From the Cover: The Fbw7 tumor suppressor regulates glycogen synthase kinase 3 phosphorylation-dependent c-Myc protein degradation (original) (raw)

Proc Natl Acad Sci U S A. 2004 Jun 15; 101(24): 9085–9090.

From the Cover

Markus Welcker

Divisions of *Clinical Research, †Human Biology, and §Basic Sciences, Fred Hutchinson Cancer Research Center, Seattle, WA 98109; ∥Department of Medicine, University of Washington School of Medicine, Seattle, WA 98104; and ¶Department of Pathology, Harvard Medical School, Boston, MA 02115

Amir Orian

Jianping Jin

Jonathan A. Grim

J. Wade Harper

Robert N. Eisenman

Bruce E. Clurman

** To whom correspondence may be addressed at: Fred Hutchinson Cancer Research Center, P.O. Box 19024, Seattle, WA 98109-1024. E-mail: gro.crchf@namnesie or gro.crchf@namrulcb.

‡M.W. and A.O. contributed equally to this work.

Contributed by Robert N. Eisenman, April 19, 2004

Supplementary Materials

Supporting Figures

GUID: F990A1E9-68BA-4A81-A311-8791612D6C3D

GUID: 9A09CF9A-E0BA-4CA0-92E8-55E4287AA6F0

GUID: 0FAEDF27-418F-4EBC-B2B7-F85617CE5A75

GUID: 858D6BE2-9069-4732-9439-3B52AB196CAD

Abstract

Myc proteins regulate cell growth and division and are implicated in a wide range of human cancers. We show here that Fbw7, a component of the SCFFbw7 ubiquitin ligase and a tumor suppressor, promotes proteasome-dependent c-Myc turnover in vivo and c-Myc ubiquitination in vitro. Phosphorylation of c-Myc on threonine-58 (T58) by glycogen synthase kinase 3 regulates the binding of Fbw7 to c-Myc as well as Fbw7-mediated c-Myc degradation and ubiquitination. T58 is the most frequent site of c-myc mutations in lymphoma cells, and our findings suggest that c-Myc activation is one of the key oncogenic consequences of Fbw7 loss in cancer. Because Fbw7 mediates the degradation of cyclin E, Notch, and c-Jun, as well as c-Myc, the loss of Fbw7 is likely to elicit profound effects on cell proliferation during tumorigenesis.

Changes in the levels of Myc family transcription factors profoundly influence cell growth, proliferation, differentiation, and apoptosis. Concordantly, cells have evolved multiple mechanisms to tightly control the levels of c-myc gene products, including regulation of the myc promoter by transcription factor binding, attenuation of RNA polymerase elongation, c-myc mRNA transport, stability and translation, and stability of the c-Myc protein. c-Myc is a rapidly degraded protein (1), and a large body of work has demonstrated the involvement of the ubiquitin pathway in c-Myc degradation (2-7).

Analysis of c-Myc mutations has implicated the c-Myc transactivation domain (TAD) in ubiquitination and degradation (4, 8). The TAD, spanning amino acids 40-150, contains the sequence PTPPLSP (residues 57-63 in human c-Myc), within which both T58 and S62 are phosphorylated (9-11). Phosphorylation of S62 mediated by the Ras/MEK/ERK kinase pathway is linked to stabilization and accumulation of c-Myc (6, 7). Phosphorylation of S62 appears to signal subsequent phosphorylation of T58 through the phosphatidylinositol 3-kinase/AKT/glycogen synthase kinase 3 (GSK-3) cascade, triggering destabilization of c-Myc. Recent work has revealed that S62 dephosphorylation is also required for degradation, and this is regulated by the Pin1 prolyl isomerase and protein phosphatase 2A (12). Furthermore, the T58 region is subject to the largest number of c-Myc mutations in human lymphoma cells (13-15). While the precise biological effects of these mutations are controversial, lymphoma-associated T58 mutations increase c-Myc protein half-life, as do similar mutations in retroviral v-myc alleles (4, 16-18).

Two recent papers have implicated the F-box protein Skp2 in c-Myc turnover in vivo (19, 20). F-box proteins are the substrate-recognition components of SCF (Skp-Cullin-F-box) ubiquitin ligases that bring together protein substrates with the catalytic core of the ubiquitin machinery. Importantly, regulation of c-Myc by Skp2 does not involve either c-Myc phosphorylation or the T58 region, and ligases other than Skp2 must mediate T58-dependent c-Myc proteolysis.

Prompted by evidence indicating genetic interaction between Drosophila myc (dmyc) and the Archipelago (Ago) ubiquitin ligase (21) we focused on the possible involvement of Fbw7/hCdc4, the mammalian ortholog of Ago, in c-Myc turnover. Conserved WD40 repeats in Fbw7 mediate recognition of phosphorylated substrates (22, 23), and Fbw7 targets phosphorylated cyclin E, Notch, and c-Jun for proteolysis. Moreover, Fbw7 is mutated in breast, endometrial, ovarian, and colon cancer cells or cell lines and functions as a tumor suppressor whose loss causes genetic instability (24-28). Deletion of Fbw7 in the mouse causes embryonic lethality with defects in cardiovascular development and increased cyclin E and Notch protein in the embryo and placenta (29, 30). Here we show that Fbw7 interacts with c-Myc and regulates its proteasome-dependent degradation in mammalian cells and its ubiquitination in vitro.

