 1- and interleukin-3-mediated Raf-1 phosphorylation and FDC-P1 cell proliferation. Additionally, an antisense c-raf oligodeoxyribonucleotide specifically inhibits bryostatin l-mediated proliferation, indicating a necessary role for Raf-1 in PKC signaling. Purified PKC can phosphorylate Raf-1 serine residues to high stoichiometry in vitro. Comparative phosphopeptide maps localize two PKC phosphorylation sites to Raf-1 phosphopeptides isolated from hematopoietic growth factor- or bryostatin l-stimulated cells. The sites of PKC- mediated Raf-1 phosphorylation are deduced to be SefiB7 and SePB. Furthermore, PKC-mediated serine phosphorylation is sufficient to activate the enzymatic function of Raf-1 in vitro. These findings demonstrate that activated PKC can promote hematopoietic cell growth by regulating the enzymatic activity of Raf-1 through direct serine phosphorylation.

The growth and development of the blood elements is regu- lated by polypeptide growth factors (1,2). The majority of these ligands, including IL-3,’ GM-CSF, and erythropoietin, bind with high affinity to specific transmembrane proteins which belong to the recently identified cytokine receptor superfamily (3). Unlike other classes of growth factor receptors, cytokine receptors apparently lack enzymatic activity (4). Thus, although receptor engagement can lead to hematopoietic cell

* This work was supported by a starter grant from the American Cancer Society, National Institutes of Health Grants CA 44649, CA 47993, CA 01679, and a grant from the R. W. Johnson Pharmaceutical Research Institute. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “uduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ Fellow of the Leukemia Society of America. 5 Scholar of the Leukemia Society of America. To whom correspond-

Bond St., Baltimore, MD 21231. ence should be addressed: The Johns Hopkins Oncology Center, 424 N.

The abbreviations used are: IL-3, interleukin-3; GM-CSF, granulo- cyte-macrophage colony-stimulating factor; IL-2, interleukin-2; PKC,

(W kinase or erk) kinase; TPA, 12-O-tetradecanoylphorbol 13-ac- protein kinase C; MAP kinase; mitogen-activated protein kinase; MEK,

etate; FSBA, p-fluorosulfonylbenzoyl 5“adenosine; PAGE, polyacrylamide gel electrophoresis; cpm, countdmin.

growth, the mechanism by which this proliferative signal is transmitted to the replicative machinery in the nucleus is unclear.

Despite the lack of enzymatic activity, recent evidence indicates that ligation of cytokine receptors results in the rapid activation of intracellular protein kinases (2). Several studies have reported that hematopoietic growth factors including IL-3, GM-CSF, IL-2, and erythropoietin can rapidly induce the phosphorylation of cellular proteins at both serine and tyrosine residues (5-10). Tyrosyl phosphorylation may occur via the non-covalent association of cytokine receptors, such as the IL-2 receptor &chain, with cytoplasmic protein kinases, such as p56ICk (11). Furthermore, protein tyrosine kinase activity may be necessary to mediate hematopoietic cell growth (12). Serine/ threonine-specific protein kinases, including PKC, Raf-1, and MAP kinase also undergo rapid activation in response to hematopoietic growth factors (6, 13-18). We and others have recently shown that Raf-1, the ubiquitously expressed product of the c-ruf-1 protooncogene, is enzymatically activated, in association with serine and tyrosine phosphorylation, after hematopoietic cell stimulation with IL-3, GM-CSF, IL-2, or erythropoietin (15-17). Furthermore, Raf-1 is required for IL-3 and erythropoietin-induced proliferation (17). Interestingly, Raf-1 appears to be involved in a cascade of post-receptor phosphorylation events involving MAP kinase-kinase (also known as MEK) and MAP kinase. Thus, Raf-1 can directly activate MEK as a result of serine and/or threonine phosphorylation, and activated MEK in turn can activate MAP kinase through direct phosphorylation of tyrosine and threonine residues (19-20). Recently, several proteins which regulate cell growth have been found to be direct substrates for activated MAP kinase. These include plasma membrane proteins such as the epidermal growth factor receptor and phospholipase A2 (21, 221, a cytoplasmic serinekhreonine ribosomal protein S6 kinase, p9OrBk (23) and nuclear transcription activators c-Jun and c-Myc (24, 25). Therefore, current evidence indicates that Raf-1, a cytosolic enzyme necessary for growth, can initiate a sequence of phosphorylation reactions, which may transmit the mitogenic signaKs) to the plasma membrane, cytoplasm, and nucleus.

The mechanism(s1 of growth factor-mediated Raf-1 activation are not known. However, recent evidence indicates that Raf-1 activation may occur as a result of the phosphorylation of serine residue(s). Several growth factors appear to mediate Raf-1 activation in association with serine phosphorylation alone (26-28). Additionally, after insulin stimulation, Raf-1 can be inactivated by a serine-specific phosphatase (26). These data suggest that a distinct protein serine kinase may directly phosphorylate Raf-1 and further, that such phosphorylation may be sufficient for Raf-1 activation. The identity of such a “Raf-1 kinase-kinase(s)” is however unknown, but one likely candidate is PKC.

Physiologic growth stimuli, including the hematopoietic growth factors IL-3, GM-CSF, and erythropoietin can rapidly activate members of the calcium, phospholipid-dependent pro-

1249

1250 PKC Activates Raf-1

tein serinelthreonine kinase family known collectively as PKC (6,13,14). We have recently demonstrated that bryostatin 1, a pharmacologic activator of PKC, can mimic the effects of IL-3 on the factor-dependent cell line FDC-P1 (6). Furthermore, PKC appears to be necessary for IL-3 and erythropoietin-induced proliferation (29). Additionally, pharmacologic activators of PKC, TPA, and bryostatin 1, can mediate Raf-1 activation (29,30), suggesting a potential physiologic interaction between these enzymes. Therefore, studies were performed to determine if PKC could directly phosphorylate and activate Raf-1, thus providing evidence for PKC as a Raf-1 kinase-kinase.

