Phosphorylation and Regulation of Raf by Akt (Protein Kinase B)

See allHide authors and affiliations

Science  26 Nov 1999:
Vol. 286, Issue 5445, pp. 1741-1744
DOI: 10.1126/science.286.5445.1741


Activation of the protein kinase Raf can lead to opposing cellular responses such as proliferation, growth arrest, apoptosis, or differentiation. Akt (protein kinase B), a member of a different signaling pathway that also regulates these responses, interacted with Raf and phosphorylated this protein at a highly conserved serine residue in its regulatory domain in vivo. This phosphorylation of Raf by Akt inhibited activation of the Raf-MEK-ERK signaling pathway and shifted the cellular response in a human breast cancer cell line from cell cycle arrest to proliferation. These observations provide a molecular basis for cross talk between two signaling pathways at the level of Raf and Akt.

The signaling pathway comprising Raf, MEK (mitogen-activated protein kinase, or ERK kinase), and ERK (extracellular signal–regulated kinase) lies downstream of the small guanine nucleotide binding protein Ras and mediates several apparently conflicting cellular responses, such as proliferation, apoptosis, growth arrest, differentiation, and senescence, depending on the duration and strength of the external stimulus and on cell type. Another pathway that lies downstream of Ras includes phosphatidylinositol (PI) 3-kinase and Akt (or protein kinase B) and also regulates these cellular responses, acting either synergistically with (1) or in opposition to (2) the Raf pathway. Coordination of the two pathways in a single cellular response may depend on cell type or the stage of differentiation (3,4).

The kinase activity of Raf (5) is regulated by phosphorylation of a highly conserved serine residue (Ser259) in the amino-terminal regulatory domain (RafNT) (6). Phosphorylation of Ser259 mediates binding of the 14-3-3 protein, resulting in Raf inactivation. Mutation of Ser259 to Ala constitutively activates the kinase activity of Raf (7). The kinases responsible for phosphorylating Ser259 are unknown but may include members of the protein kinase C family (8).

Because the region surrounding Ser259 in Raf conforms to a consensus sequence for phosphorylation by the serine-threonine kinase Akt (Fig. 1A) (9, 10), we analyzed the biochemical effect and biological function of Akt-mediated phosphorylation of Raf. Stimulation of human embryonic kidney (HEK293) cells expressing Flag epitope–tagged Raf (7) with epidermal growth factor (EGF) resulted in an increase in the kinase activity of Raf (Fig. 1B). Coexpression of a hemagglutinin epitope (HA)–tagged kinase-inactive mutant of Akt, HA-AktA179 (11), enhanced the EGF-induced increase in Raf activity. However, coexpression of an activated Akt mutant that is constitutively targeted to the plasma membrane (m/p-HA-Akt) (11) markedly inhibited the EGF-induced increase in Raf activity (Fig. 1B). A constitutively activated Raf mutant containing the Ser259→Ala mutation (7) was not regulated by these Akt proteins (12), suggesting that Akt induces phosphorylation of Raf on Ser259. In place of an exogenous stimulus, we expressed another constitutively active mutant of Raf, in which Tyr340 is mutated to Asp (13), in HEK293 cells. The kinase activity of this mutant Raf was also down-regulated by Akt-dependent phosphorylation of Ser259 (Fig. 1C).

Figure 1

Inhibition of Raf activation by Akt. (A) Comparison of the putative Akt phosphorylation site in the amino-terminal region of Raf with the sequences of phosphorylation sites of known Akt substrates (18, 19). The phosphorylated residues are shown in bold and a consensus sequence is denoted below. (B andC) Flag-Raf, the activated mutant Flag-RafD340, or the double mutant Flag-RafA259/D340 was coexpressed with a constitutively active Akt (m/p-HA-Akt) or a kinase-inactive Akt (HA-AktA179) in HEK293 cells. In (B), the cells were stimulated with EGF (10 ng/ml) for 5 min, and kinase activity of Raf toward a glutathione S-transferase fusion protein of kinase-inactive MEK (GST-MEK) was assayed. The radioactivity incorporated into GST-MEK was quantitated by image analysis (Image-Quant) and is shown as fold activation relative to the basal activity of Flag-Raf. Immunocomplexes were also subjected to immunoblot analysis with antibodies to the Flag epitope.

Endogenous Raf in HEK293 cells was also phosphorylated on Ser259 after stimulation of cells with insulin-like growth factor (IGF). The PI 3-kinase inhibitor LY294002 inhibited IGF-induced phosphorylation of Raf on Ser259 and increased the extent of activation of Raf and ERK (Fig. 2). The effects of LY294002 were mimicked by expression of the kinase-inactive Akt mutant (HA-AktA179) in HEK293 cells.

