Advertisement

The Basic Helix-Loop-Helix Transcription Factors Myogenin and Id2 Mediate Specific Induction of Caveolin-3 Gene Expression during Embryonic Development*

Open AccessPublished:August 25, 2000DOI:https://doi.org/10.1074/jbc.M001430200
      Caveolin-3 protein is the only member of the caveolin family that shows a unique muscle-specific expression pattern, and loss of its functional activity causes muscular dystrophy. Caveolin-3 mRNA levels are dramatically increased during the formation of myotubes in the C2C12 cell line. In this study, we characterized the human caveolin-3 5′-flanking region. Promoter analyses demonstrate that the proximal E box element serves as a myogenin binding site and is both necessary and sufficient to control caveolin-3 gene transcription. Transient transfection assays indicated that overexpression of myogenin activates caveolin-3 reporter gene expression, whereas Id2 overexpression inhibited caveolin-3 promoter activation by myogenin. A mutant Id2 protein lacking the HLH domain was not capable of suppressing myogenin-mediated activation. Determination of caveolin-3 transcript distribution patterns in vivo revealed that mRNA was first detectable at day 10 of gestation in the developing somites and heart. Caveolin-3 protein in myoblasts and myotubes was expressed in both the plasma membrane and vesicular structures. During skeletal myogenesis the level of Id2, an inhibitor of differentiation, decreases, allowing the induced basic helix-loop-helix transcription factor myogenin to form transcriptionally active heterodimers that bind to the caveolin-3 promoter and thereby mediate its transcription.
      MRF
      muscle regulatory factor
      HLH
      helix-loop-helix
      bHLH
      basic HLH
      DMEM
      Dulbecco's modified Eagle's minimum essential medium
      PBS
      phosphate-buffered saline
      PCR
      polymerase chain reaction
      bp
      base pair(s)
      En
      embryonic day n
      In higher eukaryotes the localized assembly of signaling complexes in discrete membrane areas plays a pivotal role in mediating the specificity of signal transduction. Lateral assemblies of lipids in the plasma membrane, referred to as “rafts,” represent such subdomains, which can be isolated as detergent insoluble glycolipid-enriched domains (DIGs) by sucrose gradient centrifugation (
      • Simons K.
      • Ikonen E.
      ,
      • Parton R.G.
      • Simons K.
      ). Several signaling molecules including Gα subunits are associated with the DIG fraction. The plasma membrane pits, caveolae, represent a specialized DIG domain characterized by the expression of caveolin-1 (
      • Rothberg K.
      • Heuser J.E.
      • Donzell W.C.
      • Ying Y.-S.
      • Glenney J.R.
      • Anderson R.G.W.
      ). Caveolae consist of small cholesterol and glycosphingolipid-enriched invaginations of the plasma membrane, which were shown to contain a number of molecules known to participate in cell signaling. These include Src-like kinases, Ha-ras heterotrimeric G-proteins, and epidermal growth factor receptors (
      • Okamoto T.
      • Schlegel A.
      • Scherer P.E.
      • Lisanti M.P.
      ,
      • Monier S.
      • Parton R.G.
      • Vogel F.
      • Henske A.
      • Kurzchalia T.
      ,
      • Li S.
      • Okamoto M.
      • Chun M.
      • Sargiacomo J.E.
      • Casanova J.E.
      • Hansen S.H.
      • Nishimoto I.
      • Lisanti M.P.
      ).
      The caveolin family comprises three members, caveolin-1, caveolin-2, and caveolin-3. Caveolin-1 and caveolin-2 show an almost identical and ubiquitous expression pattern, whereas caveolin-3 expression is restricted to striated muscle tissue (
      • Scherer P.E.
      • Okamoto T.
      • Chun M.
      • Lodish H.F.
      • Lisanti M.P.
      ,
      • Biederer C.
      • Ries S.
      • Drobnik W.
      • Schmitz G.
      ). During early muscle development, both caveolae-like structures and caveolin-3 are associated with the formation of T-tubules, suggesting involvement of caveolin-3 in the biogenesis of the T-tubule system during muscle differentiation (
      • Parton R.G.
      • Way M.
      • Zorzi N.
      • Stang E.
      ).
      The specific functions of caveolin-3 are not yet fully understood. Immunoprecipitation studies have revealed an association of caveolin-3 with the dystrophin-glycoprotein complex. This complex serves as a link between the extracellular matrix of striated muscle, the membrane, and actin located in the cytoplasm. Perturbation of this complex can lead to muscular dystrophy (
      • Campbell K.P.
      ) and consistently screening of muscular dystrophy patients resulted in the discovery of mutations in the caveolin-3 gene (
      • Minetti C.