Methods

Antibodies and Plasmids. The following antibodies were used: FLAG and tubulin (Sigma); Myc-N262, 9E10, Myc-C33, CDK2 (M2), Hsp90, and β-catenin (Santa Cruz Biotechnology); pT58 c-Myc (no. 9401) (Cell Signaling); hemagglutinin (12CA5); and GSK-3 (368662) (Calbiochem). Fbw7α, Fbw7γ, Fbw1, and Fbw6 were cloned from Wi-38 cDNA into 3pX-FLAG-myc-CMV-24 expression vector (Sigma) and sequenced. Fbw7β cloning was previously described (24). Mutations were generated by the QuikChange method (Stratagene) and sequenced. DnFbw7WD was generated by PCR starting from the 194th residue of the common region. The PGID5-6CS2MT and PGID5-6LP plasmids were provided by F. Zhang and P. Klein (University of Pennsylvania, Philadelphia). For retroviral expression, FLAG-tagged Fbw7 was cloned into pBabePuro (31). RNA interference was achieved as described (32) with these target sequences: Fbw7, ACCTTCTCTGGAGAGAGAAATGC (24) [Fbw7-1 short interfering RNA (siRNA)] or GTGTGGAATGCAGAGACTGGAGA (Fbw7-2 siRNA), and control siRNA, TATGTCAAGTTGTATAGTTA (33).

Cell Culture, Transient Transfection, Retroviral Transduction, Transcription Assays, and Immunoblotting. 293, U2OS, HeLa, and primary human foreskin fibroblasts (HFF) were grown in DMEM with 10% FCS. Cells were treated with MG-132 (5 μg/ml, Calbiochem) or LiCl (20 mM) as indicated. For Fbw7-mediated turnover, 1 μg of pCS2c-Myc plasmid was cotransfected with 0.3 μg of pFLAG-Fbw7α or 3 μg of pFLAG-Fbw7β or pFLAG-Fbw7γ plasmid by the calcium phosphate method (34). Cells were lysed, quantitated, electrophoresed, and Western blotted as described, and IP Western assays were as described (34). Viral stocks were prepared in Phoenix Ampho or Phoenix Eco cells (provided by G. Nolan, Stanford University, Stanford, CA). Cells were transduced by supernatants in DMEM and 5 mg/ml Polybrene, and selected with puromycin (1-3 μg/ml). Transcription assays were preformed as described (35), using pGLm2 (Ebox-Luciferase) reporter vector. The indicated amounts of pCS2+Fbw7 alone, or with c-Myc (75 ng) and Max (100 ng) were transfected into 293T cells.

RNA Analyses. For RT-PCR, total RNA was isolated from transduced cells and 5 μg of total RNA was used in a 20-μl reverse transcription reaction, and 2.5 μl of the resulting cDNA was used for PCR. Fbw7 primers and PCR conditions were as described (36). Control human hypoxanthine phosphoribosyltransferase (HPRT) primers were GAACGTCTTGCTCGAGGTGT and CTGCATTGTTTTGCCAGTGT. TaqMan analysis was performed by using an Fbw7 primer-probe set (available on request) with an ABI Prism 7700 sequence detection system (Applied Biosystems) using TaqMan 2× master mix (Perkin-Elmer) and human glyceraldehyde-3-phosphate dehydrogenase (GAPDH) primers (Invitrogen) in triplicate.

Pulse-Chase Experiments. As shown in Fig. 1, 293 cells were transfected with 1 μg of CMV-c-Myc and 3 μg of either pFLAGFbw7 or vector. Plates were preincubated (15 min in DMEM without methionine and cysteine plus 5% dialyzed FCS), pulsed with Tran35S-label (ICN; 300 μCi/ml; 1 μCi = 37 kBq) for 10 min, and chased (DMEM/10% FBS plus 400 mg/liter methionine) as indicated. As shown in Fig. 3_B_, HeLa cells were labeled with 500 μCi/ml Tran35S-label for 10 min. Lysates were immunoprecipitated with anti-c-Myc-N262 antibody. For cycloheximide chase experiments, 293 cells were transfected by FuGENE 6 (Roche Molecular Biochemicals) with 5 μg of dnFbw7WD and cycloheximide (10 μg/ml) treatment as indicated.

Fbw7 negatively regulates c-Myc abundance and function. (A) 293 cells were cotransfected with c-Myc (lanes 2-8) and FLAG-Fbw7 vectors as shown. Lysates were immunoblotted for c-Myc (C-33) or Fbw7 (FLAG); asterisk indicates a background band. (B) 293 cells were cotransfected with c-Myc, Max, and FLAG-Fbw7 as indicated. Lysates were blotted for c-Myc-N262 or Fbw7 (FLAG). (C) 293 cells were transfected as shown with c-Myc and Fbw7 and blotted for Fbw7 (FLAG) and c-Myc-N262 expression. MG indicates cells treated with MG-132 for2h(+) or untreated controls (-). (D) Pulse-chase analysis of 293 cells transfected with c-Myc and either Fbw7γ or empty vector (see Methods). Lane 1, untransfected cells. (E) Cells were cotransfected with c-Myc, a c-Myc reporter plasmid, and Fbw7γ as indicated. Transcriptional activation is shown (see text).

