Degradation of MyoD mediated by the SCF (MAFbx) ubiquitin ligase

,

. In addition MyoD phosphorylated on serine 200 is degraded by the proteasome and CDC34 ubiquitin-conjugating enzyme activity (E2) (19).Altogether these findings indicate that targeted degradation of MyoD following Cdk2-cyclin E phosphorylation of serine 200 during G 1 to S phase progression in myoblasts and identify a mechanism that probably requires a SCF complex.On the other hand, recent data have shown that MyoD could be degraded by the ubiquitin-proteasome pathway independently of its phosphorylation state suggesting that attachment of ubiquitin implies other(s) recognition signal(s) and/or mechanism(s).Up-regulation of the cyclin/Cdk inhibitor p57 Kip2 stabilizes MyoD by direct interaction (22), whereas specific DNA binding stabilizes MyoD (23) and in vitro MyoD is degraded via the ubiquitin-proteasome pathway in HeLa nucleoplasm (24), and in differentiated myotubes, MyoD protein turns over rapidly as observed in myoblasts (25).Altogether these data indicate that MyoD is degraded by the ubiquitin-proteasome pathway mediated by different E3 ubiquitin ligases not identified as yet.
To determine candidate molecular mediators of MyoD ubiquitination, we employed a modified version of the yeast twohybrid screen of the human skeletal muscle library to identify proteins that first interact with human Skp1.This Skp1-en-riched cDNA library was then screened to isolate F-box proteins that bind MyoD and identify positive clones.Here, we report the identification of human Atrogin-1/MAFbx (MAFbx), an E3 ubiquitin ligase up-regulated during skeletal muscle atrophy that interacts with MyoD via a novel leucine-rich interaction interface.This interface found in transcriptional coactivators mediates the binding to liganded nuclear receptors and promote its degradation.We demonstrated that the purified recombinant SCF MAFbx complex mediated ubiquitination of MyoD in vitro in a manner that appeared to depend on the presence of the lysine 133.This lysine 133 has been previously shown to play a critical role in the nuclear degradation of MyoD (26).Finally, we show that up-regulation of MAFbx in proliferating myoblasts antagonizes differentiation inducing MyoD degradation and preventing muscle specific-gene activation.

EXPERIMENTAL PROCEDURES
Yeast Two-hybrid Screen and DNA Manipulations-We used the Matchmaker Two-Hybrid Kit with a human skeletal muscle cDNA library kit (Clontech).The assays were performed as recommended by the manufacturer.The full-length cDNA clone of human Skp1 (bait) was subcloned in-frame with the Gal4 DNA-binding domain in the pGBT9 vector, resulting in pGBT9-Skp1.We screened 6 ϫ 10 6 clones (corresponding to two library titers) and selected the positive clones as suggested by the manufacturer.The liquid ␤-galactosidase assay was performed according to the manufacturer's instruction.The Skp1-enriched cDNA library was then screened against pGBT9-MyoD (bait) and we selected 100 positive clones.Oligonucleotides were synthesized by Proligo.Deletion in the F-box of MAFbx was introduced by PvuII digestion of pACT2-MAFbx and self-annealing of the resulting plasmid.The full-length coding sequences as well as mutants of MAFbx and MyoD were amplified by PCR to introduce BamHI and EcoRI sites on each side of the open reading frame and cloned as BamHI-EcoRI fragments into the eukaryotic expression vectors.The bicistronic expression plasmid for MAFbx was carrying out first by using a 1400-bp EcoRI-HindIII fragment containing the internal ribosome entry site (IRES) and GFP from the pMIGR vector and subcloned at the EcoRI and HindIII sites of pcDNA4T0 (Invitrogen).MAFbx and the ⌬F-box mutant were then subcloned at the EcoRI site of pcDNA4-IRES-GFP.
Cell Culture, Transfections, and Luciferase Assays-The mouse skeletal muscle cell line C2C12 and the fibroblast cell line 10T1/2 were maintained in growth Dulbecco's modified Eagle's medium supplemented with antibiotics and, respectively, 20 and 15% fetal calf serum.Cells were transfected by using the Jet PEI (QBiogene).Luciferase activity was determined using 10-l aliquots of cell extracts from harvested cells in 1ϫ reporter lysis buffer (Promega) with equivalent quantities of proteins in triplicate and repeated at least twice.