      • Sotgia F.
      • Bruno C.
      • Scartezzini P.
      • Broda P.
      • Bado M.
      • Masetti E.
      • Mazzocco M.
      • Egeo A.
      • Donati M.A.
      • Volonte D.
      • Galbiati F.
      • Cordone G.
      • Bricarelli F.D.
      • Lisanti M.P.
      • Zara F.
      ).
      Expression of caveolin-3 is induced in C2C12 cells upon muscle differentiation (
      • Tang Z.
      • Scherer P.E.
      • Okamoto T.
      • Song K.
      • Chu C.
      • Kohtz D.S.
      • Nishimoto I.
      • Lodish H.F.
      • Lisanti M.P.
      ). In response to serum reduction the C2C12 myoblasts withdraw from the cell cycle and fuse to form multinuclear myotubes (
      • Yun K.
      • Wold B.
      ). The myogenic basic helix-loop-helix transcription factors MyoD, myogenin, Myf5, and MRF4, known as muscle regulatory factors (MRFs),1 act at multiple stages of myogenic differentiation to establish myoblast identity and to control terminal differentiation of the muscle. MRFs also induce transcription of the myocyte enhancer-binding factor-2 family of MADS box transcription factors and thereby maintain high levels of muscle-specific gene expression (
      • Edmondson D.G.
      • Olson E.N.
      ). Mice harboring a targeted mutation in the myogenin gene are immobile and die perinatally because of deficits in skeletal muscle differentiation, although normal numbers of myoblasts are present. These results indicate an essential role of myogenin in terminal differentiation of myoblasts into myotubes (
      • Rawls A.
      • Morris J.H.
      • Rudnicki M.
      • Braun T.
      • Arnold H.H.
      • Klein W.H.
      • Olson E.N.
      ).
      bHLH transcription factors such as myogenin form heterodimers with the ubiquitously expressed E47/E12 proteins and bind to the consensus E box sequence CANNTG to mediate muscle-specific gene transcription (
      • Murre C.
      • McCaw P.S.
      • Vassin H.
      • Caudy M.
      • Jan L.Y.
      • Jan C.V.
      • Cabrera C.V.
      • Buskin J.N.
      • Hauschka S.D.
      • Lassar A.D.
      • Weintraub H.
      • Baltimore D.
      ). Growth factors inhibit myoblast differentiation by activation of protein kinase C resulting in phosphorylation of the DNA-binding site of bHLH transcription factors (
      • Li L.
      • Zhou J.
      • James G.
      • Heller-Harrison R.
      • Czech M.
      • Olson E.N.
      ). Furthermore, the inhibitor of DNA-binding Id2, a HLH protein that lacks a basic DNA-binding domain, forms heterodimers with E proteins, inhibits bHLH-mediated transcriptional activation by sequestering their dimerization partners, and thereby blocks differentiation (
      • Jen Y.
      • Weintraub H.
      • Benezra R.
      ).
      To identify and characterize important regulatory regions responsible for the tissue-specific expression pattern of caveolin-3, we cloned and characterized the human caveolin-3 promoter. In this study we demonstrate that a single E box, which is located adjacent to the 5′end of the TATA box, is required for promoter activation during muscle cell differentiation and examine the effect of myogenin and Id2 on its activity.

      RESULTS

      To investigate muscle-specific expression of caveolin-3, the C2C12 myoblast/myotube cell line was chosen as a model system. Caveolin-3 mRNA and protein levels have been reported to increase significantly during differentiation of C2C12 myoblasts (
      • Tang Z.
      • Scherer P.E.
      • Okamoto T.
      • Song K.
      • Chu C.
      • Kohtz D.S.
      • Nishimoto I.
      • Lodish H.F.
      • Lisanti M.P.
      ). Differentiation in this system is induced by serum reduction from 10% fetal bovine serum to 2% horse serum. Using a ribonuclease protection assay, we confirmed the presence of enhanced caveolin-3 mRNA 48 h after serum withdrawal (Fig.1 A). Induction of caveolin-3 protein levels visualized by immunoblots after serum withdrawal paralleled the levels of induced caveolin-3 mRNA (Fig.1 B). We also examined the subcellular localization of caveolin-3 in C2C12 cells by immunofluorescence microscopy. Caveolin-3 was only detectable in vesicular structures, concentrated at the plasma membrane of the C2C12 myotubes (Fig. 1 C). The same localization pattern was observed with a caveolin-3 GFP construct in HeLa cells (data not shown). To further confirm the localization of caveolin-3 in the plasma membrane, cell lysates of mouse skeletal muscle tissue were separated into a membrane and a cytosolic fraction. Western blot analyses revealed caveolin-3 exclusively in the membrane fraction (Fig. 1 D).