The regulation of c-Myc by Fbw7 requires both GSK-3 activity and c-Myc T58. (A) Sequence alignment of c-Myc T58 with cyclin E T380 and the Cdc4 phosphodegron (CPD) consensus. indicates where basic residues are unfavorable. (B) 293 cells were transfected as indicated, and blotted for c-Myc-N262 and Fbw7 (FLAG). (C) 293 cells were cotransfected with the indicated c-Myc constructs and either dnFbw7WD or vector. Fbw7 and c-Myc abundance in anti-FLAG immunoprecipitates and total lysates is shown. (D) In vitro translated Fbw7 or Fbw6 were bound to an immobilized c-Myc peptide (residues 51-69) when unphosphorylated, or when T58, S62, or both are phosphorylated. (E) 293 cells were transfected with c-Myc (lanes 2-6), and Fbw7γ. Cells in lanes 4 and 5 were cotransfected with the axin GSK-interacting domain (GID), or an inactive GID mutant (GID*). Cells in lane 6 were treated with LiCl.

Peptide Binding. Five microliters of in vitro translated Fbw7 or Fbw6 was incubated with immobilized c-Myc peptide encompassing residues 51-69 (KKFELLPTPPLSPSRRSGL) or the same peptide phosphorylated on T58, S62, or both, in 0.15 ml of 50 mM Tris·HCl/2 mM EDTA/100 mM NaCl/0.1% Nonidet P-40 for 1 h before washing three times with binding buffer. Bound proteins were subjected to SDS/PAGE and autoradiography using 20% of input as controls.

In Vitro Ubiquitination. Myc ubiquitination was performed by using in vitro [35S]methionine-labeled Myc translated by TNT T7/SP6 coupled reticulocyte system (2.5 μl, Promega) or TNT T7/SP6 coupled wheat germ extract system (1.25 μl, Promega) in the presence of Erk2 (25 units, New England Biolabs), insect-cell-derived GSK-3β (100 ng), E1 ubiquitin-activating enzyme (50 ng), Ubc5 (200 ng), ubiquitin (1 mg/ml), 1 μM ubiquitin aldehyde, 2.3 μl of in vitro translated FLAG-tagged F-box protein, and 4 mM ATP in a total volume of 10 μl (60 min, 30°C).

Results

Fbw7 Negatively Regulates c-Myc Turnover and Function. The Fbw7 gene encodes three protein isoforms (Fbw7α, Fbw7β, and Fbw7γ) produced by alternatively spliced mRNAs (24-26). We found that cotransfection of each isoform with c-Myc greatly decreased c-Myc expression, whereas nonfunctional Fbw7 mutants lacking the F-box (ΔF) did not (Fig. 1_A_). A cancer-associated Fbw7 inactivating mutation (R298H, 298 residues from the first shared exon) also severely crippled Fbw7-mediated c-Myc elimination (Fig. 1_B_). When the c-Myc binding partner Max was included, Fbw7 eliminated c-Myc but did not alter Max abundance (data not shown). Fbw7 also eliminated N-Myc expression in a manner analogous to c-Myc (not shown).

We next asked whether Fbw7-driven c-Myc turnover was proteasome-dependent and found that the elimination of c-Myc by Fbw7 was reversed by the proteasome inhibitor MG-132 (Fig. 1_C_). To confirm that the reduced c-Myc abundance resulted from increased c-Myc proteolysis, we measured c-Myc half-life by pulse-chase analyses and found that cotransfection of Fbw7γ with c-Myc reduced the half-life of c-Myc from ≈30 min to 13 min (Fig. 1_D_). In sum, these data show that Fbw7 promotes c-Myc turnover, and that this requires intact proteasomal and Fbw7 functions.

Consistent with its role as an oncogene (37, 38), Skp2 expression increases c-Myc transcriptional activity (19, 20). Because Fbw7 is a tumor suppressor, we predicted that it should inhibit c-Myc activity. We thus determined the impact of Fbw7 expression on c-Myc transactivation by using an established c-Myc reporter assay (35). We found that, in contrast with Skp2, Fbw7 coexpression inhibited c-Myc transcriptional activity in a dose-dependent manner (Fig. 1_E_).