RNA Extraction and Reverse Transcription-PCR Analyses-Total RNA was isolated from C2C12 cells using TRI Reagent (Sigma).For semiquantitative reverse transcription-PCR analysis, 1 g of RNA was used to perform reverse transcription with Superscript II RnaseH Reverse Transcriptase (Invitrogen) and hexanucleotides (Roche Diagnostics).2 l of each reaction served as a template for PCR analysis.Primer pairs for the amplification of the gene products were MAFbx 5Ј-GGG-GGAAGCTTTCAACAG-3Ј forward, 5Ј-TGAGGCCTTTGAAGGCAG-3Ј reverse and glyceraldehyde-3-phosphate dehydrogenase 5Ј-GAGCTGA-ACGGGAAGCTCACT-3Ј forward, 5Ј-TTGTCATACCAGGAAATGAG-C-3Ј reverse.
Immunofluorescence Staining-Cells were cultured on coverslips and fixed in 3% paraformaldehyde for 10 min at 4 °C, and permeabilized with 0.25% Triton X-100 for 30 min at room temperature.The cells were treated with 5% normal goat serum and immunostained with anti-MyoD mAb (5.A8) The Texas Red-conjugated F(abЈ) 2 fragment of donkey antimouse IgG was used to visualize the mouse monoclonal antibodies.

Identification of MAFbx as an F-box Protein Interacting with
MyoD-The CDC34 ubiquitin-conjugating enzyme activity (E2) has been implicated in MyoD ubiquitination (19) suggesting the requirement of an Skp1/Cul1/F-box protein complex (E3).To identify candidate molecular mediators of MyoD ubiquitination and degradation, we used a human skeletal muscle library enriched for F-box-expressing genes and MyoD as bait in the yeast two-hybrid screen.We identified two positive clones with the same open reading frame, one clone (clone R8, 1.6-kbinsert) coded for 350 amino acids and the second clone (clone R5, 0.95-kb insert) encoded a polypeptide of 259 amino acids (Fig. 1A) highly homologous with the mouse F-box protein Atrogin-1/MAFbx (MAFbx) (27,28).The full-length human MAFbx lacks the LRR and WD40 consensus motifs known to interact with protein substrates (29,30) In addition to the PDZ motif and cytochrome c family heme binding site located in the C terminus domain, MAFbx contains two potential nuclear localization signals, a leucine zipper domain and a leucine-charged residue-rich domain

FIG. 2. MAFbx and MyoD proteins interact via a LXXLL motif.
A, association between MAFbx and MyoD in C2C12 myoblasts and myotubes.Cells were grown in proliferation medium (Mb) and/or in differentiation medium (Mt) with 50 M MG132 (ϩ) for 2h, after which cell lysates were immunoprecipitated with anti-MAFbx 61-74 antibod-ies.The resulting precipitates were subjected to immunoblot (IB) analysis with antibodies to MyoD or MAFbx 326 -339, as indicated.B, interaction between MAFbx and various mutants of MyoD.The ability of each MyoD mutant to bind MAFbx is summarized (ϩ or Ϫ).Amino acid regions corresponding to the putative MAFbx-MyoD interacting sequence and schematic representation of mutant MyoD L164Q generated by site-directed mutagenesis.C, empty plasmid (lane 1) or expression plasmids encoding HA-MyoD L164Q (lane 2) or HA-MyoD-wt (lane 3) were cotransfected with an expression plasmid encoding FLAG-MAFbx into 10T1/2 cells.Cell lysates were processed for immunoprecipitation with anti-FLAG antibodies (M2).Immunoprecipitates were then analyzed by Western blot with anti-Flag and anti-HA, respectively (bottom panel).Aliquots of lysates were processed for Western blotting with anti-Flag and anti-HA (top panel).Note that anti-FLAG antibodies failed to immunoprecipitate HA-MyoD.D, identification of the sequence in MAFbx required for MyoD interaction.10T1/2 cells were cotransfected with MyoD and wild type (full) and/or deletion mutants of MAFbx.(ϩ) and (Ϫ) indicate whether or not the various mutants interact with MyoD.Representation of the putative MyoD-binding domain in MAFbx.Highly conserved LCD in mammals is underlined, and leucine residues are in bold.E, empty vector (lane 1) or expression plasmids encoding FLAG-MAFbx-wt (lane 2) or FLAG-MAFbx L169Q (lane 3) were transfected into 10T1/2 cells together with an expression plasmid encoding HA-MyoD.Whole cell lysates were processed for immunoprecipitation with anti-FLAG antibodies.Immunoprecipitates were then analyzed by Western blot with anti-FLAG and anti-HA, respectively (bottom panel).Aliquots of lysates were processed for Western blotting with anti-FLAG and anti-HA (top panel).WCE, whole cell extract.(LCD), which are conserved between human, rat, and mouse species (supplemental Fig. 1).The presence of various proteinprotein interacting domains suggests that MAFbx could potentially recognize different substrates.the C2C12 cell line, a well defined model for ex vivo differentiation.These cells proliferate as myoblasts in high serum concentrations and can be induced to differentiate by reducing the serum concentration from 20 to 2%.Under our conditions, C2C12 myoblasts fuse into myotubes within 48 -60 h with an efficiency of 75-80%.A semiquantitative reverse transcription-PCR analysis showed that the MAFbx mRNA was up-regulated (2-3-fold) in the course of C2C12 differentiation (Fig. 1B).Anti-MAFbx antibodies were developed in our laboratory (see "Experimental Procedures"), and Western blot analysis revealed that the MAFbx protein is barely detectable in myoblasts and accumulates in C2C12 myotubes.By this time, MyoD is up-regulated after switching cells into the differentiation medium (DM) and then decreases during differentiation; troponin T, a skeletal muscle marker, is observed only in differentiated cells (Fig. 1C).Thus, during the course of C2C12 differentiation MyoD and MAFbx showed a contrasting pattern of expression.

Expression of MAFbx and MyoD during C2C12 Muscle
Phosphorylation or Acetylation of MyoD Is Not Necessary for Interaction with MAFbx-We first determined the binding of MyoD with MAFbx in mammalian cells by using a coimmunoprecipitation assay.In C2C12 myogenic cells, endogenous MyoD coimmunoprecipitated with endogenous MAFbx, and an increase in MyoD/MAFbx association was observed when proteasome activity was inhibited (Fig. 2A).In transiently transfected 10T1/2 cells, MyoD and MAFbx have a nuclear localization, and MyoD coimmunoprecipitated with MAFbx but did not associate with other F-box proteins such as Skp2 or FBX03 (supplemental Fig. 2, A and B).Furthermore, Myf-5, the second determination factor that is also expressed in myoblasts, did not coimmunoprecipitate with MAFbx in cotransfection experiments (supplemental Fig. 2C).
Posttranslational modifications of target proteins are implicated in the recognition by certain SCF-type E3 ubiquitin ligases and subsequent degradation by the 26 S proteasome (4,31,32).On the other hand, ubiquitination of substrates requires direct binding of specific domains of the F-box protein and the substrate (33).The main posttranslational modifications of MyoD that implicate phosphorylation of Ser-5 and Ser-200 (18,20) and/or acetylation by CBP/p300 (8) and p/CAF (9) were tested for their interaction with MAFbx.10T1/2 cells were transiently transfected with plasmids encoding MAFbx alone or with MyoD-wt, MyoD S200D, which mimics constitutive serine 200 phosphorylation, MyoD S5A/S200A, and MyoD/K 3 R in which all lysines were mutated to arginines.Surprisingly MyoD-wt and the MyoD mutants all interacted with MAFbx suggesting that major phosphorylation and acetylation are not implicated in MyoD/MAFbx interaction (supplemental Fig. 2D).Altogether, these results show that independent of the phosphorylation and/or acetylation, MyoD and MAFbx form complexes in vivo LCD of MAFbx Interacts with a Highly Conserved LXXLL Motif in MyoD-The results outlined above show that MAFbx binds to MyoD, but the lack of known protein binding motifs found in other F-box proteins prompted us to search for regions of MAFbx important for substrate recognition.The domains of both proteins required for interaction were mapped by in vivo protein binding experiments.10T1/2 cells were cotransfected with expression vectors encoding MyoD-wt and/or MyoD deletion or substitution mutants together with MAFbx, and the immunoprecipitates were then examined for the presence of MAFbx by an immunoblot analysis (supplemental Fig. 3A).Stable association between MyoD and MAFbx was mediated by the helix 2 domain of MyoD (Fig. 2B).Sequence analysis of amino acids 143-172 revealed a LXXLL consensus sequence that mediates protein-protein interactions and was originally found in a variety of coactivators (34).To assess the contribution of this motif to the interaction, we mutated the leucine 164 to glutamine (L164Q) and tested its interaction with MAFbx as described above.Mutation in the LXXLL motif was sufficient to reduce MyoD interaction with MAFbx (Fig. 2C).