EXPERIMENTAL PROCEDURES Cell Lines, Culture Media, and Viral Inoculation-FDC-P1, FDC-P1/

ER, and Spodoptera frugiperdu cells were maintained as described (17, 31). Wild-type baculovirus was provided by Max D. Summers and Gale E. Smith (Department of Entomology, Texas A and M University) (31). Baculovirus, containing a recombinant expression vector which in- cludes the full coding sequence of human c-raf-1, was a generous gift of Ulf R. Rapp (National Cancer Institute, National Institutes of Health, Frederick Cancer Research and Development Center, Frederick, MD). Confluent cultures of S. frugiperda cells were inoculated with recombinant, c-raf expressing, baculovirus, or wild-type control at a multiplicity of infection of 10 for 2 h. Cells were harvested 40 h later, at the time of peak baculoviral protein accumulation, and lysed (31). Raf-1 protein expression was detected in baculovirus-infected cell lysates by Western blotting (17).

Chemicals and Radioisotopes-Radioisotopes were obtained commer- cially (Amersham Corp.). Recombinant erythropoietin was provided by Ortho-Biotech. Synthetic mouse IG3 was kindly furnished by I. Clark- Lewis (University of British Columbia). PKC was purified from rat brain (6). Bryostatin 1 was provided by R. Pettit (University ofArizona). Oligodeoxyribonucleotides were prepared as described (17). Raf-1 peptides were custom synthesized and purified (Bachem or Johns Hopkins University).

Metabolic Labeling and Immune Precipitation-Cultures equilibrated with [32P]orthophosphate were treated for 15 min with bryostatin 1 , l pg/d IL-3 (FDC-Pl), or 4 unitdml erythropoietin (FDC- Pl/ER), washed, and lysed (15). Immunoprecipitates were prepared from whole cell lysates as described (29). Where indicated, 32P-labeled Raf-1 was excised from polyacrylamide gels and subjected to phosphoamino acid and phosphopeptide analyses (6,32).

Protein Kinase Inhibitor Studies-FDC-P1 cells were incubated without mitogen at 37 "C in RPMI-1640 containing 1% (v/v) fetal bovine serum together with or without H7 (33) or staurosporine (34). ARer 16 h, control cells were washed three times with RPMI-1640 to remove the inhibitor. Cells pretreated with inhibitors were aliquoted into 96-well plates at 5 x lo4 celldwell and treated with M bryostatin 1,1 p g / d IL-3 (50 unitdml), or 0.1% MezSO as a control. After an 8-h incubation, cells were pulsed for 16 h with 1 pCi/well L3H1thymidine. [3H1Thymidine incorporation was determined as described (17). 50% inhibitory concen- trations (IC5,,) were estimated from data averaged over three quadruplicate experiments. For Raf-1 phosphorylation assays, cells were equilibrated with [3zPlorthophosphate in the presence or absence of inhibitor. Labeled cultures were treated for 15 min with 0.1% MezSO control, M bryostatin-1, or 1 pg/ml IL-3. Raf-1 immune precipitates were prepared and analyzed as above. Raf-1 phosphorylation was assessed by optical densitometry.

Deatment of Cells with Olig&oxyribonucleoti&s-FDC-Pl cells were incubated in serum-free RPMI-1640 alone or with 10 p~ antisense or "scrambled" nonsense control oligomer for 19 h in 96-well microliter plates as described (17). Cultures were treated for 6 h with 1 p & d IL-3 or M bryostatin 1 and pulsed with 1 pCi/well L3H1thymidine for 16 h. [3H]Thymidine incorporation was determined as above.

In Vitro Kinase Reactions-Recombinant human Raf-1 was isolated by immune precipitation from infected S. frugiperda cell lysates as described (29). Washed Raf-1 immunoprecipitates were treated with 500 p~ p-fluorosulfonylbenzoyl 5"adenosine (FSBA) or 1% MezSO as a control for 30 min at 4 "C in 50 m Tris, pH 7.0, containing 10% glycerol and 100 m magnesium chloride (35,361. Immune complexes, containing Raf-1 or the FSBA-inactivated control, were resuspended in 50 pl of kinase buffer containing 50 m Tris, pH 7.2,50 m sodium chloride, 1% (v/v) glycerol, 10 m magnesium chloride, and 2 n m dithiothreitol with or without 100 p ATP, 10 pCi of [y-32PlATP, 2 pl of rat brain PKC, 40 p&ml phosphatidyl serine, 1 p~ TPA or 100 n~ bryostatin 1, 800 p~ calcium chloride, and 2 nm EGTA. Mixtures were incubated for 30 min

at 37 "C. Reactions were terminated by the addition of 25 m EDTAand immune complexes were washed five times in lysis buffer. After "priming" Raf-1 using unlabeled ATP and PKC, or after the treatment of FDC-P1 cells with bryostatin 1, Raf-1 immune precipitates were used in an in vitro kinase reaction, containing 20 pCi of [+"P]ATP with or without 0.5 mg/d histone 111s (Sigma), as an exogenous substrate (15). Raf-1 and histone phosphorylation were assessed after SDS-PAGE by Cerenkov counting of excised bands (17). Nonspecific histone phosphorylation was controlled using Raf-1 immune precipitates prepared with excess peptide antigen (20 pg) (15). In order to assess PKC-mediated Raf-1 phosphorylation, primary reactions including [y-32PlATP were resolved by 7.5% SDS-PAGE. The protein content was estimated by comparing Coomassie staining of Raf-1 bands to known concentra- tions of bovine serum albumin as standards. Phosphorylated Raf-1 bands were excised from the gels and subjected to phosphoamino acid and phosphopeptide analysis as described (15, 32).

Raf-1 peptides were dissolved in kinase buffer in 50 pl of reaction mixtures containing 2 pl of purified rat brain PKC, 100 p~ ATP, 10 pCi of [y-3zPlATP with or without 1 p~ TPA, 40 pg/ml phosphatidylserine, 800 p~ calcium chloride, or 2 m~ EGTA for 1, 5, 15, 30, or 60 min at 37 "C. R1, R2, R4, and R5 phosphorylation was assessed after adsorp- tion of the peptide to phosphocellulose discs (26). R3 phosphorylation was determined by Cerenkov counting after trichloroacetic acid precipitation. Michaelis constants for PKC-mediated peptide phosphorylation were determined using double-reciprocal plots. Phosphorylated peptides were purified for further analysis (32). Briefly, 32P-labeled peptides were visualized by autoradiography after thin layer electrophoresis (Rl, R2, R4, R5) or SDS-PAGE (R3) and excised. Phosphopeptides were eluted from silica gel using 500 pl of aqueous buffer containing 4.4% (v/v) formic acid and 7.8% (v/v) acetic acid and from polyacrylamide using 50 m aqueous ammonium bicarbonate. Lyophilized peptides were digested with trypsin for two-dimensional phosphopeptide analysis or analyzed for phosphate release by manual Edman degradation as described (32).