Figure 2

Effects of inhibition of Akt on phosphorylation of Ser259 of Raf and on Raf and ERK activities. HEK293 cells were deprived of serum, incubated for 20 min with the indicated concentrations of LY294002, and then stimulated with IGF (50 ng/ml) for 4 min. Cells expressing kinase-inactive Akt (HA-AktA179) were similarly treated with IGF. Endogenous Raf protein was immunoprecipitated and in vitro kinase assays were performed as in Fig. 1B. Cell lysates were also subjected to immunoblot analysis with antibodies specific for Raf phosphorylated on Ser259, for activated Akt, or for activated ERK, or with antibodies to the corresponding unmodified proteins (20).

In vivo 32P-labeling experiments revealed that activation of Akt by an oncogenic form of Ras (RasV12) or by IGF resulted in RafNT phosphorylation (4.7- and 2.2-fold increases, respectively) (Fig. 3). However, coexpression of the kinase-inactive Akt mutant inhibited the increase in RafNT phosphorylation in response to either RasV12 or IGF. Phosphorylation induced a decrease in the electrophoretic mobility of RafNT. Almost all of the 32P-labeled phosphate was incorporated into Raf at Ser259, as revealed by the observation that the RafNTA259 mutant was phosphorylated to a much lesser extent than was RafNT (Fig. 3).

Figure 3

Akt-induced phosphorylation of Raf on Ser259 in vivo. HEK293 cells were transfected with the indicated constructs, deprived of serum, and metabolically labeled with [32P]- orthophosphate (0.25 mCi/ml) as described (17). They were then stimulated with IGF (100 ng/ml) for 15 min, after which Flag-RafNT proteins were immunoprecipitated, resolved by SDS–polyacrylamide gel electrophoresis, and subjected to image analysis for detection of radioactivity and to immunoblot analysis with antibodies to Flag. The incorporation of 32P was normalized for protein content and is shown as fold increase compared with the basal phosphorylation of Flag-RafNT. Aktwt, wild-type Akt.

When expressed as a GST fusion protein, RafNT was directly phosphorylated by Akt on Ser259 in vitro (Fig. 4A); the GST-RafNTA259 mutant was not phosphorylated by Akt under these conditions (Fig. 4B). The Akt substrate BAD (14) was included in these experiments as a positive control (12). The carboxyl-terminal domain of Raf (RafCT) fused to GST was also not phosphorylated by Akt (12).

Figure 4

Phosphorylation of Raf on Ser259 by Akt in vitro. HEK293 cells expressing the indicated HA-Akt proteins were deprived of serum, stimulated for 20 min with IGF (100 ng/ml), lysed, and subjected to immunoprecipitation with antibodies to HA. The resulting precipitates were then subjected to an in vitro kinase assay with GST-RafNT (A) or GST-RafNTA259 (B) as substrate. Autoradiograms are shown in the upper panels, and immunoblot analysis with antibodies to GST is shown in the lower panels (21). The amounts of Akt in the different reaction mixtures were similar (12).

Akt and Raf coimmunoprecipitated from HEK293 cells overexpressing these proteins (Fig. 5A). Akt associated with both RafNT and RafCT. Two Raf mutants that do not bind 14-3-3 protein, RafNTA259 and RafCTA621 (7), also co-immunoprecipitated with Akt, indicating that 14-3-3 protein does not serve as an adapter between Raf and Akt. In addition, Raf proteins in which residues that constitute putative lipid binding sites are mutated (15), RafNTG143/E144 and RafCTG398/E399, also associated with Akt (Fig. 5A).

Figure 5

Association of Akt and Raf. (A) Lysates prepared from HEK293 cells transfected with the indicated constructs were subjected to immunoprecipitation with antibodies to Flag with or without an anti-Flag antibody competing peptide (FLAG peptide), and the resulting precipitates were subjected to immunoblot analysis with antibodies to Flag or to HA. The amount of HA-Akt in direct lysates (DL) was also assessed by immunoblot analysis with antibodies to HA (lower panel) (22). (B) Serum-deprived MCF-7 cells were stimulated with IGF (100 ng/ml) for 5 min, after which endogeneous Raf and Akt were immunoprecipitated (IP) with their cognate antibodies in the absence or presence of the immunizing peptide for the antibodies to Akt (22). The resulting immunoprecipitates were then subjected to immunoblot analysis with the same antibodies. IgG, immunoglobulin G.