      Figure thumbnail gr1
      Figure 1Induction of caveolin-3 mRNA and protein during differentiation of C2C12 myoblasts. A, ribonuclease protection assay of caveolin-3 mRNA levels in myoblasts (lane 3) and 56 and 72 h differentiated C2C12 myocytes (lane 4 and 5, respectively). Lane 1 shows the caveolin-3 probe without RNA, and lane 2displays the t-RNA control. A GAPDH cRNA probe was included as internal loading control in each reaction (bottom panel).B, caveolin-3 protein levels analyzed by Western blot analysis after serum reduction. β-Actin expression served as loading control. C, immunofluorescence analysis of caveolin-3 expression in myoblasts and myotubes 48 h after serum withdrawal. Caveolin-3 antibody incubation was followed by incubation with a fluorescein isothiocyanate-conjugated secondary anti-mouse antibody.Arrows mark vesicles and the plasma membrane. D, Western blot of mouse skeletal muscle tissue separated into membrane (mem) and cyto-plasmatic (cyto) fractions immunoprobed with the caveolin-3 antibody.
      To study in detail the expression pattern of caveolin-3 mRNA during embryonic development, we performed in situ hybridizations to paraffin-embedded murine embryo specimens. To avoid signals resulting from cross hybridization with other members of the highly homologous caveolin family, a specific cRNA probe was prepared from the 3′-untranslated region of the mouse caveolin-3 mRNA. In mouse embryos dissected at stages E10, E11, E12, and E15 (Fig.2) caveolin-3 expression was detected in skeletal and heart muscle. The earliest signal was found in the myocardial wall of the common ventricular heart chamber (Fig.2 A, arrowhead). Fig. 2 B displays a sagittal section of a mouse embryo of gestational stage E11. Here, caveolin-3 mRNA is detectable in the heart, most prominently in the wall of the bulbus cordis (asterisk). Also faint signals are visible in the right atrial chamber. The somites at this time clearly show caveolin-3 expression (arrowheads). During the next days of development (Fig. 2, C and D) caveolin-3 was expressed with increasing intensity in all recognizable skeletal muscles, very prominently in the tongue, and even in the small muscles of the facial whisker pads. The entire myocardium at day 15 showed strong signals (Fig. 2 D). Interestingly, no caveolin-3 expression was detected in tissues containing smooth muscle such as the urinary bladder, stomach and intestinal wall (Fig. 2 D,asterisk). In summary, results of in situhybridizations revealed that in vivo caveolin-3 expression is restricted to striated muscle, both in cardiac and skeletal muscle tissues, and is induced late during differentiation.
      Figure thumbnail gr2
      Figure 2In situ hybridizations of murine embryo tissue sections with a caveolin-3-specific probe. A, transverse section of stage E10. Weak signals are visible in the myocardial wall of the ventricular heart chamber (arrowhead). B–D, sagittal section of stages E11, E12, and E15. Note signals in all later stages of heart development and in somites (arrowheads in B). Very strong expression is observed in tongue starting at E12 and in all other skeletal muscles but not in smooth muscles (asteriskin D). ht, heart; sm, skeletal muscle;tg, tongue.
      To investigate the transcriptional regulation of caveolin-3 expression, we cloned and sequenced the 5′-flanking region of the human caveolin-3 gene (Fig. 3 A). Previous 5′ rapid amplification of cDNA ends-PCR analysis indicated a single transcription start for the caveolin-3 gene (
      • Biederer C.
      • Ries S.
      • Drobnik W.
      • Schmitz G.