RESULTS AND DISCUSSION

Activated PKC Mediates R a f l Phosphorylation and Enzy- matic Activation-Hematopoietic growth factors, including IL-3, GM-CSF, and erythropoietin, can activate PKC, probably by inducing the generation of diacylglycerols from membrane phospholipids (6, 13, 14, 37). The macrocyclic lactone bryostatin 1 is a potent activator of PKC which can support the growth of factor-dependent FDC-P1 cells in lieu of IL-3 (6). Therefore, activation of PKC may be sufficient for hematopoietic growth factor-induced proliferation. Additionally, rapid enzymatic activation of Raf-1 occurs in close association with hematopoietic growth factor-mediated Raf-1 hyperphospho- rylation (15, 17), and Raf-1 appears to be necessary for hematopoietic cell growth (17). Studies were performed to determine a potential role for activated PKC in Raf-1 regulation. Bryostatin 1 induces both phosphorylation and activation of Raf-1 in FDC-P1 cells in a dose-dependent manner (Fig. 1). Bryostatin 1-mediated Raf-1 activation is maximal by 10 min, thus mimicking the activation kinetics seen for growth factors including IL-3, GM-CSF, IL-2, erythropoietin, insulin, colony- stimulating factor-1, epidermal growth factor, and platelet-de- rived growth factor (15-17,26-28,30). These findings demonstrate that bryostatin 1-mediated Raf-1 phosphorylation and enzymatic activation are closely associated with the mitogenic effects of bryostatin 1. The degree of bryostatin 1-induced Raf-1 phosphorylation generally correlates with enzymatic activation, suggesting that phosphorylation may be involved in the Raf-1 activation mechanism(s). However, some discordance between phosphorylation and activation is apparent at low con- centrations of bryostatin 1. Thus, 1 x 10"O M bryostatin 1 induces only 8% of the maximal increase in Raf-1 phosphorylation (as estimated by optical densitometry) and 45% of the maximal increase in Raf-1-mediated histone phosphorylation (Fig. 1). The explanation for this difference is unclear, however, several possibilities exist. Since mitogen-induced Raf-1 phosphorylation appears to occur at multiple sites (Figs. 6 and 81, it is possible that only one or a subset of these events participate

PKC Activates Raf-1 1251

116.

97.

66.

9 10

T

”1 T T T

5 10 15 30 60

rnin

FIG. 1. Bryostatin 1 mediates Raf-1 phosphorylation and enzymatic activation in FDC-P1 cells. IG3 was withdrawn from FDC-P1 cells for 16 h. Upper panel, cells were equilibrated with [”P]orthophosphate and treated with the indicated concentration of bryostatin 1 (BRYO) for 15 min. Raf-1 immune precipitates were resolved by 7.5% SDS-PAGE and analyzed by autoradiography. The arrowhead indicates the position of Raf-1. Center and lower panels, cells were treated for 15 min with bryostatin 1 at the indicated concentration (center) or incubated with 100 n.~ bryostatin 1 for the times shown (lower). Raf-1 and control immune precipitates were added to in vitro kinase reaction mixtures containing histone 111s and [Y-~~PIATP as described under “Experimental Procedures.” Phosphorylated histone bands were excised from polyacrylamide gels and counted by the method of Cerenkov. Re- sults are expressed as the mean percentage = S.E. of unstimulated control values (mean 1.6 x lo3 cpm) for three separate experiments after correcting for nonspecific background (mean 7 x lo2 cpm).

in the activation mechanism(s). Preferential mitogen-induced phosphorylation of this putative activation site(s) could explain the observed difference between overall Raf-1 phosphorylation

CONT BRYO PKC FIG. 2. Bryostatin 1- and PKC-mediated Rat-1 phosphorylation

occurs at serine residues. “P-Labeled Raf-1 from control (CONT) or bryostatin 1-stimulated (BRYO) FDC-P1 cells or from Sf9 cells after an in vitro kinase reaction with purified PKC was excised from polyacryl-

under “Experimental Procedures.” Markers indicate the locations of amide gels and subjected to phosphoamino acid analysis as described

phosphoamino acid and free phosphate standards. PO,, free phosphate; Pser, phosphoserine; Pthr, phosphothreonine; Ptyr, phosphotyrosine.

and activation. Alternatively, additional mechanisms of mitogen-induced enzymatic activation may exist which are distinct from Raf-1 phosphorylation.

Bryostatin 1-induced phosphorylation of Raf-1 occurs exclusively a t serine residues (Fig. 2) and as such is distinct from the serine ana’ tyrosine phosphorylation observed with IL-3, GM- CSF, and erythropoietin (15, 17) or IL-2 (16). Therefore, these results suggest that PKC-mediated serine phosphorylation could be sufficient for Raf-1 activation. In support of these data, Morrison et al., (30) recently found that TPA can also stimulate Raf-1 activation in NIH3T3 cells.

PKC Inhibitors Block Both Bryostatin 1- and ZL-3-mediated Proliferation as Well as Raf-1 Phosphorylation-Since PKC activation by bryostatin 1 (6) or TPA (38) is associated with hematopoietic cell proliferation, we evaluated the effects of the widely used specific PKC inhibitors, H7 (33) and staurosporine (34), on FDC-P1 cell growth as well as Raf-1 phosphorylation and activation. Results indicate that both H7 and staurosporine inhibit IL-3- and bryostatin 1-mediated proliferation a t doses compatible with a PKC-specific effect (Fig. 3). Indeed, the ICso are very similar for these two mitogens (Table I). These findings support a role for PKC activation in hematopoietic growth signaling. Additionally, both IL-3- and bryostatin l-mediated Raf-1 phosphorylation can be inhibited by these drugs (Fig. 3). No significant effects on Raf-1 phosphorylation or [3H]thymidine incorporation are seen when the drugs are removed immediately prior to mitogenic stimulation, ruling out any irreversible cytotoxicity that could represent a trivial explanation for these findings (Fig. 3). Importantly, the IC50 for both H7 and staurosporine with respect to Raf-1 phosphorylation is similar to that observed for inhibition of mitogenesis, consistent with a regulatory role for PKC in Raf-1 activation and growth signaling (Table I). Thus, Raf-1 phosphorylation appears to represent a biochemical correlate of mitogenicity.