Fbw7 and c-Myc Physically Interact. To determine whether Fbw7 and c-Myc associate in vivo, we immunoprecipitated either c-Myc or Fbw7 from lysates of cells expressing Myc, Fbw7, or both (Fig.2 and data not shown). c-Myc and Fbw7 efficiently coprecipitated when c-Myc turnover was prevented by proteasome inhibition (Fig. 2 A, compare lanes 3 and 7). In contrast, two other F-box proteins that also contain WD40 repeats (Fbw1/β-TRCP and Fbw6) neither bound to, nor eliminated, c-Myc (Fig. 2 A). We extended these findings by determining whether endogenous c-Myc binds to near-physiologic levels of Fbw7 by developing retroviral vectors that express low levels of either Fbw7α or Fbw7γ. Both Fbw7α and Fbw7γ coprecipitated with endogenous c-Myc, and in each case the binding was sensitive to proteasome inhibition (Fig. 2_B_). These data indicate that c-Myc and Fbw7 specifically interact and exhibit stable binding that is sensitive to proteasome function.

The physical interaction of Fbw7 and c-Myc is sensitive to proteasomal function. (A) 293 cells were transfected as indicated; vec, vector alone. Upper two gels, c-Myc and F-box protein abundance in lysates. Lower two gels, c-Myc and F-box protein abundance in anti-F-box protein (anti-FLAG) immunoprecipitates (IP). MG-132 treatment is indicated. (B) HeLa cells were transduced with retroviruses expressing FLAG-Fbw7α or FLAG-Fbw7γ and treated with MG-132 as shown, and endogenous c-Myc and Fbw7 abundance in lysates and anti-Fbw7 (FLAG) immunoprecipitates is shown. Asterisk indicates a background band.

Fbw7-Catalyzed c-Myc Turnover Requires c-Myc Threonine-58 and GSK-3. Fig. 3_A_ shows an alignment between the cyclin E T380 region [a high-affinity Fbw7-binding site (24, 26, 33)] and the c-Myc T58 region. There is striking homology between these regions, and c-Myc T58 completely conforms to the high-affinity consensus phospho-binding motif for Cdc4, termed the Cdc4 phosphodegron, or CPD (22). Like c-Myc T58 (7, 10, 39, 40), cyclin E T380 is also phosphorylated by GSK-3 (33), suggesting that cyclin E and c-Myc contain homologous Fbw7-binding motifs that are regulated by shared mitogenic signal transduction pathways.

We examined the role of T58 in Fbw7-driven c-Myc turnover, and we found that Fbw7 did not reduce the abundance of cotransfected c-MycT58A (in which T58 is changed to alanine, Fig. 3_B_). Mutation of S62, or both T58 and S62, also prevented Fbw7-mediated c-Myc elimination (Fig. 3_B_, compare lanes 4 and 5 with lanes 8 and 9). However, because T58 phosphorylation requires S62 phosphorylation, we cannot distinguish between the effects of S62 phosphorylation per se and defective T58 phosphorylation in the S62A mutant.

The requirement for c-MycT58 phosphorylation in Fbw7 binding was tested by two methods. We first cotransfected either c-Myc or c-MycT58A with dnFbw7WD, a dominant-negative Fbw7 mutant that contains only the WD40 repeats (41), and found that whereas WT c-Myc efficiently coprecipitated with dnFbw7WD, c-MycT58A did not (Fig. 3_C_). We used inactive Fbw7 in these experiments to separate binding from turnover. Conversion of T58 to S58 restored Fbw7 binding, although it was reduced (Fig. 3_C_, lanes 4 and 8), consistent with the report (22) that the cyclin E CPD interacts more efficiently with Cdc4 by means of phosphothreonine than phosphoserine. Finally, we found that the ability of dnFbw7WD to stabilize c-Myc correlated well with binding, and that WT c-Myc and c-MycT58S levels were increased by dnFbw7WD, but that c-MycT58A was not (Fig. 3_C_ Bottom).

We also used synthetic peptides corresponding to the c-Myc CPD to further define the role of T58 phosphorylation in mediating Fbw7 binding. The binding of these peptides to Fbw7 was strongly dependent upon T58 phosphorylation (Fig. 3_D_). Interestingly, although Ras-mediated S62 phosphorylation prevents c-Myc turnover in vivo (7), we found that the T58S62 doubly phosphorylated peptide still bound to Fbw7, indicating that the mechanism of S62 stabilization is not likely to involve inhibition of Fbw7 binding. In contrast, the Fbw6 protein did not exhibit phosphorylation-dependent binding to the c-Myc peptide. These data indicate that T58 phosphorylation regulates the physical interaction of Fbw7 with the c-Myc CPD and is required for Fbw7-driven c-Myc turnover.

There is substantial evidence that GSK-3 regulates c-Myc stability and is the primary c-Myc T58 kinase (7, 39). To test the role of endogenous GSK-3 in Fbw7-mediated c-Myc turnover, we inhibited GSK-3 by overexpression of a peptide corresponding to the axin GSK-3-interaction domain (GID) that binds to and inhibits GSK-3 (42) (Fig. 3_E_, compare lanes 3 and 4, and Fig. 6, which is published as supporting information on the PNAS web site), and by treatment with lithium, a pharmacologic GSK-3 inhibitor (Fig. 3_E_, compare lanes 3 and 6). In both cases we found that GSK-3 inhibition prevented Fbw7-driven c-Myc elimination. In contrast, a control GID peptide bearing a point mutation that prevents GSK-3 interaction did not prevent c-Myc turnover (Fig. 3_E_, lane 5). These data indicate that Fbw7-driven c-Myc turnover requires endogenous GSK-3 activity and strongly support the model that Fbw7-driven c-Myc turnover requires phosphorylation of c-Myc on T58 by GSK-3.