Exposure of the MCF-7 breast cancer cell line to IGF induces cell proliferation as a result of activation of the PI 3-kinase–Akt pathway and transient activation of the Raf-MEK-ERK pathway (3). However, prolonged activation of the Raf cascade inhibits growth in these cells (3). Thus, regulation of the Raf-MEK-ERK pathway in MCF-7 cells determines whether the response is proliferation or growth arrest. Endogenous Akt and endogenous Raf co-immunoprecipitated from MCF-7 cells that had been stimulated with IGF for 5 min (Fig. 5B). This interaction was disrupted with a peptide that competed with Akt for the immunoprecipitating anti-Akt antibodies (Fig. 5B). Serum deprivation of the cells or stimulation with IGF for 5 min or 2 hours did not alter the amounts of the co-immunoprecipitating proteins (12).

On exposure of MCF-7 cells to IGF, both Raf and ERK activities peaked after 2 to 3 min and then decreased (Fig. 6A). In contrast, Akt was activated rapidly and remained fully active for up to 20 min. Inhibition of Akt activation by LY294002 increased Raf and ERK activities (Fig. 6A).

Figure 6

Activation of Raf and ERK promoted by inhibition of PI 3-kinase and Akt. (A) Serum-deprived MCF-7 cells were treated with LY294002 and then exposed to IGF (100 ng/ml) or vehicle (dimethyl sulfoxide) for the indicated times. Lysates were subjected to in vitro kinase assays of Raf activity as in Fig. 1 or to immunoblot analysis with antibodies specific for activated ERK or Akt; the blot was stripped and reprobed with antibodies to the unmodified proteins (20, 23). (B) Serum-deprived MCF- 7 cells were treated as indicated, lysed, and subjected to immunoblot analysis with antibodies to p21Cip1 (24). PD, PD98059 (20 μm); LY, LY294002 (20 μm). (C) MCF-7 cells transiently expressing the indicated Akt proteins were treated with PMA (100 ng/ml) for 4 hours and subjected to indirect double-immunofluorescence analysis to reveal Akt-expressing cells (red) and p21Cip1 expression (green) (24). Overlays of representative confocal microscopic images are shown. Bars, 10 μm. (D) MCF-7 cells (1.5 × 105) were treated with IGF (100 ng/ml), PMA (100 ng/ml), or PMA (100 ng/ml) plus PD98059 (PD, 20 μM) for 3 days, and then counted. Data are means from two independent experiments performed in duplicate. Error bars indicate SEM. (E) A model for the Akt-Raf interaction in the context of the respective signaling pathways. Phosphorylation of Raf by Akt leads to inhibition of the Raf-MEK-ERK cascade and modulation of the cellular response.

Tumor promoters such as phorbol 12-myristate 13-acetate (PMA) induce sustained activation of ERK as well as the expression of the cyclin-dependent kinase inhibitor p21Cip1 in MCF-7 cells (3). The MEK inhibitor PD98059 partially inhibited the PMA-induced increase in the amount of p21Cip1, whereas LY294002 had no such effect, implicating the Raf-MEK-ERK pathway in regulation of p21Cip1 expression (Fig. 6B). Expression of membrane-targeted, constitutively active Akt (m/p-HA-Akt) in MCF-7 cells prevented PMA-induced expression of p21Cip1 in the nucleus; expression of membrane-targeted, kinase-inactive Akt (m/p-HA-AktA179) had no such inhibitory effect (Fig. 6C). The growth-inhibitory effect of PMA on MCF-7 cells was also partially reversed by the MEK inhibitor PD98059 (Fig. 6D) (3).

Our results demonstrate that Akt antagonizes Raf activity by direct phosphorylation of Ser259. This modification creates a binding site for 14-3-3 protein, a negative regulator of Raf. Similarly, phosphorylation of BAD or the forkhead transcription factor FKHRL1 by Akt also promotes binding of 14-3-3 protein (14, 16). In all three instances, phosphorylation by Akt inactivates the function of its substrate. Cross talk between the Raf-MEK-ERK and the PI 3-kinase–Akt pathways, mediated by direct interaction of Akt with and its phosphorylation of Raf, may switch the biological response from growth arrest to proliferation, as shown for MCF-7 cells, and may also modulate senescence or differentiation as shown for myoblast differentiation (25), depending on the cellular system (Fig. 6E).

  • * To whom correspondence should be addressed. E-mail: moelling{at}


View Abstract

Navigate This Article