Raf-1 Is Required for Bryostatin 1-mediated Proliferation of FDC-PI Cells-A c-raf antisense oligodeoxyribonucleotide can be used to selectively deplete intracellular Raf-1 (17). To directly examine a role for Raf-1 in bryostatin 1-mediated growth signaling, cells were incubated with 10 p~ antisense c-raf for 19 h as described (17). Results demonstrate that the mitogenic effect of bryostatin 1 is inhibited by 71% (Fig. 4). By contrast, no significant effect is observed in cells incubated with a control

PKC Activates Raf-1

97 +

FIG. 3. PKC inhibitors block both bryostatin-1- and &%mediated proliferation as well aa Raf-1 phosphorylation. FDC-PI cells were incubated without mitogen with the indicated concentration of inhibitor for 16 h. Cultures were treated with 100 n~ bryostatin 1 ( le f t ) or 1 pg/ml IL3 (right) and assessed for ["]thymidine incorporation and Raf-1 phosphorylation as described under "Experimental Procedures." Asterisks indicate control samples in which the indicated concentration of inhibitor was removed from the culture after the 16-h incubation, prior to mitogen addition. Results shown are the mean of three separate experiments performed in triplicate ([3H]thymidine incorporation) or representative of three independent experiments (Raf-1 phosphorylation).

TABLE I Estimated IC,, for mitogen-induced PHlthymidine incorporation and

Raf-1 phosphorylation in FDC-PI cells

Inhibitor Mitogen ICso PHldT ICso [32PlRaf-1

H7 Bryostatin 1 10-25 1.1~ 10-25 p~ IL3 25-50 1.1~ 25-50 1.1~

Staurosporine Bryostatin 1 10-50 n~ 5-10 nM IL3 10-50 rn 10-50 rn

nonsense oligomer, demonstrating specificity for Raf-1. These findings suggest that Raf-1 may be positioned downstream of PKC in the growth signal pathway and, further, that PKC- initiated growth may be dependent on physiologic intracellular levels of Raf-1.

PKC Can Directly Phosphorylate Raf-1-Results demonstrate that PKC may act proximal to Raf-1 in the growth signal pathway(s). Therefore, we investigated a possible direct role for PKC in the mechanism(s) of Raf-1 phosphorylation. Purified rat brain PKC, which contains the classical PKC isoforms a, p, and y (391, can phosphorylate immunopurified, baculoviral human Raf-1 in vitro (Fig. 5, upper panel, lane 3). Phosphoryla- tion is found to be a calcium, phosphatidylserine, and PKC activator-dependent process, indicating a role for a classical PKC isofonn (40) (lane 4) . Baculoviral Raf-1 alone displays weak autophosphorylation in vitro, consistent with unstimulated Raf-1 enzyme activity (lane 5). Treatment of Raf-1 immunoprecipitates with 0.5 mM FSBA for 30 min inhibits autoki- nase activity by 85% as determined by Cerenkov counting (Fig. 5, lower panel, data not shown). However, PKC can effectively phosphorylate Raf-1 which has been irreversibly inactivated by FSBA, demonstrating that this is a direct effect of activated PKC and not the result of Raf-1 autophosphorylation. (Fig. 5, lower panel). The stoichiometry of maximal PKC-mediated Raf-1 phosphorylation was estimated to be 2.7. Furthermore,

'"1

25

T

Ns As Ns As

L BRYO 11-3 A

cally inhibits the mitogenic effects of bryostatin 1 in FDC-PI FIG. 4. An antisense c-raf oligodeoxyribonucleotide specifi-

cells. Growth factor-deprived FDC-PI cells were treated for 19 h with 10 1.1~ antisense (As) or nonsense control (Ns) oligomers, stimulated with 1 n~ bryostatin 1 or 1 pg/ml IL3 and assessed for L3H1thymidine incorporation as described under "Experimental Procedures." Results are expressed as percent inhibition compared with mean control values (IL3 = 31,000 cpm; bryostatin 1 = 7,200) and represent the mean * S.E. of three separate experiments performed in quadruplicate.

phosphoamino acid analysis reveals that PKC modifies Raf-1 exclusively at serine residues similar to Raf-1 phosphorylation observed when bryostatin 1 is added to cells (Fig. 2). These results demonstrate that Raf-1 is a direct substrate for PKC in vitro, suggesting a potential physiologic role for serine phosphorylation. Comparative tryptic phosphopeptide analysis carried out after the phosphorylation of FSBA-inactivated Raf-1 indicates that activated PKC phosphorylates serine residues in vitro which are contained within three of the seven major Raf-1 phosphopeptides isolated from I L 3 or bryostatin 1-treated

4200 ’. 4 1 1

-87

PKC Activates Raf-1

IN VIVO

1253

D D

L

. -, . . . , 4 4 3

W f RAF 1 2 3 4 5 6 7

IN VITRO

C o K n O L

11s - QI-

MIX

66-

1 2 3 4

FTG. 5. Upper panel, PKC can directly phosphorylate Raf-1. 1 x lo6 Sf9 cells were inoculated with wild-type (WT) or recombinant c-raf cDNA containing (W) baculoviruses. WT (lane 1 ) or RAF (lanes 2 4 ) infeded cell lysates were prepared 48 h aRer inoculation and immune precipitated with Raf-1 antiserum with (lane 2) or without (lanes I , 3 4 ) excess Raf-1 peptide. Washed immune precipitates were resw- pended in an in vitro kinase reaction mixture containing 2 pl of purified rat brain PKC (lanes 14), 2 nm EGTA (lane 4) , 100 rn bryostatin, 100 p~ ATP 800 p~ CaCI2, 10 m~ MgCI2, 40 pg/d phosphatidylserine, and 20 pCi of [-ps2PIATP. The PKC preparation without immune precipitate was added to an in vitro kinase mixture as above with (lane 8) or without (lane 7) 2 nm EGTA. After 30 min at 37 “C, reactions were terminated and the immune precipitates were washed five times in 50 column volumes of ice-cold lysis buffer. Proteins were resolved by 7.5% SDS-PAGE and analyzed by autoradiography. Additionally, samples of infected Sf9 cell lysates, equivalent to 50 pg of protein, were simulta- neously resolved by 7.5% SDS-PAGE, transferred to a nitrocellulose membrane, and probed with the Raf-1 antiserum (far left). Raf-1 was detected with 1261-protein A. Lower panel, Raf-1 containing immune precipitates were prepared as above and treated with 0.5 nm FSBA (lanes 3 and 4 ) or 1% Me2S0 as a control (lanes 1 and 2) as described under ”Experimental Procedures.” ARer washing, the treated immune