Regulation of Endogenous c-Myc by Endogenous Fbw7. We developed retroviral vectors that inhibit Fbw7 by stably expressing siRNA targeting the common region of Fbw7. Because antibodies that recognize endogenous Fbw7 are not available, we confirmed the efficacy of the Fbw7 siRNA by demonstrating that the siRNA decreased expression of retroviral or transfected FLAG-tagged Fbw7, using semiquantitative PCR and quantitative real-time PCR analyses of endogenous Fbw7 mRNA (Fig. 7, which is published as supporting information on the PNAS web site, and data not shown). The siRNAs led to a 50% or greater reduction in Fbw7 expression, and we found that stable Fbw7 knockdown by siRNA is rapidly selected against during cell culture (data not shown). Fbw7 siRNA increased endogenous c-Myc abundance in U2OS cells and primary human fibroblasts (Fig. 4_A_ Left and Center). To control for possible siRNA effects on genes other than Fbw7, we developed an additional Fbw7 common region siRNA (siFbw7-2), and this also increased endogenous c-Myc abundance (Fig. 4_A_ Right) and delayed c-Myc turnover in pulse-chase analyses (Fig. 4_B_). In contrast, a number of control proteins were unaffected by Fbw7 siRNA (Fig. 4_A_).

Endogenous c-Myc abundance is regulated by endogenous Fbw7. (A) U2OS cells, HFF, and HeLa cells were transduced with retroviral vectors encoding control (C) siRNA or two different Fbw7 siRNAs (U2OS and HFF cells, Fbw7-1 siRNA; HeLa cells, Fbw7-2 siRNA). Lysates were blotted for c-Myc, tubulin, GSK-3, Cdk2, and Hsp90. (B) Pulse-chase analysis of endogenous c-Myc stability in HeLa cells transduced with either si-Fbw7-2 or control retroviruses. (C) 293 cells were transfected with increasing amounts of dnFbw7WD and the amount of endogenous c-Myc protein is shown. (D) 293 cells were transfected with vector or dnFbw7WD and lysates were analyzed for total c-Myc (9E10) or pT58-c-Myc (p-Myc) protein after treatment with cycloheximide (CHX) for the indicated times (min).

The role of endogenous Fbw7 on endogenous c-Myc turnover was further examined by inhibiting endogenous Fbw7 activity with dnFbw7WD. dn-Fbw7WD expression increased c-Myc abundance and delayed endogenous c-Myc turnover in decay in the presence of cycloheximide and metabolic pulse-chase analyses (Fig. 4 C and D and data not shown). The effect of dnFbw7WD on endogenous T58-phosphorylated c-Myc turnover was also examined with a specific phospho-T58 antibody (p-Myc, Fig. 4_D_). The turnover of pT58-c-Myc was delayed by dnFbw7WD to an even greater extent than total c-Myc turnover, indicating that T58-phosphorylated c-Myc constitutes the fraction of total c-Myc that is most sensitive to dnFbw7WD.

Importantly, Fbw7 inhibition by siRNA or dnFbw7WD delays, but does not abrogate, c-Myc turnover. This finding is quite similar to what has been observed when c-Myc cannot be phosphorylated on T58. That is, c-MycT58A is more stable than WT c-Myc, but is still labile, indicating that T58 and Fbw7-independent mechanisms also degrade c-Myc (4, 7, 17, 39). In summary, the siRNA and dnFbw7WD approaches indicate that endogenous Fbw7 regulates endogenous c-Myc abundance and turnover.

Fbw7 Promotes c-Myc Ubiquitination in Vitro. Because our data predict that Fbw7 functions to recruit the SCF to c-Myc and promote its ubiquitination, we determined whether the SCFFbw7 could directly stimulate c-Myc ubiquitination in vitro. _In vitro_-translated c-Myc or c-MycT58A was incubated with reticulocyte extracts containing either Fbw7 or Fbw2, as well as other required ubiquitination components (including UbcH5 as the E2 enzyme). We added exogenous GSK-3 into these reactions because we found that c-Myc was not T58-phosphorylated in these extracts, but that T58 phosphorylation was greatly stimulated by GSK-3 (Fig. 8, which is published as supporting information on the PNAS web site). Fbw7 specifically converted WT c-Myc into higher molecular weight products but Fbw2 did not (Fig. 5). Moreover, this conversion required both T58 and GSK. Similar results were obtained when we used wheat germ extract instead of reticulocyte extract as the source of c-Myc (Fig. 5_C_). Consistent with the multiple ubiquitination pathways that target c-Myc, we found a small amount of background activity that depended upon the amount of reticulocyte extract but not T58 phosphorylation. Thus Fbw7 directly catalyzed c-Myc ubiquitination in a T58- and GSK-3-dependent manner.