      ). We determined the exact position of the caveolin-3 start site by RNase protection assay using an antisense RNA probe derived from genomic DNA spanning almost 200 bp upstream of the translational initiation codon and corresponding to −197 to +77 bp. This probe was hybridized to 10 μg of human skeletal muscle RNA or tRNA as negative control and RNase-treated to cleave single-stranded RNA. A sequencing reaction was loaded on the same gel to allow the accurate determination of the product size (Fig.3 B). One strong signal was detectable confirming one major transcription start 81 nucleic acid residues upstream of the translational start codon. As displayed in Fig. 3 A, a consensus TATA box was identified 34 bp 5′ of the transcriptional start site. In a computer homology search for other cis-acting regulatory motifs that have been implicated in skeletal muscle gene regulation, we identified four consensus E box elements that represent putative binding sites for basic helix-loop-helix transcription factors (Fig. 3 A, underlined). To investigate which of these motifs are required for activation of the caveolin-3 promoter in response to skeletal muscle cell differentiation, we cloned a series of human caveolin-3 promoter deletion constructs into the promoterless luciferase reporter plasmid pGL3 Basic (Promega). These reporter constructs were transiently transfected into C2C12 cells, and luciferase activity was measured 48 h after transfection (Fig.4, bars a–f). Cotransfection of a thymidine kinase promoter driven Renilla luciferase reporter construct (Promega) was used to normalize for transfection efficiency in each experiment. As shown in Fig. 4, all promoter constructs L1 to L5, but not L6, supported high levels of reporter gene expression in C2C12 cells following serum withdrawal for 48 h. These results suggested that L6 lacks a cis-regulatory element mediating muscle-specific caveolin-3 expression. Because the most 3′ E box at position −64 is deleted in the L6 construct, we further investigated a specific function for this element followingin vitro mutagenesis. Consistently, deletion of the E boxes at positions −793, −585, and −5417 showed no effect on the transcriptional activity of the promoter reporter constructs (data not shown), but deletion of the E box 1 in the L5 promoter construct resulted in significant decrease of promoter activity (Fig. 4,bar g). To determine whether one of the distal E boxes can substitute for the loss of E box 1, we generated a full-length promoter construct harboring a specific deletion of E box 1 (L1mut). As shown in Fig. 4 (bar h) deletion of E box 1 led to almost complete inactivation of caveolin-3 promoter function in C2C12 cells. These experiments highlight the importance of E box 1 (CAGCTG) located at position −64 adjacent to the TATA box.
      Figure thumbnail gr3
      Figure 3Promoter sequence and mRNA start site of the human caveolin-3 gene. A, the caveolin-3 promoter sequence was cloned using the genome walker kit libraries (CLONTECH). Four E box motifs are marked inbold type and underlined. The TATA box is displayed in bold type, the cytidine at the mRNA start site is printed in italic bold type, and the ATG start codon is in bold type and double-underlined. B, the transcriptional start site (TS) was mapped performing an RNase protection assay using a cRNA probe derived from a genomic fragment spanning from 197 bp upstream of the initiation start codon to +77 bp downstream. Lane 1 shows the size of the protected fragment, and lane 2 shows the cRNA probe alone. The sequencing reaction is displayed at the right as a separate part because of different exposure times.
      Figure thumbnail gr4
      Figure 4Deletion analysis of caveolin-3 promoter fragments controlling luciferase reporter plasmids in C2C12 and NIH3T3 cells. Deletion constructs L1–L6 are depicted schematically underneath. Luciferase activity measured in C2C12 myoblasts (black bars) and 48 h differentiated myocytes (gray bars) is shown above. Thymidine kinase drivenRenilla reporter served as control for transfection efficiency. The E-box 1-mutated promoter constructs and the corresponding luciferase activities are shown in bars g andh, and the full-length promoter construct in reverse orientation in bar i. The L5 and L6 promoter constructs were transfected into NIH3T3 cells cultivated in 2% (black bars) or 10% serum (gray bars; bars j andk).
      A luciferase construct containing the entire 5′-flanking region in the reverse orientation revealed no transcriptional activity (Fig. 4,bar i). To ensure that the transcriptional induction of the caveolin-3 promoter activity is due to the differentiation process and not to serum withdrawal, the L5 and L6 constructs were transfected into NIH 3T3 fibroblasts and cultured in medium supplemented with 2% or 10% fetal calf serum. As shown in Fig. 4 (bars j andk), both reporter plasmids were inactive under either condition. In summary, we concluded from these results that the E box 1 harbors an important cis-regulatory element that controls muscle-specific transcription of the caveolin-3 gene in differentiated C2C12 cells.
      Because E box elements are known to serve as binding sites for myogenic bHLH transcription factors, we examined the influence of the muscle regulatory factor myogenin, which is is induced in C2C12 upon serum withdrawal and plays an essential role in myoblast terminal differentiation (
      • Rawls A.
      • Morris J.H.
      • Rudnicki M.
      • Braun T.
      • Arnold H.H.
      • Klein W.H.
      • Olson E.N.
      ), on caveolin-3 promoter activity. The luciferase reporter construct L1, which includes the entire 5′-flanking region was cotransfected with a myogenin expressing plasmid. Myogenin expression was verified by Western blot analysis (data not shown). As displayed in Fig. 5 overexpression of myogenin dramatically increased the caveolin-3 promoter activity in undifferentiated C2C12 myoblasts. In the differentiated C2C12 myocytes, myogenin was sufficient to induce caveolin-3 expression.
      Figure thumbnail gr5
      Figure 5Cotransfection of myogenin, Id2, and the caveolin-3 promoter construct L1 into C2C12 and NIH3T3 cells. 1 μg of L1 reporter plasmid was transiently transfected with 100 ng of each expression plasmid as indicated below the bars. Id2 dHLH represents an HLH-domain mutant Id2 that fails to interact with HLH proteins. Immediately after transfection one fraction of C2C12 cells was cultured in 2% horse serum containing medium (gray bars), whereas the other fraction was continuously maintained in 10% fetal bovine serum medium (black bars). Luciferase activity was determined 48 h after transfection. Expression of myogenin, Id2, and the mutant Id2 was verified by Western blot analysis (data not shown).
      Overexpression of myogenin in NIH3T3 fibroblasts causes a myocyte phenotype (
      • Davis R.L.
      • Weintraub H.
      • Lassar A.B.
      ). Therefore, we also investigated the consequence of myogenin overexpression in 3T3 cells on the L1 caveolin-3 promoter activity. Consistently, myogenin transfection into NIH3T3 cells resulted in induction of caveolin-3 promoter activation (Fig. 5). Serum starvation of myogenin-transfected NIH3T3 cells (0.1% serum) resulted in an even enhanced transactivation of the caveolin-3 promoter (data not shown). This effect may be explained by reduced levels of Id2 expression levels in starved NIH3T3 cells.
      Id genes function as inhibitors of differentiation by forming inactive heterodimers with basic HLH proteins and thereby blocking bHLH-mediated transactivation of differentiation-specific gene expression (
      • Jen Y.
      • Weintraub H.
      • Benezra R.
      ). Cotransfection of the L1 construct with an Id2 expression plasmid (
      • Florio M.
      • Hernandez M.C.
      • Yang H.
      • Shu H.K.
      • Cleveland J.L.
      • Israel M.A.
      ) in C2C12 cells revealed suppression of L1 reporter gene transcription, following serum withdrawal. Cotransfection of myogenin expression plasmid also failed to induce caveolin-3 promoter activity in the presence of Id2 (Fig. 5). A mutant Id2, which lacks the HLH domain and is defective in heterodimerization with bHLH factors, was not capable of suppressing myogenin-induced activation of the caveolin-3 promoter or activation ocurring as a result of serum withdrawal (Fig. 5). These results demonstrate that activation of the human caveolin-3 promoter occurs in the presence of the myogenic bHLH transcription factor myogenin and can be blocked by HLH interactions with inhibitory dimerization partners such as Id2.
      To analyze DNA-protein complexes interacting with E box 1, synthetic oligonucleotides forming a double-stranded E box 1-binding site were annealed and used for electromobility shift assays (Fig.6). In addition, a second oligomeric-binding site was synthesized containing point mutations in the E box 1 consensus sequence as indicated under “Experimental Procedures.” The wild type E box 1 oligomere formed a single band shift that was present exclusively in C2C12 myotube but not in myoblast nuclear extracts (Fig. 6, lanes 2–4). A significant portion of this band shift was further retarded upon addition of myogenin antibody (Fig. 6, lane 7) but not upon addition of an unrelated c-Fos antibody that served as a negative control (Fig.6, lane 8). Further competition experiments with 50-fold molar excess of unlabeled wild type and mutated oligomeric-binding sites were performed to control the specificity of the band shift experiments. Incubation of the binding reaction with unlabeled wild type binding site completely inhibited detection of the retarded complex (Fig. 6, lane 5), whereas a 50-fold molar excess of mutated binding site had no effect on the the band shift (Fig. 6,lane 6). Taken together these data define thecis-regulatory element E box 1 located at position −64 in the caveolin-3 promoter as a myogenin-binding site.
      Figure thumbnail gr6
      Figure 6Gel shift assays with C2C12 nuclear extracts and the E box 1-binding site of the caveolin-3 promoter. Lane 1 shows the binding site without addition of nuclear extracts. Differentiation of C2C12 myoblasts (lane 2) was initiated by culturing the cells in medium containing 2% horse serum for 24 h (lane 3) or 48 h (lanes 4–8).Lanes 5 and 6 show competition with 50-fold excess cold wild type (lane 5) and mutant (lane 6) oligonucleotides. 2 μl of myogenin antibody (DAKO, lane 7) or c-fos antibody (Geneka, lane 8) were added to the reaction containing 48 h differentiated C2C12 nuclear extracts. RC, retarded complex; SS, supershift.
      Finally, we determined by Western blot analysis the expression of MoyD, myogenin, and Id2 in undifferentiated C2C12 myoblasts and after serum withdrawal. As shown in Fig. 7 (top panel) MyoD was expressed in both undifferentiated and differentiated C2C12 cells, whereas myogenin expression was specifically detected in differentiated C2C12 cells 48 h after serum withdrawal. In parallel the Id2 level decreased during the muscle differentiation process (bottom panel), strongly suggesting a physiological role as an inhibitor of differentiation.
      Figure thumbnail gr7
      Figure 7Expression of transcriptional regulators of the myogenic differentiation program in C2C12 cells. Western blot analysis of whole cell extracts from undifferentiated or 24 h or 48 h differentiated C2C12 cells. 20 μg of protein lysates were immunoprobed with affinity-purified antibodies against MyoD (mouse monoclonal, DAKO), myogenin (mouse monoclonal, Pharmingen), or Id2 (rabbit polyclonal, Santa Cruz). Equal loading of the lanes was confirmed by reprobing with an antibody against β-actin (mouse monoclonal; Sigma).
      To further determine whether Id2 functions as a negative regulator of caveolin-3 promoter activity in vivo, we compared expression patterns of Id2 and caveolin-3 in the dermatomyotome of developing somites and differentiated skeletal muscle tissue by in situhybridization to murine embryos. As shown in Fig.8, we observed very strong Id2 signals in undifferentiated muscle precursor cells at gestational age E10 but not in differentiated skeletal muscle at gestational age E16. Vice versa, caveolin-3 expression was absent in Id2-positive myoblasts but very high in Id2-negative differentiated muscle tissues. These results further point to myogenin as an important regulator of caveolin-3 expression because induction of myogenin and concomitant decrease in the level of the myogenic inhibitor Id2 parallel the pattern of caveolin-3 expression during differentiation of myoblasts.
      Figure thumbnail gr8
      Figure 8Reciprocal expression patterns of Id2 and caveolin-3 in myoblasts and differentiated myocytes. In situ hybridizations with Id2 (A and C) and caveolin-3-specific probes (B and D) to horizontal sections of mouse embryo sections at day E10 (Aand B) and sagitto-lateral sections through the neck and shoulder girdle of E16 embryos (C and D).nt, neural tube; my, myotome; d, dermis; pn, peripheral nerve; sm, skeletal muscle.

      DISCUSSION

      We have previously isolated and determined the entire cDNA sequence of the human caveolin-3 gene (
      • Biederer C.
      • Ries S.
      • Drobnik W.
      • Schmitz G.
      ). In this report we cloned and characterized the human caveolin-3 promoter and identified an importantcis-regulatory E box element controling caveolin-3 gene transcription during skeletal muscle differentiation. Our results provide conclusive evidence that the formation of an active transcription complex on the caveolin-3 promoter in mouse C2C12 myotubes is critically dependent on binding of the bHLH transcription factor myogenin to the E box 1 located adjacent to the TATA box.
      For our study we used the C2C12 myoblast/myotube system because these cells display many of the morphological and genetic characteristics of the skeletal muscle differentiation program. In these cells caveolin-3 expression parallels the in vivo expression pattern observed during embryonic differentiation. Formation of myotubes from C2C12 myoblasts induced by serum withdrawal is accompanied by transcriptional induction of muscle-specific genes (
      • Edmondson D.G.
      • Olson E.N.
      ). A complex set of muscle-specific and ubiquitous transcription factors has been identified that cooperatively controls muscle-specific transcription (
      • Weintraub H.
      • Dwarki V.J.
      • Verma I.
      • Davis R.
      • Hollenberg S.
      • Snider L.
      • Lassar A.
      • Tapscott S.J.
      ). Well characterized muscle-specific transcription factors of the MRF family, including MyoD (
      • Davis R.L.
      • Weintraub H.
      • Lassar A.B.
      ), myogenin (
      • Wright W.E.
      • Sassoon D.A.
      • Lin V.K.
      ), Myf5 (
      • Braun T.
      • Buschhausen-Denker G.
      • Bober E.
      • Tannich E.
      • Arnold H.H.
      ), and MRF4 (
      • Rhodes S.J.
      • Konieczny S.F.
      ), bind to E box motifs consensus to the sequence CANNTG (
      • Ludolph D.C.
      • Konieczny S.F.
      ) present in muscle-specific gene promoters.
      Although the 5′-flanking sequence of the caveolin-3 gene contains several E box elements, promoter deletion experiments revealed that the loss of three E-boxes did not significantly affect luciferase expression in C2C12 myotubes. However, in vitro mutagenesis of the E box 1 adjacent to the TATA box resulted in strong inhibition of caveolin-3 gene transcription in C2C12 myocytes. EMSA analysis using a myogenin antibody identified E box 1 as a specific myogenin-binding site occupied during C2C12 differentiation (Fig. 6). Consistently, Western blot analysis confirmed that myogenin expression was induced upon serum withdrawal, subsequently leading to differentiation (
      • Wright W.E.
      • Sassoon D.A.
      • Lin V.K.
      ). Structural analyses of the MyoD bHLH domain-DNA complex by x-ray crystallography has revealed limited specificity of MyoD and myogenin for the central two base pairs of the E box consensus sequence (
      • Ma P.C.
      • Rould M.A.
      • Weintraub H.
      • Pabo C.O.
      ). These two myogenic transcription factors bind preferentially to the sequence CAGCTG in contrast to other bHLH factors (
      • Biederer C.
      • Ries S.
      • Drobnik W.
      • Schmitz G.
      ). In good agreement, E box 1 serving as a myogenin-binding site matches precisely the preferred sequence. The important role of myogenin is further supported by our results from cotransfection studies using myogenin and Id2 expression plasmids, which led either to activation or to inhibition of caveolin-3 mRNA induction. These experiments clearly underline that the myogenic transcription factor myogenin and the inhibitor of differentiation Id2 direct caveolin-3 expression during skeletal muscle cell differentiation and provide further evidence that myogenin and MyoD do not exert entirely redundant functions. Recently, specific expression patterns of all three caveolin genes have been detected in adult brain tissues (
      • Ikezu T.
      • Ueda H.
      • Trapp B.D.
      • Nishiyama K.
      • Sha J.F.
      • Volonte D.
      • Galbati F.
      • Byrd A.L.
      • Bassell G.
      • Serizawa H.
      • Lane W.S.
      • Lisanti M.P.
      • Okamoto T.
      ). Therefore, it can be speculated that the distal E box elements may serve as enhancer element in neuronal or glial cells.
      The spatial and temporal patterns of caveolin-3 mRNA distribution during mouse embryonic development again points to an important role of myogenin in regulating caveolin-3 mRNA expression. The earliest detection of myogenin transcripts is at 8.5 days in all somites (
      • Wright W.E.
      • Sassoon D.A.
      • Lin V.K.
      ) compared with 11 days for caveolin-3 mRNA (Fig. 2). Thus, myogenin expression precedes caveolin-3 induction during the skeletal muscle differentiation program in vivo, and significant expression of caveolin-3 is not observed prior to down-regulation of Id2.
      Our data indicate that the bHLH transcription factor myogenin binds to the proximal E box 1 in close proximity to the TATA element. Intriguingly, the desmin promoter is regulated in a similar way. Both the distance between the E box and the TATA box and the distance from the mRNA start site have been shown to be important for transactivation of the desmin promoter by MyoD and myogenin (
      • Li H.
      • Capetanaki Y.
      ). Therefore, it is possible that direct interactions between myogenin and the basal transcription machinery, potentially mediated by TATA-binding protein-associated factors (TAFs) specifically expressed in differentiated myotubes (
      • Takeda S.
      • North D.L.
      • Diagana T.
      • Miyagoe Y.
      • Lakich M.M.
      • Whalen R.G.
      ), depend on restricted spatial requirements that are crucial for caveolin-3 gene transcription.

      Acknowledgments

      We thank Dr. Osamu Tetsu for help with the site-directed mutagenesis and Dr. Clodagh O'Shea for helpful advice with immunofluorescent microscopy. We also thank Gerd Schmitz for supporting part of the experimental work.

      REFERENCES

        • Simons K.
        • Ikonen E.
        Nature. 1997; 387: 569-572
        • Parton R.G.
        • Simons K.
        Science. 1995; 269: 1398-1399
        • Rothberg K.
        • Heuser J.E.
        • Donzell W.C.
        • Ying Y.-S.
        • Glenney J.R.
        • Anderson R.G.W.
        Cell. 1992; 68: 673-682
        • Okamoto T.
        • Schlegel A.
        • Scherer P.E.
        • Lisanti M.P.
        J. Biol. Chem. 1998; 273: 5419-5422
        • Monier S.
        • Parton R.G.
        • Vogel F.
        • Henske A.
        • Kurzchalia T.
        Mol. Biol. Cell. 1995; 6: 911-927
        • Li S.
        • Okamoto M.
        • Chun M.
        • Sargiacomo J.E.
        • Casanova J.E.
        • Hansen S.H.
        • Nishimoto I.
        • Lisanti M.P.
        J. Biol. Chem. 1995; 270: 15693-15701
        • Scherer P.E.
        • Okamoto T.
        • Chun M.
        • Lodish H.F.
        • Lisanti M.P.
        Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 131-135
        • Biederer C.
        • Ries S.
        • Drobnik W.
        • Schmitz G.
        Biochim. Biophys. Acta. 1998; 1406: 5-9
        • Parton R.G.
        • Way M.
        • Zorzi N.
        • Stang E.
        J. Cell Biol. 1997; 136: 137-154
        • Campbell K.P.
        Cell. 1995; 80: 675-679
        • Minetti C.
        • Sotgia F.
        • Bruno C.
        • Scartezzini P.
        • Broda P.
        • Bado M.
        • Masetti E.
        • Mazzocco M.
        • Egeo A.
        • Donati M.A.
        • Volonte D.
        • Galbiati F.
        • Cordone G.
        • Bricarelli F.D.
        • Lisanti M.P.
        • Zara F.
        Nat. Genet. 1998; 18: 365-368
        • Tang Z.
        • Scherer P.E.
        • Okamoto T.
        • Song K.
        • Chu C.
        • Kohtz D.S.
        • Nishimoto I.
        • Lodish H.F.
        • Lisanti M.P.
        J. Biol. Chem. 1996; 271: 2255-2261
        • Yun K.
        • Wold B.
        Curr. Opin. Cell Biol. 1996; 8: 877-889
        • Edmondson D.G.
        • Olson E.N.
        J. Biol. Chem. 1993; 268: 755-758
        • Rawls A.
        • Morris J.H.
        • Rudnicki M.
        • Braun T.
        • Arnold H.H.
        • Klein W.H.
        • Olson E.N.
        Dev. Biol. 1995; 172: 37-50
        • Murre C.
        • McCaw P.S.
        • Vassin H.
        • Caudy M.
        • Jan L.Y.
        • Jan C.V.
        • Cabrera C.V.
        • Buskin J.N.
        • Hauschka S.D.
        • Lassar A.D.
        • Weintraub H.
        • Baltimore D.
        Cell. 1989; 58: 537-544
        • Li L.
        • Zhou J.
        • James G.
        • Heller-Harrison R.
        • Czech M.
        • Olson E.N.
        Cell. 1992; 71: 1181-1194
        • Jen Y.
        • Weintraub H.
        • Benezra R.
        Genes Dev. 1992; 6: 1466-1479
        • Blau H.M.
        • Chiu C.
        • Webster C.
        Cell. 1983; 32: 1171-1180
        • Wright W.E.
        • Sassoon D.A.
        • Lin V.K.
        Cell. 1989; 56: 607-617
        • Florio M.
        • Hernandez M.C.
        • Yang H.
        • Shu H.K.
        • Cleveland J.L.
        • Israel M.A.
        Mol. Cell. Biol. 1998; 18: 5435-5444
        • Buettner R.
        • Kannan P.
        • Imhof A.
        • Bauer R.
        • Yim S.O.
        • Glockshuber R.
        • Van Dyke M.W.
        • Tainsky M.A.
        Mol. Cell. Biol. 1993; 13: 4174-4185
        • Davis R.L.
        • Weintraub H.
        • Lassar A.B.
        Cell. 1987; 51: 987-1000
        • Weintraub H.
        • Dwarki V.J.
        • Verma I.
        • Davis R.
        • Hollenberg S.
        • Snider L.
        • Lassar A.
        • Tapscott S.J.
        Genes Dev. 1991; 5: 1377-1386
        • Braun T.
        • Buschhausen-Denker G.
        • Bober E.
        • Tannich E.
        • Arnold H.H.
        EMBO J. 1989; 8: 701-709
        • Rhodes S.J.
        • Konieczny S.F.
        Genes Dev. 1989; 3: 2050-2061
        • Ludolph D.C.
        • Konieczny S.F.
        FASEB J. 1995; 9: 1595-1604
        • Ma P.C.
        • Rould M.A.
        • Weintraub H.
        • Pabo C.O.
        Cell. 1994; 77: 451-459
        • Ikezu T.
        • Ueda H.
        • Trapp B.D.
        • Nishiyama K.
        • Sha J.F.
        • Volonte D.
        • Galbati F.
        • Byrd A.L.
        • Bassell G.
        • Serizawa H.
        • Lane W.S.
        • Lisanti M.P.
        • Okamoto T.
        Brain Res. 1998; 804: 177-192
        • Li H.
        • Capetanaki Y.
        EMBO J. 1994; 13: 3580-3589
        • Takeda S.
        • North D.L.
        • Diagana T.
        • Miyagoe Y.
        • Lakich M.M.
        • Whalen R.G.
        J. Biol. Chem. 1995; 270: 15664-15670