In vivo reciprocal binding experiments were performed to define the sequence in MAFbx that contributes to complex formation with MyoD.To avoid confusion between the IgG band and the MAFbx mutants, we used a mammalian expression plasmid encoding GST-MyoD (supplemental Fig. 3B).Overlapping deletion mutant analysis showed that the central part of MAFbx (amino acids 140 -222) was implicated in the MyoD binding (Fig. 2D).This region contains a LCD that shares a consensus inverted LXXLL motif flanked by LXXL and LXXXL sequences highly conserved in mammals.We found that mutation (L169Q) in the inverted LXXLL motif dramatically reduced the binding to MyoD (Fig. 2E).These results demonstrated that complex formation depends upon the integrity of the inverted LXXLL motif in MAFbx.
Accelerated Degradation of MyoD by MAFbx-The effect of MAFbx expression on the MyoD half-life was investigated after blocking protein synthesis with cycloheximide and/or in pulsechase experiments.As expected the empty vector did not affect MyoD stability (half-life was to ϳ45 min).Overexpression of MAFbx resulted in enhanced degradation of MyoD by decreasing its half-life to ϳ30 min.By contrast Skp2 and/or FBL5 was ineffective on MyoD degradation.In cells with compromised SCF MAFbx activity, MyoD degradation was affected; the MAFbx mutant ⌬F-box (deleted of the F-box domain required for Skp1 interaction) showed an increased MyoD half-life (ϳ150 min) (Fig. 3A).In cells with compromised LXXLL motif interaction between MyoD and MAFbx, the degradation rate of MyoD remained similar to that of MyoD-wt alone (Fig. 3, B and C).It appears that destabilization of MyoD is increased in cells with SCF MAFbx E3-ligase activity and highlights the LXXLL motif that mediates MyoD/MAFbx interaction.
Overexpression of MAFbx in C2C12 myotubes results in atrophy (28), and on the other hand, during the course of C2C12 differentiation MyoD and MAFbx showed a contrasting pattern of expression (Fig. 1C).Together these data suggested that MAFbx-inducing C2C12 atrophy implicated MyoD degradation.We observed that increasing amounts of MAFbx on MyoD-mediated myogenic conversion of 10T1/2 cells caused a gradual decrease of MCK-Luc activity and MyoD protein abundance (Fig. 4A).To locate and assess the effect of MAFbx overexpression on MyoD proteins, C2C12 myoblasts were transfected with pcDNA4-MAFbx-IRES-GFP or the mutant MAFbx-⌬F-box-IRES-GFP, and 48 h later C2C12 myoblasts (Fig. 4B) were stained with anti-MyoD antibodies and/or induced to differentiate in low serum medium (Fig. 4C).In C2C12 myoblasts transfected with MAFbx (green), MyoD staining was selectively lost, whereas in C2C12 myoblasts transfected with the empty vector and/or the ⌬F-box mutant, MyoD staining was observed in the nucleus (Fig. 4B).Furthermore overexpression of MAFbx-IRES-GFP was never observed during myotube formation, and the great majority of the transfected cells were lost.The few MAFbx-expressing C2C12 myoblasts (green) maintained the mononucleated, non-differentiated phenotype and did not express MyoD.In contrast, C2C12 cells transfected with empty vector expressed MyoD and fused into multinucleated myotubes, with a higher level in the presence of the mutant ⌬F-box (Fig. 4C).Altogether these data suggest that high levels of MAFbx are incompatible with MyoD protein expression.
Mediation of MyoD Ubiquitination by the SCF MAFbx -Because MAFbx mRNA was identified as a marker that is upregulated in multiple models of skeletal muscle atrophy in mice, MAFbx protein expression was first tested in two models of in vivo atrophy, aging and food deprivation.Then the association of MAFbx with the essential Skp1 and Cul1 proteins, specific components of an E3 ubiquitin-protein ligase, was tested by using MyoD as a substrate in the ubiquitination assays.The level of MAFbx protein (42 kDa) in skeletal muscle FIG. 6.The SCF MAFbx recombinant complex ubiquitinates MyoD.A, the recombinant SCF MAFbx complex purified from Sf9 insect cells lysates was analyzed by SDS-PAGE and then by immunoblotting (IB) and/or staining with Coomassie Brilliant Blue (CBB).B, the recombinant SCF MAFbx complex was assayed for the ability to mediate the ubiquitination of HA-MyoD in the presence of E1, Cdc34, and ubiquitin (Ub).C, effect of mutation in the LXXLL motif on in vitro ubiquitination of MyoD.D, mutation of the lysine 133 in arginine represses MyoD ubiquitination by the recombinant SCF MAFbx complex.increased during aging and/or food deprivation (Fig. 5A).Total extracts from the hind leg muscles were then analyzed for the association of Cul1 and Skp1 with MAFbx.Coimmunoprecipitation of MAFbx, Cul1, and Skp1 were observed in normal mouse skeletal muscle, and increased amounts were found in atrophic mice.These complexes were not observed after immunoprecipitation with a preimmune serum (Fig. 5B).In an in vitro ubiquitination assay using a MyoD mutant with the amino-terminal three amino acids deleted (HA-tagged MyoD) as the physiological substrate, the SCF MAFbx from normal skeletal muscle exhibited ubiquitination activity, which was higher in atrophic muscle (Fig. 5C).Finally to determine in vitro whether the SCF MAFbx complex mediates ubiquitination of MyoD, we expressed the four subunits of the complex (Skp1, Cul1, Rbx1, and MAFbx) in Sf9 cells.We then purified the recombinant complex to near homogeneity (Fig. 6A) and assayed it for its ability to catalyze the ubiquitination of HA-tagged MyoD in vitro in the presence of E1, Cdc34 (E2), and ubiquitin.A significant ubiquitination was detected with E1, E2, ubiquitin, and SCF MAFbx (Fig. 6B) but the polyubiquitination of the HAtagged MyoD mutant L164Q was dramatically reduced (Fig. 6C).We also examined whether the SCF MAFbx complex mediated the ubiquitination of a MyoD mutant K133R that has been recently shown to play a critical role in the nuclear degradation of MyoD (26).As observed in Fig. 6D, the MyoD mutant K133R showed a dramatically reduced polyubiquitination strongly suggesting the implication of this internal lysine in MyoD ubiquitination by the SCF MAFbx .
Finally, to test whether MAFbx could promote MyoD ubiquitination in vivo, expression vectors encoding MAFbx or the mutant ⌬F-box were cotransfected with MyoD and HA-ubiquitin into 10T1/2 cells.MyoD protein was immunoprecipitated with anti-MyoD antibodies and was probed with anti-HA to detect ubiquitinated MyoD proteins.In the presence of MAFbx, MyoD showed a higher degree of ubiquitination compared with the mock transfectant (Fig. 7A).The mutant ⌬F-box did not promote the ubiquitination of MyoD, although it remained associated with MyoD.In C2C12 myoblasts, overexpression of MAFbx increased MyoD ubiquitination whereas the MAFbx mutant L169Q like the empty vector were ineffective (Fig. 7B).This was also observed in 10T1/2 cells transiently cotransfected by expression plasmids encoding HA-tagged MyoD-wt, FLAGtagged MAFbx-wt, and/or the MAFbx mutant L169Q (Fig. 7C, lanes 1-3).The ubiquitination levels of the mutant Myo DL164Q were roughly similar to those of MyoD in the absence or presence of MAFbx (Fig. 7C, lanes 1, 4, and 5).These data suggest that MAFbx via the LXXLL core motif may control the destabilization of MyoD.Altogether, these experiments thus demonstrated that MyoD is a target of the SCF MAFbx complex.

DISCUSSION
By using a modified version of the yeast two-hybrid screen to identify F-box proteins that are able to interact with MyoD, we identified MAFbx, a muscle-specific ubiquitin ligase that dramatically increases during skeletal muscle atrophy.We have provided direct biological and biochemical evidence that MAFbx interacts with MyoD and thereby induces its ubiquitindependent proteolysis.MAFbx contains a LCD in which the central part is an inverted LXXLL motif that mediates direct binding to a LXXLL motif found in all MyoD species.The ␣-helical LXXLL is described as a signature motif, which mediates the recruitment of coactivators by the nuclear hormone receptors (34,35) well as ligand-independent recruitment of steroid receptor coactivators to estrogen receptor by cyclin D1 (36).Recently the cardiac-enriched nuclear receptor ERR␣ has been shown to be coactivated by the transcriptional coactivator peroxisome proliferator-activated receptor-␥ coactivator-1␣ (PGC-1␣).The PGC-1␣/ERR␣ binding interface revealed that among the three potential LXXLL motifs (L1-L3), only L3, which is an inverted motif, plays a major role in binding the nuclear receptor and represents a novel LXXLL-type of nuclear receptor binding motif on PGC-1␣ (37).The LXXLL motif in MyoD (EGLQALLR) is not observed in the E proteins basic helix-loop-helix factors and E12 did not coimmunoprecipitate with MAFbx (data not shown).Myf-5 did not coimmunoprecipitate with MAFbx, although it contains the motif (ESLQELLR) suggesting that variant residues in LXXLL core motifs influence the affinity and selectivity of MAFbx for the substrates (38).MAFbx contains other three domains known to be implicated for protein-protein interaction, the PDZ-binding motif, cytochrome c family heme-binding site, and the leucine zipper domain (Fig. 1A).It seems likely that these domains are dispensable for binding of MyoD and could mediate the interaction with additional substrates.MAFbx is the first example of an E3 ligase that involves a conserved LCD domain containing a LXXLL inverted motif for substrate recognition.
MyoD is degraded in the nucleus by the ubiquitin-proteasome system implicating at least two distinct pathways, an N terminus-dependent pathway (39) and a lysine-dependent pathway (40).In vivo studies also implicate a link between the phosphorylation of MyoD Ser-200 and the ubiquitin-proteasome degradation pathway in myoblasts (17,19,20).Altogether these data suggest that MyoD may be targeted by different E3 ligases.To avoid confusion between the N terminusdependent pathway and the lysine-dependent pathway, we introduced modifications to the N terminus of MyoD protein (the first three amino acids have been replaced by three HA epitopes), and we have recently shown that in MyoD, lysine 133 is targeted for ubiquitination and rapid degradation in the lysine-dependent pathway and plays an integral role in compromising MyoD activity in the nucleus (26).We showed that the SCF MAFbx recombinant complex and the SCF MAFbx from the skeletal muscles induced ubiquitin-dependent proteolysis of HA-tagged MyoD, leading to the conclusion that MAFbx is probably not implicated in the N terminus-dependent ubiquitination of MyoD.Indeed we observed that the mutation K133R in MyoD that dramatically reduced its ubiquitination by SCF-MAFbx strengthened the implication of this internal lysine in MyoD ubiquitination.Altogether our data show that MyoD is ubiquitinated by the SCF MAFbx via the internal lysines in which lysine 133 may be the major ubiquitination site.On the other hand, the ⌬F-box mutant binds to and represses the ubiquitination and degradation of MyoD probably by blocking and/or masking the binding sites for other ligase(s).It appears that the ubiquitin-proteasome pathway via phosphorylationdependent and -independent mechanisms regulates the turnover of MyoD suggesting that MyoD ubiquitination on the N terminus and/or on lysine 133 could be mediated by different ubiquitin ligases during myogenesis.
The expression of MAFbx in normal skeletal muscle tissues (27,28) and its up-regulation during the course of C2C12 differentiation suggest that MAFbx is not involved in regulating typical cell cycle progress.MAFbx could help maintain myotubes in a postmitotic state and/or regulate proteins in pathways that typically couple signal transduction to cell cycle control in proliferating cells.MyoD is known to control the exit from the cell cycle as well as muscle-specific gene expression (41).It would be critical for myofibers that maintain MyoD and signaling pathways that control proliferation in undifferentiated myoblasts to have mechanisms protecting against attempts to enter the cell cycle (42).In healthy muscles, there is a continual process of muscle protein production and breakdown.Skeletal muscle contains not only differentiated cells but also undifferentiated quiescent cells (satellite cells) that retain differentiation potential and contribute not only to the myotubes but also to the satellite pools (43).In the quiescent state, satellite cells are negative not only for differentiation markers but also for MyoD, and in the activated state they proliferate with high levels of MyoD protein in their nuclei.MyoD expression is induced from satellite cells and is required for these cells to proliferate and to reinitiate skeletal muscle differentiation necessary for the repair process (14).MyoDϪ/Ϫ satellite cells are differentiation-defective, highlighting the dependence of MyoD expression in the muscle fiber regeneration (44).Accelerated proteolysis via the ubiquitin-proteasome pathway is a major cause of muscle atrophy induced by denervation, disuse, and various pathological conditions (45).In atrophy, the muscle-specific ubiquitin-ligase MAFbx dramatically increases, and protein breakdown occurs more rapidly than protein production, leading to loss of muscle weight.The subsequent effect on MyoD turnover in myogenesis and regeneration after injuries suggests that up-regulation of MAFbx is likely to affect skeletal muscle regulation and regeneration.MAFbx is strongly regulated at an early stage of muscle wasting, even before muscle weight loss is detectable, and its expression is maintained during muscle atrophy.This suggests a role of these proteins in both initiation and maintenance of the accelerated proteolysis (27,28).Our data strongly suggested that overexpression of MAFbx would be required to suppress all MyoD functions inhibiting the formation of new myofibers.Thus, direct inhibition of MAFbx expression may prove beneficial in reducing the muscle wasting associated with cachexia in cancer patients as well as with other disorders.

FIG. 1 .
FIG. 1. Characterization of human MAFbx protein.A, schematic representation of the putative domains present in human skeletal muscle cDNA clones R5 and R8 encoding MAFbx.B, reverse transcription-PCR analysis of MAFbx mRNA expression in proliferating myoblasts and during the course of differentiation of C2C12 cells (24, 48, and 72 h after switching into differentiation medium).C, Western blot analysis of MAFbx and MyoD proteins during C2C12 differentiation.Asterisk marks a nonspecific band.LZ, leucine zipper; GAPDH, glyceraldehyde-3-phosphate dehydrogenase, NLS, nuclear localization signal.

FIG. 3 .
FIG. 3. MAFbx increases MyoD turnover.A, 10T1/2 cells were cotransfected using the expression vector encoding HA-MyoD and various FLAG-tagged F-box proteins as indicated.The cells were treated with cycloheximide (CHX) 36 h after transfection to block protein synthesis.Cell lysates were prepared at the indicated time and analyzed by Western blotting with anti-HA and anti-FLAG, respectively (left panel).Quantitation of MyoD turnover following cycloheximide treatment based on densitometric scanning of experiments is shown (right panel).Note that the ⌬F-box mutant stabilized MyoD.B, effects of mutations of LXXLL motif in MAFbx on MyoD degradation.MyoD turnover was analyzed in 10T/1/2 cells transfected with the indicated expression vectors followed by pulse chase to measure MyoD stability in the presence of MAFbx wild type and/or the mutant L169Q.Quantitation of MyoD turnover following pulse chase based on densitometric scanning is shown (right panel).C, pulse-chase analysis was used to measure MyoD stability in wild type and mutant MyoD L164Q in the absence or in the presence of MAFbx as in B.
FIG. 4. Overexpression of MAFbx degrades MyoD and suppresses myogenic differentiation.A, effect of MAFbx on MyoD-dependent transcriptional transactivation of muscle-specific genes.10T1/2 cells were cotransfected with 0.5 g of MCK-Luc reporter plasmid together with 2 g of pEMSV-MyoD in combination with 0.5 and 2 g of expression vector encoding FLAG-tagged MAFbx.Expression vector without insert was included to normalize DNA in all transfections.Luciferase levels were determined 48 h after transfections in proliferating medium and represent the average of three independent experiments (errors bars, S.D.).Protein concentrations were equalized by Bradford assay, and aliquots were analyzed by Western blotting for MyoD and MAFbx expression using specific antibodies (upper panel).B, overexpression of MAFbx suppressed MyoD in C2C12 myoblasts (a-o).Myoblasts were transfected with empty pcDNA4-IRES-GFP vector (a-e) and either MAFbx-IRES-GFP (f-j) or the mutant MAFbx ⌬F-IRES-GFP (k-o).Forty-eight hours later, cells were fixed and stained with anti-MyoD and revealed with goat anti-mouse antibody conjugated to Texas Red (d, i, and n).Hoechst 33258 dye shows nuclei (b, g, and i).Merged images are shown in e, j, and o.Images of a representative field were obtained by indirect immunofluorescence microscopy (400ϫ).C, overexpression of MAFbx suppresses myogenic differentiation.C2C12 myoblasts were transfected with pcDNA4-IRES-GFP expression plasmid (a-e), the mutant MAFbx ⌬F-IRES-GFP (f-j), or MAFbx-IRES-GFP (k-o).Twenty-four hours later, cells were switched to differentiation medium (DM) for 72 h, fixed, and stained with anti-MyoD as in (B) and with Hoechst 33258 dye (b, g, and l).Merged images are shown in e, j, and o.Images of a representative field were obtained by indirect immunofluorescence microscopy (400ϫ).DAPI, 4,6-diamidino-2-phenylindole.

FIG. 5 .
FIG. 5. Accumulation of MAFbx during atrophy induced by aging and fasting promotes ubiquitin conjugation of MyoD in vitro.A, total muscle lysates were prepared from hind limb skeletal muscle of control (N, lanes 1 and 3) and 3 days food-deprived mice (Atro, lanes 2 and 4) and immunoprecipitated (IP) with anti-MAFbx 61-74 antibodies.Immunoprecipitates were then analyzed by immunoblot (IB) with anti-MAFbx 326 -339 antibodies.B, total muscle lysates from hind limb skeletal muscle of control (N, lanes 1 and 2) and 3 days food-deprived mice (Atro, lane 3) were immunoprecipitated with anti-MAFbx 61-74 antibodies.Immunoprecipitates were then analyzed by Western blot with anti-MAFbx 326 -339, anti-Skp1, and anti-Cul1 antibodies, respectively.C,35 S-labeled HAtagged MyoD-wt subjected to the in vitro ubiquitination reaction in the presence of normal and/or atrophic mouse skeletal muscle lysates.

FIG. 7 .
FIG. 7. MAFbx increases MyoD ubiquitination in vivo.A, 10T1/2 cells were transiently transfected with expression plasmids encoding HA-ubiquitin and MyoD in the absence (lane 2) or in the presence of the mutant ⌬F-box (lane 3) or MAFbx (lane 4).Transfected cells were treated for 2 h with 50 M MG132 before harvesting.Cells were lysed in denaturation buffer containing 1% SDS.Aliquots (10%) were analyzed by Western blotting with anti-HA antibodies (upper panel).Cell lysates were then diluted in immunoprecipitation buffer, subjected to immunoprecipitation with anti-MyoD, and analyzed by Western blotting with anti-HA antibodies (lower panel).After exposure blots were stripped and reprobed with anti-FLAG and anti-MyoD.B, MAFbx but not the mutant MAFbx L169Q increased polyubiquitination of endogenous MyoD in C2C12 myoblasts.C, mutation of the LXXLL interaction motif impaired MyoD ubiquitination.