Fbw7 catalyzes phosphorylation-dependent c-Myc ubiquitination in vitro.(A) c-Myc or c-MycT58A was translated in vitro by using reticulocyte extracts in the presence of GSK-3 and subjected to in vitro ubiquitination reactions (see text). Reticulocyte extract containing Fbw7 was added as indicated. (B) In vitro ubiquitination of c-Myc in reticulocyte-based extracts as described above except that GSK-3 and Erk were added as indicated, and all samples contained Fbw7. (C) c-Myc or c-MycT58A was translated in wheat germ extracts containing either Fbw2 or Fbw7 as indicated, and in vitro ubiquitination reactions were performed.

Discussion

We have shown that Fbw7 regulates phosphorylation-dependent c-Myc degradation. Fbw7 had previously been found to promote the degradation of cyclin E after phosphorylation within the CPD (24-26, 33). We now demonstrate that the c-Myc protein is also targeted for degradation by Fbw7 subsequent to phosphorylation within a domain that is highly related to the cyclin E CPD. c-Myc turnover is augmented upon Fbw7 overexpression, and Fbw7-mediated c-Myc degradation is impaired by mutation of either the Fbw7 F-box or β-propeller regions, or by pharmacologic proteasome inhibition. The role of endogenous Fbw7 in endogenous c-Myc protein degradation was studied by siRNA targeting Fbw7, as well as a parallel approach employing dnFbw7. These experiments established that endogenous Fbw7 inhibition by either approach augmented c-Myc abundance and delayed c-Myc turnover. Thus Fbw7 overexpression or inhibition caused c-Myc abundance to decrease or increase, respectively. We also found that c-Myc specifically interacts with Fbw7, and this is regulated by proteasomal function. These data fit well with the work of Moberg et al. (21), who report that Ago, the Drosophila ortholog of Fbw7, binds dMyc and regulates its abundance and cell growth function.

The c-Myc T58 region has been previously identified as a crucial signal for c-Myc degradation dependent on phosphorylation of T58 by GSK-3 (4, 7, 10, 40). Our data show that (i) T58 and endogenous GSK-3 activity are required for Fbw7-mediated Myc degradation in vivo; (ii) T58 phosphorylation regulates the binding of Fbw7 to the c-Myc CPD; and (iii) Fbw7 directly promotes c-Myc ubiquitination in vitro, and this requires both T58 and GSK-3. In sum, these data indicate that Fbw7 mediates T58 phosphorylation-dependent c-Myc turnover.

That two distinct SCF receptor proteins bind c-Myc and promote its degradation suggests that different aspects of c-Myc function are being regulated. Skp2 binds the basic-helix—loop—helix-leucine zipper (bHLHZ) and Myc box II (MBII) regions of c-Myc (19, 20). Both of these regions are critical for Myc transcriptional activity, consistent with the model that ubiquitination of short-lived transcription factors is tightly coupled to their transcriptional activities (43). However, Skp2 is unlikely to be the sole regulator of c-Myc stability. Unlike most SCF complexes, which are specialized to recognize phosphorylated substrates, Skp2 binds to domains of c-Myc that lack known phosphorylation sites, and Skp2-mediated c-Myc turnover is phosphorylation independent (19, 20). Thus Skp2 cannot regulate T58-dependent c-Myc turnover.

In contrast to Skp2, Fbw7-mediated c-Myc turnover requires T58 phosphorylation within a classic CPD. T58 phosphorylation by GSK-3 regulates c-Myc stability during the mitogenic response (6, 7, 10, 39). Furthermore, whereas Skp2 increases c-Myc transcriptional activity (19, 20), we show here that Fbw7 acts to decrease it. These opposing effects on c-Myc transcriptional activity are consistent with the observations that Skp2 functions as a dominant oncogene, but Fbw7 functions as a recessive tumor suppressor gene. It is therefore likely that the SCFSkp2 and SCFFbw7 pathways perform very different functions with respect to c-Myc regulation. We speculate that Skp2 promotes c-Myc turnover in concert with its transcriptional activation, whereas Fbw7 negatively regulates c-Myc in contexts where c-Myc phosphorylation is regulated by GSK-3 and Ras.

The fbw7 gene is found within 4q32, a chromosomal region that is mutated in many human cancers, and documented Fbw7 mutations in cancer cells supported its identification as a human tumor suppressor (25-27). The observation that c-Myc and cyclin E are both Fbw7 substrates likely reflects coordinate regulation of these proteins in normal cells. Indeed, cyclin E and c-Myc are each absent from quiescent cells and exhibit periodic expression after mitogenic stimulation. Moreover, cyclin E and c-Myc are each phosphorylated by GSK-3 within their CPD motifs, and both proteins become resistant to Fbw7-driven turnover when GSK-3 is inhibited (33). Thus c-Myc and cyclin E are key regulators of cell growth and division that share elements within the signal transduction and proteolytic pathways that regulate their expression. In addition, Fbw7 regulates Notch and c-Jun abundance (41, 44). Thus cancer-associated mutations that disrupt Fbw7 would be expected to deregulate cyclin E, c-Myc, c-Jun, and Notch signaling in a single mutation, and Fbw7 loss in cancers may simultaneously deregulate the pathways that regulate cell division, cell growth, apoptosis, and cell differentiation. In conclusion, Fbw7 is part of an evolutionarily conserved pathway permitting phosphorylation-dependent control of c-Myc in response to extracellular signals.