ATP, 10 pCi of [ys2P1ATP, 800 p~ CaCI2, 40 p g / d phosphatidylserine, precipitates were added to an in vitro kinase reaction containing 100 p~

1 p~ TPA, 10 m~ MgC12 with (lanes 2 and 4 ) or without (lanes 1 and 3) 2 pl of PKC for 30 min at 37 “C. After the reaction, labeled immune precipitates were washed, eluted, and resolved by 7.5% SDS-PAGE as described under Txperimental Procedures.”

cells (Fig. 6). These findings support a physiologic role for Raf-1 phosphorylation directly by PKC in vivo. However, PKC does not phosphorylate Raf-1 in vitro at all the sites detected after mitogenic stimulation of cells, suggesting that additional cellular factors may be necessary for PKC to phosphorylate these residues in vivo or that a distinct kinase(s) may be responsible for phosphorylation of these sites. Conversely, purified PKC does phosphorylate several peptides which appears to be

FIG. 6. Comparative phosphopeptide analysis of Rat-1 from whole cells and after direct phosphorylation by PKC in vitro. S2P-Labeled Raf-1 from control, IL-3, or bryostatin 1-stimulated FDC-P1 cells (IN W O ) or aRer the in vitro reaction with or without (control) PKC (IN VITRO) was excised from.polyacrylamide gels and digested with ~1-tosylamide-2-phenylethyl chloromethylketone- treated trypsin. Raf-1 was inactivated by FSBA prior to phosphorylation as described under “Experimental Procedures.” Washed digests were resolved by thin layer electrophoresis a t 800 V for 1 h in buffer containing by volume 1 part pyridine, 2 parts acetic acid, 8 parts ac- etone, and 40 parts water (vertical). Digests were developed in the

volume 15 parts n-butanol, 3 parts acetic acid, 10 parts pyridine, and 12 second dimension by ascending chromatography in buffer containing by

parts water (horizontal). Phosphopeptides were detected by autoradiography. The origins are located at the upper Tight corner of each au- toradiograph. For direct comparison, digests from in vivo phosphorylated Raf-l(100 cpm) were mixed with in vitro PKC-phosphorylated, FSBA-inactivated Raf-l(100 cpm) prior to two-dimensional resolution (MIX). Arrowheads denote the comigrating in vivo and in vitro phosphopeptides.

unique to the in vitro reaction and may be an artifact in that phosphorylation of these sites is not required for Raf-1 activation in vivo (Fig. 6).

Several studies have demonstrated that the mobility of Raf-1 in SDS-PAGE is retarded upon mitogen-induced phosphorylation in whole cells (15-17, 26-28, 30). Bryostatin 1 also mediates a decrease in the mobility of Raf-1 (Fig. 1). Indeed this “shift” in apparent molecular weight is dependent on Raf-1 phosphorylation (30). The physiologic significance of this mobility shift, if any, is unknown. Furthermore, it is unclear whether this mobility shift is dependent on the stoichiometry of Raf-1 phosphorylation per se, or the phosphorylation of a distinct site(s). However, stoichiometric phosphorylation by PKC in vitro results in no significant increase in the apparent molecular weight of Raf-l as assessed by SDS-PAGE (Fig. 5), suggesting that an additional phosphorylation event(s) may be required for this effect. Since IG3 and bryostatin 1 appear to induce the phosphorylation of Raf-1 at several sites not phosphorylated by PKC in vitro (Fig. 6), it is possible that phosphorylation of one or more of these “non-PKC” sites may be necessary to decrease the mobility of Raf-1 in SDS-PAGE.

Since it was impossible to isolate and purify the picomolar


1-2 I L S V S R 21- n

~3 KS P W ~ G S Q Q V L Q P T C S V L W M A ? I V I I - x

1.4 I ~ A S I P S L U I abel 2,

Id K L T D S S K I S N I I I 47-53 N.D.

N- c cR3 cR2 cR1

t t Y srra 89'" 89- rnl) N o w

FIG. 7. Synthetic Rat4 peptides containing candidate phoe phorylation sites for PKC. Upper panel, synthetic Raf-1 peptide se"

derlined residues conform to the consensus sequence for PKC quences are denoted using single-letter amino acid designations. Un-

substrates. Asterisks appear above deduced Raf-1 phosphorylation sites as described under "Results." The K,,, values for PKC were determined as described under 'Experimental Procedures" using double-reciprocal plots. Lower panel, the location of the deduced Raf-1 phosphorylation sites and corresponding synthetic peptides are denoted by arrows using a schematic of the Raf-1 protein structure. CR, conserved in raf family proteins; N, amino terminus; C, carboxyl terminus.

amounts of labeled Raf-1 from cells necessary for direct phoe- phopeptide sequencing, an alternate method for determining the PKC site was employed. Thus, a series of Raf-1 peptides conforming to the PKC substrate consensus sequences were synthesized and used to assess the PKC phosphorylation sites (Fig. 7) (41). The Michaelis constant (K,) of each unique p e p tide was calculated for the in vitro kinase reaction with PKC. Peptides R2, R3, and R4 are all avid PKC substrates (i.e. K, values of 77-250 w) (Fig. 7). Phosphoamino acid analyses of in vitro phosphorylated peptides R1, R2, R3, and R4 demonstrate that serine only was phosphorylated while R5 reveals that both serine and a minor amount of threonine phosphorylation oc- curred in vitro (data not shown). These data support a physiologic role for serine phosphorylation of Raf-1 by PKC at sites contained within R2, R3, and R4. Comparative phosphopeptide analysis was carried out and results indicate that the tryptic digests of phosphorylated R1, R3, and R4 peptides comigrate with tryptic digests of Raf-1 immunoprecipitated from cells stimulated with IG3, erythropoietin, or bryostatin 1 (Figs. 6 and 8). The maps of R1 and R4 each demonstrate two phosphopeptides secondary to incomplete tryptic digestion (data not shown). Furthermore, R3 and R4 digests also comigrate with two of the Raf-1 phosphopeptides observed after in vitro phosphorylation of baculoviral Raf-1 by PKC (Fig. 6). These data strongly suggest that peptides R1, R3, and R4 contain at least three distinct mitogen-induced Raf-1 serine phosphorylation sites. Additionally, results suggest that R3 and R4 contain physiologic serine phosphorylation sites which are stoichio- metrically modified by activated PKC.

The sites of PKC-mediated Raf-1 serine phosphorylation are contained within peptides R3 and R4 in the conserved kinase domain (CR3) of Raf-1 (Fig. 7). Although peptide R3 contains 4 serine residues, several lines of evidence suggest that Sei'97 is the preferred site of the PKC-mediated phosphotransfer reaction. First, only Sei'94 and Sei'97 conform to a consensus motif for PKC substrates, i.e. within 1-3 residues of a lysine or argi- nine residue (Fig. 7) (41). Second, residues homologous to Sei'94 and Sei'97 but not Sei'99 are conserved throughout the raf kinase family (A-raf, B-raf, and c-raf) (43) suggesting that these sites are likely to be functionally important. Indeed, resi-

IN VIVO

IL3 ER)

MIX

4 0 e

4dK - HCF

Fro. 8. Comparative analysis of Rat-1 phqhopeptides iao- lated from growth factor treated whole cells and synthetic Rat-1 peptides phosphorylated in v i t r o by PKC. Phosphopeptide maps were prepared from 32P-labeled Raf-1 isolated from erythropoietin- stimulated FDC-Pl/ER or IL3-treated FDC-P1 cells as described under 'Experimental Procedures" (IN W O ) . Synthetic Raf-1 peptides were phosphorylated by PKC in vitro, r e p d e d , and digested with L-1-tosylamide-2-phenylethyl chloromethyl ketone-treated trypsin. Washed peptide digests were resolved by electrophoresis and ascending chromatography. Numbered arrowheads 1-5 identify the products of tryptic digestion for the corresponding 32P-labeled Raf-1 peptides, Rl-R5. The tryptic maps of R1, R4, and R5 demonstrate two distinct phosphopeptides secondary to incomplete digestion (data not shown) (SYNTHETIC PEPTIDES). Mixtures of in vivo digests (100 cpm) and synthetic phosphopeptide digests (25 cpm each) were subjected to two-dimensional analysis (MIX) . Numbered arrowheads denote phosphopeptides which comigrate with the corresponding synthetic Raf-1 peptides after in vitro

sketch depicts the positions of Raf-1 phosphopeptides isolated from he- phosphorylation and tryptic digestion (1 = R1, 3 = R3, 4 = R4). The

matopoietic growth factor or bryostatin 1-treated cells. The origin is marked by an asterisk. Shaded figures illustrate phosphopeptides which comigrate with Raf-1 peptides phosphorylated in vitro by PKC. Open jigures illustrate phosphopeptides which do not comigrate with peptides phosphorylated by PKC in vitro. Numbered arrowheads denote phosphopeptides which comigrate with the corresponding synthetic Raf-1 peptides, as above. EPO, erythropoietin; HGF, hematopoietic growth factor.

dues homologous to Raf-1 Se#g7 are known physiologic sites of phosphorylation in other protein kinases, e.g. CAMP-dependent protein kinase (Thrlg7) and the protein tyrosine kinase p60"" ( Q f ? (44). Phosphorylation of such sites may, in part, regulate enzymatic activity. Thus, substitution of TYpl6 with phe- nylalanine in p60"" significantly decreases protein tyrosine kinase activity (45), indicating that the phosphate acceptor moiety is crucial for the regulation of enzymatic activity. There- fore by analogy, phosphorylation of the homologous residue by PKC, i.e. SelAg7, may regulate Raf-1 kinase activity. Third,

PKC Activates Raf-1

“1 T

1255

Receptor e! BRYoSTATlN . . . . . . . . . . . . . . . . . . . . . .

1 2 3 4 5 6

“1 = J 0 3 0 0 T

1 2 3 4 5 6

FIG. 9. PKC activates Rat-1. Baculoviral Raf-1 or control immune precipitates were treated for 30 min with 0.5 rn FSBA (lunes 3 and 4 ) or 1% Me2S0 control lanes ( I , 2,5, and 6), washed, and combined in an in vitro kinase priming reaction containing 1 p~ TPA, 40 pg/ml phosphatidylserine, and 800 p~ calcium chloride (lunes I 4 and 6). 2 p~ EGTA (lune 5), 100 p~ ATP (lunes 1-51, and 2 pl of purified PKC (lunes 2, 4-6) a t 37 “C for 30 min as described under “Experimental Proce- dures”. After priming Raf-1 immune precipitates were washed five times in ice-cold lysis buffer containing 1% Triton X-100 and 2 m~ EGTA, twice in Raf-1 kinase buffer and combined in a 50-111 Raf-1 kinase assay containing 100 p~ ATP, 20 pCi of [y32PlATF’ with (lower) or without (upper) 0.5 mg/ml histone 111s for 30 min a t 37 “C. Proteins were resolved by SDS-PAGE. Histone (lower) and Raf-1 (upper) phosphorylation were assessed by Cerenkov counting. Results are expressed as the mean percentage S.E. of untreated control (lune I ) values (mean: upper, 2.7 x lo2 cpm; lower, 5.6 x lo2 cpm). Lower values were also controlled for background histone phosphorylation (mean: lane 1, 3.2 x lo2 cpm; lane 2, 5.1 x lo2 cpm; lune 3, 2.4 x lo2 cpm; lune 4, 4.7 x lo2 cpm; lune 5, 4.3 x lo2 cpm; lune 6, 4.9 x lo2 cpm).

Edman degradation reveals that less than 5% of the phosphate is released after cycle 2, suggesting that the primary location(s) of PKC-mediated serine phosphorylation is C-terminal to Ser494, i.e. Ser497, Ser499, or Ser508 (data not shown). Unfortu- nately, manual Edman degradation could only be successfully camed out for four cycles, making a direct determination of the extent of S e P 7 phosphorylation impossible by this method (data not shown). Since the submission of this article, Kolch e t al. (46) have reported that PKC can phosphorylate Raf-1 at a site contained within peptide R3. These authors demonstrated that alanine substitution of Sefls4, Ser497, or Ser499 markedly reduced PKC-mediated Raf-1 peptide phosphorylation in v i t ro , suggesting that all of these residues are critical for efficient phosphate transfer. Using a synthetic peptide this phosphorylation site was further localized to either or Ser499. Additionally, the authors demonstrated by mutational analysis

MAPK-K + + MAPK

MYC Jun

PROLIFERATION

FIG. 10. A cascade of sequential phosphorylation events is involved in hematopoietic growth signal transduction. Hematopoi- etic growth factor receptor engagement is envisioned to stimulate phospholipid hydrolysis in the plasma membrane, resulting in the generation of diacylglycerol, a physiologic activator of PKC. PKC phosphorylates Raf-1 on SeF” and SeP9, activating this enzyme which is required for proliferation. Activated Raf-1 propagates a sequence of phosphorylation reactions in the cytosol involving MAP kinase-kinase (MAPK-K) and MAP kinase Lth4.l“). Activated MAF’K may phosphorylate and potentially regulate the function of transcriptional activators in the nucleus such as the products of c-myc (Myc) and cjun (Jun).

that Se#= is necessary for PKC-mediated enzymatic activation of Raf-1 and for the transformation of NIH3T3 cells con- transfected with PKCa. However, the consequences of Sei’97 mutation were not reported. Therefore, although these findings strongly support Ser499 as a candidate PKC phosphorylation site, a precise determination will await successful phosphopeptide sequencing. R4 contains the additional site of PKC-mediated Raf-1 phosphorylation. The predicted site, based on the substrate consensus sequences for PKC, is Ser619. Manual Ed- man degradation reveals that approximately 90% of the labeled phosphate is released after cycle 2, confirming that Ser6I9 is the preferred site of PKC-mediated RA phosphorylation (data not shown). As is the case for Ser497, this PKC phosphorylation site is conserved throughout the raf kinase family (43), suggesting a role for phosphorylation in the Raf regulatory mechanism(s).

Bryostatin 1 and hematopoietic growth factors appear to induce the phosphorylation of several Raf-1 sites not phosphorylated by PKC, a t least in v i t ro , suggesting the involvement of serine kinases distinct from PKC in the Raf-1 phosphorylation mechanisms (Figs. 6 and 8). The identity of these putative additional kinases is unknown. Digests of phosphorylated R1 peptide comigrate with two of these “non-PKC” phosphopeptide spots due to incomplete digestion (Fig. 8, data not shown). The Raf-1 phosphorylation site in R1 is Ser43 (Fig. 7). Ser43, near the N terminus, is by analogy, situated in a portion of Raf-1 similar to the negative regulatory “pseudosubstrate” domain of PKC (Fig. 7) (42). As such, phosphorylation of this site could be involved in relieving steric constraints at the substrate binding site(s) of Raf-1. The identities and functions of the additional non-PKC sites detected by phosphopeptide mapping are unknown. PKC Activates Rafl-Based on the above results, we per-

formed experiments to determine if PKC could “prime” (i.e. activate) the Raf-1 enzyme. Results indicate that such priming of Raf-1 results in a marked stimulation of the Raf-1 protein kinase activity in v i t ro , while mock priming (i.e. in the absence of PKC) failed to activate Raf-1 (Fig. 9). FSBA, an ATP analog, reacts with the lysine residue necessary for phosphate transfer


to inactivate protein kinases (35,36). Raf-1 treated with FSBA displays minimal kinase activity, indicating that the phos- photransferase activity observed after priming with PKC is intrinsic to Raf-1 and does not result from any residual PKC (Fig. 9). Furthermore, priming reactions performed with PKC in the absence of ATP also failed to activate Raf-1, indicating that a phosphate transfer reaction during the priming step is required for Raf-1 enzymatic activation. These results demonstrate that PKC-mediated serine phosphorylation is sufficient for Raf-1 kinase activation in vitro and position PKC as a physiologic Raf-1 kinase-kinase. These data also confirm recently reported findings that Raf-1 can be activated by the classical isoforms of PKC in vitro (47).

Finally, a growing body of data supports a sequential phosphorylation cascade, such as PKC + Raf-1 + MAP kinase- kinase + MAP kinase, as a mechanism for mitogenic signal transduction. A variety of hematopoietic growth factors including IL-3, GM-CSF, and erythropoietin rapidly activate PKC and Raf-1 following receptor engagement (6, 13-15, 17). Our results shown here suggest that such an activation cascade may be necessary for post-receptor growth signaling. Thus, by identifying a direct role for PKC in Raf-1 activation, the current study links molecular events generated at the plasma membrane to a cascade of cytosolic activation events which may culminate in nuclear transcription (Fig. 10).

Acknowledgments-We are grateful to Dr. Ulf R. Rapp for kindly providing the baculovirus containing the human c-raf expression vector, Dr. Takahiko It0 for graphic design, and Constance V. Stewart for manuscript preparation.

REFERENCES

2. Cross, M., and Dexter, T. M. (1991) Cell 64,271-280 1. Metcalf, D. (1991) Science 264,529-533

3. Bazan, J. F. (1990) Proc. Natl. Acad. Sei. U. S. A. 87,6934-6938 4. Itoh, N., Yonehara S. , Schreurs, J., Gorman, D. M., Maruyama, K., Ishii, A.,

5. Isfort, R. A,, Stevens D. A,, May, W. S. , and Ihle, J. N. (1988) Prcc. Natl. Acad. Yahara, I., Arai, K., and Miyajima, A. (1990) Science 247,324-327

6. Fields, A. P., Pincus, S . M., Kraft, A. S . , and May, W. S . (1989) J. Biol. Chem. Sci. U. S. A 86,7982-7986

264,21896-21901 7. Gomez-Cambronero, J., Yamazaki, M., Metwally, F., Molski, T. F. P., Bonak, V.

A,, Huang, C., Becker, E. L., and Sha'afi, R. I. (1989) Proc. Natl. Acad. Sci.

8. Mills, G. B., May, C., McGill, M., Fung, M., Baker, M., Sutherland, R., and U. S. A 86,35694573

Greene, W. C. (1990) J. Biol. Chem. 266,35614567 9. Farrar, W. L., and Ferris, D. K. (1989) J. BioZ. Chem. 264.12562-12567

10. Quelle, F. W., and Wojchowski, D. M. (1991) J. Bid . Chem. 266,609-614 11. Hatakeyama, M.. Kono. T., Kobayashi, N., Kawahara, A., Levin, S . D.,

Perlmutter, R. M., and Taniguchi, T. (1991) Science 252, 1523-1528 12. Sherr, C. J. (1990) Blood 75, 1-12 13. Adunyah, S . E., Unlap, T. M., Wagner, F., and &&,A. S . (1991)J. B i d . Chem.

268,5670-5675

14. Spangler, R., Bailey, S . C., and Sytkowski, A. J. (1991) J. Biol. Chem. 286,

15. Carroll, M. P., Clark-Lewis, I., Rapp, U. R., and May, W. S . (1990) J. Biol.

16. Turner, B., Rapp, U., App, H., Greene, M., Dobashi, K, and Reed, J. (1991)

17. Carroll, M. P., Spivak. J. L., McMahon, M., Weich, N., Rapp, U. R., and May,

18. Okuda, K, Sanghera, J. S . , Pelech, S . L., Kanakura, Y., Hallek, M., Griffin, J.

19. Howe, L. R., Levers, S . J., Gomez, N., Nakielny, S., Cohen, P., and Marshall,

20. Kyriakis, J. M., App, H., Zhang, X.-F., Baynejee, P., Brautigan. D. L., Rapp, U.

21. Northwood, I. C., Gonzalez, F. A., Wartmann, M., Raden, D. L., and Davis, R.

22. Nemenoff, R. A, Winitz, S. , Qian, N.-X., Van Putten, V., Johnson, G. L., and

23. Chung, J., Pelech, S. L., and Blenis, J. (1991)pToc. Natl. Acad. Sci. U. S . A. 88,

24. Baker, S. J., Kerppola, T. K, Luk, D., Vandenberg, M. T., Marshak, D. R.,

25. Seth, A,, Alvarez, E., Gupta, S . , and Davis, R. J. (1991) J. Biol. Chem. 288,

26. Kovacina, K. S . , Yonezawa. K., Brautigan, D. L., 'lbnks, N. K., Rapp. U. R., and

27. Baccarini, M., Sabatini, D. M., App, H., Rapp, U. R., and Stanley, E. R. (1990)

28. App, H., Hazan, R., Zilberstein, A., Ullrich, A,, Schlessinger, J., and Rapp, U.

30. Momson. D. K., Kaplan, D. R., Rapp, U. R., and Roberta, T. M. (1988) Proc. 29. May, W. S . , and Carroll, M. P. (1993) Colloq. INSERM 229, in press

31. Summers, M. D., and Smith, G. E. (1988)A Manual ofMethods for Baculouirus Natl Acad. Sci. U. S. A. 85,8855-8859

Vectors and Insect Cell Culture Procedures, pp. 10-36, Texas A & M Uni-

32. Boyle, W. J., van der Geer, P., and Hunter, T. (1991) Methods Ewymol. 201, versity, College Station, TX

110-149 33. Hidaka, H., Inagaki, M., Kawamoto, S., and Sasaki, Y. (19841 Biochemistry 23,

5036- 5041 34. Tamaoki, T., Nomoto, H.. Takahashi, T., Kato, Y., Morimoto, M., and 'lbmita, F.

(1986) Biochem. Biophys. Res. Commun. 135,397402 35. Zoller, M. J., and Taylor, S . S . (1979) J. Biol. Chem. 254,836343368 36. Adams, P. D., and Parker, P. J. (1992) J. Biol. Chem. 267,13135-13137 37. Dumnio, V., Nip, L., and Pelech, S . L. (1989) Biochem. Biophys. Res. Commun.

38. Hogans, B. B., and Spivak, J. L. (1988) J. Cell. Phys. 137, 346352 39. Knopf, J. L., Lee, M. -H., Sultzman, L. A,, Kriz, R. W., Loomis, C. R., Hewick,

40. Nishizuka, Y. (1992) Science 2M). 607414 41. Kennelly, P. J., and Krebs, E. G. (1991) J. Biol. Chem. 266, 15555-15558 42. Orr, J. W., Keranen, L. M., and Newton, A. C. (1992) J. Biol. Chem. 287,

43. Sithanandam, G.. Kolch, W., Duh, F.-M., and Rapp, U. R. (1990) Oncogene 5,

44. Hanks, S. K, qui^. A. M., and Hunter, T. (1988) Science 241,42-52 45. Piwnica-Worms, H., Saunders, K B., Roberts, T. M., Smith, A. E., and Cheng,

S . H. (1987) Cell 49, 75-82 46. Kolch, W., Heidecker, G., Kochs, G., Hummel, R., Vahldl, H., Mischak, H.,

Finkenzeller, G., Marme, D., and Rapp, U. R. (1993) Nature 3&1,24%252 47. Szeri, O., Vollmer, K, Liyanage, M., Frith, D., Kour, G., Mark, G. E., and

Stabel, S . (1992) Oncogene 7,225%2262

681484

Chem. ulb 19812-19817

P m . Natl. Acad. Sci U. S. A. 88, 1227-1231

W. S . (1991) J. Bwl. Chem. 268, 14964-14969

D., and Drucker, B. J. (1992) Blood 79,2880-2887

C. J. (1992) Cell 71,335342

R., and Avruch, J. (1992) Nature 358,417421

J. (1991) J. Biol. Chem. 266, 1526615276

Heasley, L. E. (1993) J. Bwl. Chem. 268, 196&1964

4981- 4985

Curran, T., and Abate, C. (1992) Mol. Cell. Biol. 12,46944705

23521-23524

Roth, R. A. (1990) J. Biol. Chem. 266, 12115-12118

EMBO J. 9,36493657

(1991) Mol. Cell. Biol. 11, 913-919

164,804-808

R. M., and Bell, R. M. (1986) Cell 46,491-502

15263-15266

1775-1780

protein kinase c-mediated serine phosphorylation directly

Documents