Supplementary Material

Acknowledgments

We thank Ken Moberg and Iswar Hariharan for prepublication information, Lee Madrid and Carla Grandori for advice and reagents, and Jherek Swanger and Jason Yada for technical help. This work was supported by National Cancer Institute Grants R01CA84069 and R01CA102742-01 (B.E.C.) and R01CA20525 (R.N.E), National Institutes of Health Grant AG11085 (J.W.H.), the Leukemia and Lymphoma Society (M.W.), the Human Frontiers Science Program (A.O.), and the Department of Defense (J.J.). B.E.C. is a W. M. Keck Distinguished Young Scholar, and R.N.E. is an American Cancer Society Research Professor.

Notes

Abbreviations: GSK-3, glycogen synthase kinase 3; SCF, Skp-Cullin-F-box; siRNA, short interfering RNA; CPD, Cdc4 phosphodegron; GID, GSK-interacting domain.

See Commentary on page 8843.

References

2. Ciechanover, A., DiGiuseppe, J. A., Bercovich, B., Orian, A., Richter, J. D., Schwartz, A. L. & Brodeur, G. M. (1991) Proc. Natl. Acad. Sci. USA 88**,** 139-143. [PMC free article] [PubMed] [Google Scholar]

3. Gross-Mesilaty, S., Reinstein, E., Bercovich, B., Tobias, K. E., Schwartz, A. L., Kahana, C. & Ciechanover, A. (1998) Proc. Natl. Acad. Sci. USA 95**,** 8058-8063. [PMC free article] [PubMed] [Google Scholar]

6. Sears, R., Leone, G., DeGregori, J. & Nevins, J. R. (1999) Mol. Cell 3**,** 169-179. [PubMed] [Google Scholar]

7. Sears, R., Nuckolls, F., Haura, E., Taya, Y., Tamai, K. & Nevins, J. R. (2000) Genes Dev. 14**,** 2501-2514. [PMC free article] [PubMed] [Google Scholar]

9. Henriksson, M., Bakardjiev, A., Klein, G. & Luscher, B. (1993) Oncogene 8**,** 3199-3209. [PubMed] [Google Scholar]

11. Pulverer, B. J., Fisher, C., Vousden, K., Littlewood, T., Evan, G. & Woodgett, J. R. (1994) Oncogene 9**,** 59-70. [PubMed] [Google Scholar]

12. Yeh, E., Cunningham, M., Arnold, H., Chasse, D., Monteith, T., Ivaldi, G., Hahn, W. C., Stukenberg, P. T., Shenolikar, S., Uchida, T., et al. (2004) Nat. Cell Biol. 6**,** 308-318. [PubMed] [Google Scholar]

13. Bhatia, K., Huppi, K., Spangler, G., Siwarski, D., Iyer, R. & Magrath, I. (1993) Nat. Genet. 5**,** 56-61. [PubMed] [Google Scholar]

14. Yano, T., Sander, C. A., Clark, H. M., Dolzeal, M. V., Jaffe, E. S. & Raffeld, M. (1993) Oncogene 8**,** 2741-2748. [PubMed] [Google Scholar]

15. Bhatia, K., Spangler, G., Gaidano, G., Hamdy, N., Dalla-Favera, R. & Magrath, I. (1994) Blood 84**,** 883-888. [PubMed] [Google Scholar]

16. Gavine, P. R., Neil, J. C. & Crouch, D. H. (1999) Oncogene 18**,** 7552-7558. [PubMed] [Google Scholar]

17. Bahram, F., von der Lehr, N., Cetinkaya, C. & Larsson, L. G. (2000) Blood 95**,** 2104-2110. [PubMed] [Google Scholar]

18. Law, W. & Linial, M. L. (2001) Oncogene 20**,** 1118-1127. [PubMed] [Google Scholar]

19. Kim, S. Y., Herbst, A., Tworkowski, K. A., Salghetti, S. E. & Tansey, W. P. (2003) Mol. Cell 11**,** 1177-1188. [PubMed] [Google Scholar]

20. von der Lehr, N., Johansson, S., Wu, S., Bahram, F., Castell, A., Cetinkaya, C., Hydbring, P., Weidung, I., Nakayama, K., Nakayama, K. I., et al. (2003) Mol. Cell 11**,** 1189-1200. [PubMed] [Google Scholar]

21. Moberg, K. H., Mukherjee, A., Veraksa, A., Artavanis-Tsakonas, S. & Hariharan, I. K. (April 29, 2004) Curr. Biol. 14**,** 10.1016/S0960982204003070. [PubMed]

22. Nash, P., Tang, X., Orlicky, S., Chen, Q., Gertler, F. B., Mendenhall, M. D., Sicheri, F., Pawson, T. & Tyers, M. (2001) Nature 414**,** 514-521. [PubMed] [Google Scholar]

23. Orlicky, S., Tang, X., Willems, A., Tyers, M. & Sicheri, F. (2003) Cell 112**,** 243-256. [PubMed] [Google Scholar]

24. Koepp, D. M., Schaefer, L. K., Ye, X., Keyomarsi, K., Chu, C., Harper, J. W. & Elledge, S. J. (2001) Science 294**,** 173-177. [PubMed] [Google Scholar]

25. Moberg, K. H., Bell, D. W., Wahrer, D. C., Haber, D. A. & Hariharan, I. K. (2001) Nature 413**,** 311-316. [PubMed] [Google Scholar]

26. Strohmaier, H., Spruck, C. H., Kaiser, P., Won, K. A., Sangfelt, O. & Reed, S. I. (2001) Nature 413**,** 316-322. [PubMed] [Google Scholar]

27. Spruck, C. H., Strohmaier, H., Sangfelt, O., Muller, H. M., Hubalek, M., Muller-Holzner, E., Marth, C., Widschwendter, M. & Reed, S. I. (2002) Cancer Res. 62**,** 4535-4539. [PubMed] [Google Scholar]

28. Rajagopalan, H., Jallepalli, P. V., Rago, C., Velculescu, V. E., Kinzler, K. W., Vogelstein, B. & Lengauer, C. (2004) Nature 428**,** 77-81. [PubMed] [Google Scholar]

29. Tetzlaff, M. T., Yu, W., Li, M., Zhang, P., Finegold, M., Mahon, K., Harper, J. W., Schwartz, R. J. & Elledge, S. J. (2004) Proc. Natl. Acad. Sci. USA 101**,** 3338-3345. [PMC free article] [PubMed] [Google Scholar]

30. Tsunematsu, R., Nakayama, K., Oike, Y., Nishiyama, M., Ishida, N., Hatakeyama, S., Bessho, Y., Kageyama, R., Suda, T. & Nakayama, K. I. (2004) J. Biol. Chem. 279**,** 9417-9423. [PubMed] [Google Scholar]

32. Grandori, C., Wu, K. J., Fernandez, P., Ngouenet, C., Grim, J., Clurman, B. E., Moser, M. J., Oshima, J., Russell, D. W., Swisshelm, K., et al. (2003) Genes Dev. 17**,** 1569-1574. [PMC free article] [PubMed] [Google Scholar]

33. Welcker, M., Singer, J., Loeb, K., Grim, J., Bloecher, A., Gurien-West, M., Clurman, B. E. & Roberts, J. M. (2003) Mol. Cell 12**,** 381-392. [PubMed] [Google Scholar]

34. Clurman, B. E., Sheaff, R. J., Thress, K., Groudine, M. & Roberts, J. M. (1996) Genes Dev. 10**,** 1979-1990. [PubMed] [Google Scholar]

35. Laherty, C. D., Yang, W.-M., Sun, J.-M., Davie, J. R., Seto, E. & Eisenman, R. N. (1997) Cell 89**,** 349-356. [PubMed] [Google Scholar]

36. Reed, S. E., Spruck, C. H., Sangfelt, O., Van Drogen, F., Mueller-Holzner, E., Widschwendter, M., Zetterberg, A. & Reed, S. I. (2004) Cancer Res. 64**,** 795-800. [PubMed] [Google Scholar]

37. Gstaiger, M., Jordan, R., Lim, M., Catzavelos, C., Mestan, J., Slingerland, J. & Krek, W. (2001) Proc. Natl. Acad. Sci. USA 98**,** 5043-5048. [PMC free article] [PubMed] [Google Scholar]

38. Latres, E., Chiarle, R., Schulman, B. A., Pavletich, N. P., Pellicer, A., Inghirami, G. & Pagano, M. (2001) Proc. Natl. Acad. Sci. USA 98**,** 2515-2520. [PMC free article] [PubMed] [Google Scholar]

39. Gregory, M. A., Qi, Y. & Hann, S. R. (2003) J. Biol. Chem. 278**,** 51606-51612. [PubMed] [Google Scholar]

41. Wu, G., Lyapina, S., Das, I., Li, J., Gurney, M., Pauley, A., Chui, I., Deshaies, R. J. & Kitajewski, J. (2001) Mol. Cell. Biol. 21**,** 7403-7415. [PMC free article] [PubMed] [Google Scholar]

42. Hedgepeth, C. M., Deardorff, M. A., Rankin, K. & Klein, P. S. (1999) Mol. Cell. Biol. 19**,** 7147-7157. [PMC free article] [PubMed] [Google Scholar]

43. Salghetti, S. E., Caudy, A. A., Chenoweth, J. G. & Tansey, W. P. (2001) Science 293**,** 1651. [PubMed] [Google Scholar]

44. Oberg, C., Li, J., Pauley, A., Wolf, E., Gurney, M. & Lendahl, U. (2001) J. Biol. Chem. 276**,** 35847-35853. [PubMed] [Google Scholar]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences