Syndecan-1 Expression Is Down-regulated during Myoblast Terminal Differentiation

Syndecan-1 is an integral membrane proteoglycan involved in the interaction of cells with extracellular matrix proteins and growth factors. It is transiently expressed in several condensing mesenchymal tissues after epithelial induction. In this study we evaluated the expression of syndecan-1 during skeletal muscle differentiation. The expression of syndecan-1 as determined by Northern blot analyses and immunofluorescence microscopy is down-regulated during differentiation. The transcriptional activity of a syndecan-1 promoter construct is also down-regulated in differentiating muscle cells. The decrease in syndecan-1 gene expression is not dependent on the presence of E-boxes, binding sites for the MyoD family of transcription factors in the promoter region, or myogenin expression. Deletion of the region containing the E-boxes or treatment of differentiating cells with sodium butyrate, an inhibitor of myogenin expression, had no effect on syndecan-1 expression. Basic fibroblast growth factor and transforming growth factor type β, which are inhibitors of myogenesis, had little effect on syndecan-1 expression. When added together, however, they induced syndecan-1 expression. Retinoic acid, an inducer of myogenesis, inhibited syndecan-1 expression and abolished the effect of the growth factors. These results indicate that syndecan-1 expression is down-regulated during myogenesis and that growth factors and retinoic acid modulate syndecan-1 expression by a mechanism that is independent of myogenin.

Heparan sulfate is present on the cell surface of most, if not all, adherent cells in the form of heparan sulfate proteoglycans. The best characterized cell surface proteoglycans are the syndecans, a family of transmembrane heparan sulfate proteoglycans that are implicated in a variety of interactions with the pericellular microenvironment (for review, see Refs. 1 and 2). The first member of this family to be discovered, syndecan-1, is a hybrid proteoglycan bearing both heparan sulfate and chon-droitin sulfate glycosaminoglycans (3). This proteoglycan binds a variety of extracellular matrix constituents, such as fibronectin (4); thrombospondin (5); tenascin (6); and collagen types I, III, and V (7) as well as basic fibroblast growth factor (bFGF) 1 (8). Syndecan-1 has been suggested to be a morphoregulatory molecule because its expression during embryonic development is dynamically regulated and corresponds to morphological boundaries (9). Changes in syndecan-1 expression correlate with morphologic changes during interactions between epithelial and mesenchymal cells in tooth (10), kidney (11), female reproductive tract (12), and limb bud (13) development.
The molecular mechanisms that regulate syndecan expression are only beginning to be explored. Sequences upstream of the first exon of murine syndecan-1 gene have promoter activity. This region contains an array of consensus transcription factors binding sites (14). These sites, which include Antennapedia, NF-B, SP1 (GC and GT box), TATA and CAAT boxes, and five E-boxes (CANNTG), provide potential mechanisms for regulation of syndecan-1 expression.
Skeletal muscle cells are a useful model for studying cell differentiation. The fusion of mononucleated myoblasts to form multinucleated myotubes is a central event in skeletal muscle development. The onset and progression of this process are controlled by a complex set of interactions between myoblasts and their environment. Some of the regulatory proteins that control this process have been identified. Thus, when myogenesis begins, myogenic regulatory genes of the MyoD family (myogenin, mrf5, mrf4) are activated (15). These factors bind to specific DNA consensus sites called E-boxes, which function as transcriptional enhancers of muscle differentiation genes (for review, see Ref. 16). Recent data suggest that factors of the MEF2 family are also required to specify muscle fate or to direct muscle differentiation (17). The expression and activity of these master genes are regulated by polypeptide growth factors including bFGF (18), transforming growth factor ␤ (TGF-␤) (19), and insulin-like growth factor (20), as well as by retinoic acid (RA) (21). One or more of these growth factors, when present at high levels, hold myoblasts in the undifferentiated state (for review, see Ref. 22), whereas RA induces myoblast differentiation (21).
It has been demonstrated that heparan sulfate proteoglycans are necessary for the modulation of terminal myogenesis (23), probably by acting as low affinity receptor for bFGF (24,25), a potent inhibitor of myogenesis (18). We have shown that the expression of glypican (26,27) and perlecan (28) varies during skeletal muscle differentiation of the C 2 C 12 myoblast cell line. Syndecan-1 is expressed transiently during limb development (13) but is absent in adult skeletal muscle (29), making this proteoglycan a good candidate for modulation of bFGF signaling during myoblast differentiation. The presence of E-boxes in the syndecan-1 promoter has led to the suggestion that MyoD or a related protein binds to these sites to cause down-regulation of syndecan-1 expression. The decrease in syndecan-1 expression would attenuate bFGF signaling and promote differentiation of myoblasts (30).
In this study we show that expression of syndecan-1 is downregulated during skeletal muscle differentiation of C 2 C 12 myoblasts but by a myogenin-and E-box-independent pathway. Syndecan-1 expression is controlled by a proximal region of the promoter which contains putative SP1, NF-B, and RA response element binding sites and is strongly influenced by bFGF, TGF-␤, and RA.

EXPERIMENTAL PROCEDURES
Cell Culture-The mouse skeletal muscle cell line C 2 C 12 (31) was grown as described by Brandan et al. (32). Three days after plating (80% confluence), the medium was changed to differentiation medium (Dulbecco's modified Eagle's medium supplemented with 5% horse serum). Two days later, 0.1 mM cytosine-␤-D-arabinofuranoside was added to the culture medium. Thereafter the incubation medium was changed daily. NMuMG mouse mammary epithelial cells and C3H10T1/2 mouse fibroblasts (ATCC) were routinely cultured in Dulbecco's modified Eagle's medium containing 10% fetal calf serum (FCS) as described (33). For experiments, cells were plated at equal density and grow to 60 -70% confluence.
RNA Isolation and Northern Blot Analysis-Total RNA was isolated from cell cultures at the indicated times by guanidium thiocyanate/ phenol/chloroform extraction and isopropyl alcohol precipitation using RNAzol B (Cinna/Biotecx Laboratories, Inc., Houston, TX) (34). RNA samples (15 g/lane) were electrophoresed through 1.2% agarose/formaldehyde gels, transferred to Nylon membranes (Sigma), and hybridized with probes for myogenin, syndecan-1, and creatine kinase. The probes were prepared as follows. For myogenin, a conserved 672-bp fragment of human myogenin cDNA was amplified by PCR (35). For syndecan-1 and creatine kinase a fragment of 529 bp (from 1505 to 2033 in the cDNA) and 486 bp from 389 to 874 in the cDNA), respectively, were amplified by reverse transcriptase PCR from total RNA obtained from C 2 C 12 cells. Blots were hybridized with random primed labeled probes in a buffer containing 1.0 M NaCl, 1% sodium dodecyl sulfate, 10% dextran sulfate, and 100 g/ml denatured salmon testes DNA at 65°C overnight. Hybridized membranes were washed twice at 65°C in 0.2 ϫ SSC, 0.1% sodium dodecyl sulfate for 5 min and exposed to Kodak x-ray film.
Immunofluorescence Microscopy-Cells to be analyzed by immunofluorescence microscopy were grown on glass coverslips, as described previously (27). For cell surface staining the live unfixed cells were incubated in the first antibody solution (1:20 of affinity-purified antisyndecan-1 (36) in 5% non-fat milk, 0.1 M sodium chloride, 0.02 M sodium phosphate, pH 7.4) for 30 min on ice. After removing this solution the cells were fixed in 3% paraformaldehyde, and the bound antibodies were detected by incubating the cells with affinity-purified Texas Red-conjugated secondary antibodies. Nuclei were stained by incubating the cells in 0.1 g/ml DAPI, 20 mM Tris-HCl, pH 7.5, 0.1 M NaCl for 20 min. After rinsing, the slides were viewed with a Nikon upright microscope equipped for epifluorescence.
Immunoblot Analysis-Conditioned medium from approximately 6 ϫ 10 8 cells was concentrated in a Q-Sepharose column (Sigma) (2.5 ϫ 15 cm) equilibrated with 50 mM Tris-HCl, pH 7.4. After application of the conditioned medium the column was washed with the equilibration buffer until the A 280 of the effluent was 0. Bound proteins were eluted with the same buffer containing 1.0 M NaCl until phenol red was eluted. After dialysis against the equilibration buffer the sample was subjected to high performance liquid chromatography on a 7.5 ϫ 75-mm column of DEAE (Beckman Spherogel TSK DEAE-5PW) equilibrated with the same buffer. The bound proteins were eluted with a linear gradient from 0 to 1 M NaCl in the equilibration buffer at a flow rate of 1 ml/min. For immunoblot analysis aliquots were subjected to sodium dodecyl sulfate-gel electrophoresis in 7.5% polyacrylamide gels, electrophoretically transferred to Immobilon membranes (Millipore, Bedford, MA), and stained with affinity-purified rabbit anti-syndecan-1 antibodies (36) and visualized by enhanced chemiluminescence (Pierce).
Cell Transient Transfections-Plasmids were purified by Wizard Plus Maxiprep kits (Promega). The cells were plated in growth medium 1 day before transfection at a density of 8,000/cm 2 in 60-mm plates (Corning). For transfection the cells were incubated for 6 h with Opti-MEM I (Life Technologies, Inc.) containing 5 g of plasmid DNA and 20 g of LipofectAMINE (Life Technologies, Inc.). After transfection the cells were incubated for 14 h in Opti-MEM I containing 10% FCS and 0.5% chick embryo extract. The cells were then washed twice with Hanks' balanced salt solution and incubated for 2 days in growth medium, followed by 1, 2, or 3 days in differentiation medium. To analyze the effects of growth factors and RA the cells were incubated for 2 days in differentiation medium containing 10 ng/ml bFGF, 2.5 ng/ml TGF-␤, or the indicated concentrations of RA. All transfections were performed at least three times with at least two plasmid DNA preparations.
Chloramphenicol Acetyltransferase and ␤-Galactosidase Assays-At various times the cells were harvested and assayed for chloramphenicol acetyltransferase (CAT) and ␤-gal activities. CAT activity was measured as follows. Cell extracts were incubated for 14 h at 37°C in a reaction mixture containing 14 C-labeled chloramphenicol (NEN Life Science Products) and n-butyryl-CoA (Promega). The reaction products were extracted with xylene, and the organic phase was counted in a scintillation counter. For the ␤-gal assay the cell extracts were incubated for 14 h at 37°C with a buffer containing the substrate ONPG (Promega), and the absorbance was read at 420 nm with a spectrophotometer.
Construction of Expression Vectors-For the construction of the syndecan-1 reporter (p-667CAT), the published mouse promoter sequence (14) was used to design oligonucleotides that were used to amplify by nested PCR a 667-bp upstream fragment that begins 50 bp downstream from the TATA box and contains potential transcription binding sites for SP1 (GC and GT boxes), MyoD (E-boxes) and NF-B. The fragment was subcloned into the pCAT-basic vector (Promega). For the preparation of a reporter construct without E-boxes (p-244CAT) a 244-bp fragment was amplified by PCR with sense primer (CCT AGG AGG CGT AGA AGG) and antisense primer (CTG CGT TAG GCT CTG TCT CC) using p-667CAT plasmid as template.
For construction of the p-myoCAT reporter plasmid, a fragment of 688 bp corresponding to a region of the myogenin promoter which contains the MEF1 and E-boxes sites (from ϩ62 to Ϫ626 (37)) was amplified by nested PCR using a mouse BALB/c genomic library (ATCC) as template. The amplified fragment was ligated to the pCATbasic vector as described above. All the pCAT vectors were sequenced using the Sequenase kit. A ␤-gal expression plasmid (RSV-␤gal) was obtained from ATCC. This plasmid was co-transfected with the CAT constructs to monitor transfection efficiencies.

RESULTS
Myoblasts Synthesize Syndecan-1-To evaluate the expression of syndecan-1 in myoblasts, total RNA was isolated from C 2 C 12 myoblasts and from 10T1/2 fibroblasts and NMuMG epithelial cells, two cell lines known to express syndecan-1. Fig.  1A shows that C 2 C 12 myoblasts express both the major 2.5kilobase and minor 3.1-kilobase forms of syndecan-1 mRNA at levels comparable to the mouse epithelial cell line (29). Fig. 1B shows Western blot analyses of conditioned medium from myoblasts. The proteoglycans were eluted from a DEAE column at different NaCl concentrations (lane 2, 0.85 and lane 3, 1.0 M NaCl) and stained with affinity-purified anti-syndecan-1 antibodies (38). The antibodies recognized a high molecular weight smear in the conditioned medium of myoblasts, which was eluted from the DEAE column at high salt concentration, consistent with staining of a proteoglycan. For comparison, lane 1 shows an immunoblot of syndecan-1 present in conditioned medium of Schwann cells transfected with syndecan-1 cDNA (36). Together, these results indicate that myoblasts synthesize and release syndecan-1.
Syndecan-1 Expression Decreases during Myogenesis-To evaluate the expression of syndecan-1 during myoblast differentiation total RNA was isolated from myoblasts and from myoblasts induced to differentiate for 2 and 5 days and evaluated by Northern blot analysis. Fig. 2 shows a significant decrease in the amount of syndecan-1 mRNA during differentiation, which was evident by day 2 of differentiation and essentially complete by day 5. Fig. 2 also shows the increase of mRNAs coding for creatine kinase and myogenin, skeletal muscle-specific markers that increase during differentiation.
To evaluate further the expression of syndecan-1 during skeletal muscle differentiation cell cultures were stained with anti-syndecan-1 antibodies. Fig. 3 shows that myoblasts (panel A) express syndecan-1 that can be detected on the cell surface by immunofluorescent staining. Cells stained 2 days after initiation of differentiation (panel C) showed less immunoreactive syndecan-1. Essentially no staining was observed after 5 days of differentiation (myotubes) (panel E). Together, these results clearly demonstrate that the synthesis of syndecan-1 during differentiation of C 2 C 12 skeletal muscle cells is significantly diminished.
To characterize the mechanism underlying this down-regulation syndecan-1 transcriptional activity was measured in transient transfection experiments using a reporter vector consisting of 667 bp of the rat syndecan-1 promoter linked to a CAT reporter gene (p-667CAT). This promoter region contains a putative TATA box sequence and consensus binding sites for several transcriptions factors, including two E-boxes (414 and 289 bp upstream of TATA box). Fig. 4 shows the CAT activity in myoblasts transiently transfected with p-667CAT and induced to differentiate. The transcriptional activity decreases significantly in differentiating cells so that by 3 days the activity is essentially abolished. The inset in Fig. 4 shows transcriptional activity obtained with an expression construct that contains a myogenin promoter fused to a CAT reporter gene. As expected, and in contrast to syndecan-1 promoter activity, transcription driven by the myogenin promoter increases significantly after differentiation is triggered. These results demonstrate that the decrease in syndecan-1 expression which occurs during skeletal muscle differentiation results from a decreased rate of transcription.
The Decrease in Syndecan-1 Expression during Myogenesis Is Myogenin-and E-box-independent-The syndecan-1 promoter contains several E-boxes, which are DNA binding sites for MyoD and related proteins, including myogenin. Upon induc- FIG. 1. Syndecan-1 is synthesized by C 2 C 12 myoblasts. Panel A, Northern blot analysis. 15 g of total RNA isolated from myoblasts (C 2 C 12 ), fibroblasts (10T1/2), and epithelial cells (NMuMG) were analyzed by Northern blot with a 32 P-labeled syndecan-1 (syn-1) cDNA probe. The EtBr-stained gel is shown in the bottom panel. Panel B, immunoblot analysis. Culture medium from C 2 C 12 myoblasts was concentrated on a Q-Sepharose column followed by a chromatography on a DEAE column eluted with a linear gradient of 0 -1 M NaCl in 0.1 M Tris-HCl, pH 7.5. Equal amounts of fractions eluted with 0.85 and 1.0 M NaCl (lanes 2 and 3, respectively) were analyzed by immunoblot analysis using affinity-purified anti-syndecan-1 antibodies. In lane 1, culture medium from syndecan-1-expressing Schwann cells was applied as a positive control. Numbers and arrows to the left indicate the migration positions of molecular mass standards (in kDa).
FIG. 2. Syndecan-1 expression is down-regulated during myoblast terminal differentiation. 15 g of total RNA isolated from myoblasts (mb) and myoblasts induced to differentiate for 2 (d2) or 5 (d5) days were analyzed by Northern blot with a 32 P-labeled syndecan-1 cDNA probe (syn-1). Numbers and arrows to the right indicate syndecan-1 mRNA sizes (in kilobases). The same membrane was hybridized with probes for creatine kinase (ck) and myogenin (myo). The EtBrstained gel is shown in the bottom panel.
FIG. 3. The synthesis of syndecan-1 decreases during skeletal muscle differentiation. Myoblasts were grown on glass coverslips and induced to differentiate as described under "Experimental Procedures." Myoblasts (panels A and B) and cells incubated in differentiation medium for 2 days (panels C and D) or 4 days (panels E and F) were stained with affinity-purified rabbit anti-syndecan-1 and Texas Redconjugated goat anti-rabbit IgG antibodies (panels A, C, and E) to reveal cell surface syndecan-1 staining. The cells were then fixed and stained with DAPI to reveal the nuclei (panels B, D, and F). The bar corresponds to 25 m. tion of muscle differentiation these proteins bind to E-box regions and activate the transcription of skeletal muscle genes, such as creatine kinase (39). To evaluate the role of myogenin in the expression of syndecan-1, myoblasts were incubated for 2 days in differentiation medium with or without sodium butyrate (40), an agent known to block transcriptional activity of myogenin. The level of syndecan-1 mRNA was evaluated by Northern blot analyses. Fig. 5A shows that the decrease in the syndecan-1 mRNA level was independent of the presence of butyrate. In contrast, the expression of creatine kinase was totally abolished by the treatment. These results suggest that myogenin expression is not involved in the decrease of syndecan-1 transcription observed during differentiation.
To evaluate directly the role of myogenin and other MyoD proteins on the expression of syndecan-1, myoblasts were transiently transfected with a syndecan-1 promoter that contained the binding sequences for transcription factors SP1 (GC and GT boxes) and NF-B, but lacked E-boxes (p-244CAT). The expression driven by this promoter was compared with the expression from the p-667CAT vector during differentiation. Fig. 5B shows that the reporter expression obtained with both constructs was identical, and both were strongly inhibited during differentiation. Together these results suggest that the decrease in syndecan-1 expression observed during myogenesis is not related to the presence of E-boxes in the promoter region.
Syndecan-1 Expression during Myogenesis Is Growth Factordependent-In vitro induction of myoblast differentiation can be obtained by removal of growth factors from the medium. To evaluate the role of growth factors on the expression of syndecan-1, myoblasts were induced to differentiate in the presence and absence of defined growth factors (bFGF and TGF-␤) and growth medium (10% FCS), and the expression of p-244CAT was evaluated in transient transfection experiments. Fig. 6 shows that treatment of the cells for 2 days with bFGF (10 ng/ml) increased CAT activity slightly compared with differentiation medium (5% horse serum), whereas treatment with TGF-␤ (2.5 ng/ml) had no effect on syndecan-1 expression.
Under these conditions a strong inhibitory effect of bFGF and TGF-␤ on myogenin expression was observed (data not shown) (18,19). However, when bFGF and TGF-␤ were added together, activation of syndecan-1 promoter activity was obtained to values similar to those achieved when 10% FCS was used. These results suggest that growth factors like bFGF and TGF-␤ are necessary to maintain the expression of syndecan-1 and that the sequences contained in the proximal region of the promoter are sufficient to exert such regulation.
RA Inhibits Syndecan-1 Expression-RA, a vitamin A metabolite, plays a major role in skeletal muscle development (41) and skeletal muscle differentiation (21). As shown in Fig. 7A, the presence of RA strongly inhibited the transcriptional activity of p-244CAT. Maximal inhibitory activity was observed at 10 Ϫ6 M RA. Interestingly, as shown in Fig. 7B, the synergistic stimulatory effect observed for bFGF and TGF-␤ on syndecan-1 expression was totally abolished by RA. RA treatment also blocked the increase in syndecan-1 transcriptional activity caused by 10% FCS (Fig. 7B). These results indicate that RA, an inducer of skeletal muscle differentiation, inhibits the expression of syndecan-1 and abolishes the stimulatory effect of bFGF and TGF-␤ or serum. The sequences responsible of this inhibitory effect are contained in the proximal region of the promoter gene contained in p-224CAT. DISCUSSION The results presented in this paper demonstrate that the expression of syndecan-1, a transmembrane heparan sulfate proteoglycan (1), decreases during differentiation of skeletal muscle cells. This conclusion is based on analysis of syndecan-1 mRNA levels, immunofluorescent staining of cells, and the activity of a reporter construct containing a portion of the rat syndecan-1 promoter. This phenomenon correlates well with the observation that in adult skeletal muscle tissue syndecan-1 is absent (29) and with the loss of syndecan-1 in later stages of limb development (13).
The function of syndecan-1 in myoblasts and the consequences of syndecan-1 down-regulation during differentiation are not known. However, several roles for syndecan-1 and its down-regulation can be postulated. It is well known that proliferation of myoblasts is strongly stimulated by bFGF. At the same time this growth factor is a strong inhibitor of skeletal muscle differentiation (18). Heparan sulfate proteoglycans act as co-receptors for bFGF (24), allowing the binding of the growth factor to its signaling receptor on the plasma membrane and stimulating its biological effects. Therefore, the presence of syndecan-1 in the membrane might be critical for bFGF activity. Previously, we demonstrated that the expression of perlecan, a basal lamina-associated heparan sulfate proteoglycan, also decreases during differentiation (28) and that the synthesis of glypican (26), a lipid-anchored membrane-associated heparan sulfate proteoglycan, increases during differentiation (27).
These observations are potentially important in the context of the specificity of action of different heparan sulfate proteoglycans. It has been shown that addition of perlecan but not soluble syndecans or glypican restores bFGF signaling and biological activity to heparan sulfate-deficient fibroblasts (25). On the other hand, recombinant membrane-associated syndecans or glypican have been shown to promote bFGF signaling in a hematopoietic cell line that expresses low levels of heparan sulfate proteoglycans (42). It can be speculated that downregulation of syndecan-1 expression, together with the bFGF receptor down-regulation (43), makes the cells refractory to the presence of bFGF and therefore allows differentiation of skeletal muscle (44). Experiments to test the effects of constitutive FIG. 4. p-667CAT reporter activity is down-regulated during myoblast differentiation. C 2 C 12 myoblasts were transiently co-transfected with 5 g of p-667CAT and RSV-␤gal plasmids. The cells were then grown for 48 h in growth medium (day 0) and then changed to differentiation medium for 1, 2, or 3 days. Then the cells were harvested and the levels of CAT and ␤-gal activity determined. Activity (CAT/␤gal) is expressed as percentage of the activity detected in myoblasts. The results for the same experiment but using the p-myoCAT reporter are presented in the inset. The results correspond to two different experiments done in duplicate. syndecan-1 expression and inhibition of endogenous syndecan-1 expression in myoblasts are in progress.
Another potential role for syndecan-1 on the surface of myoblasts is the binding of extracellular matrix adhesive proteins. Syndecan-1 has been shown to bind several extracellular matrix adhesive molecules, including fibronectin (4); thrombospondin (5); tenascin (6); and collagen types I, III and V (7). Cell adhesion to these matrix proteins might not be required after differentiation when individual muscle fibers are in direct contact with basement membrane extracellular matrix. Furthermore, it has been shown that syndecan-1 can influence cell invasion (45). Myoblasts are able to migrate through basal lamina during early stages of differentiation (46). It is tempting to speculate that the presence of syndecan-1 on the surface of myoblasts may influence their migratory pathway to give rise to slow or fast primary myotubes (47).
Syndecan-1 expression in myoblasts appears to be regulated at the level of transcription. It has been suggested that downregulation of syndecan-1 expression during muscle differentiation could result from the presence of E-boxes, target sites for the action of myogenic regulators such MyoD and myogenin (39), in the syndecan-1 promoter (30). We tested this by transiently transfecting myoblasts with CAT reporter vectors containing the rat syndecan-1 promoter that contained or lacked E-boxes. Our results clearly demonstrate that the presence of the E-boxes is not required for the decrease in expression of syndecan-1 which is observed after differentiation is triggered. These results are supported by the finding that treatment of differentiating myoblasts with sodium butyrate, an agent known to inhibit myogenin expression (40,48), had no effect on the pattern of syndecan-1 expression. Furthermore, treatment of the myoblasts with bFGF or TGF-␤ strongly inhibited myogenin expression (18,19, and data not shown), without affecting syndecan-1 expression. Together, these results suggest that syndecan-1 expression in myoblasts is not directly regulated by myogenin.
In this study we also found that promoter activity was not significantly affected by bFGF or TGF-␤. However, exposure to both growth factors resulted in a significant increase in syndecan-1 gene activity. These results are similar to the finding of Elenius et al. (33) on syndecan-1 expression in fibroblasts. On the other hand, treatment of the cells with RA, an inducer of skeletal muscle differentiation (21), inhibited the activity of the syndecan-1 reporter. Furthermore, RA was able to abolish the stimulatory effect of bFGF and TGF-␤ as well as FCS. This is particularly interesting because both growth factors and RA are known to be expressed in the vicinity of condensing mesenchymal cells (49,50) and limb buds during early developmental stages (41).
These results indicate that the main regulatory elements responsible for the transcriptional regulation of syndecan-1 expression during myoblasts differentiation are contained in the proximal 277-bp segment of the syndecan-1 promoter. This region contains a putative TATA box sequence and consensus binding sites for several transcription factors such as SP1 and FIG. 5. Down-regulation of syndecan-1 expression is myogenin-and E-box-independent. Panel A, butyrate treatment does not affect syndecan-1 expression. 15 g of total RNA isolated from myoblasts (mb) and myoblasts induce to differentiate for 2 days (d2) in the presence or absence of butyrate was subjected to Northern blot analysis with 32 P-labeled syndecan-1 (syn-1) and creatine kinase (ck) cDNA probes. The EtBr-stained gel is shown at the bottom. Panel B, p-667CAT and p-244 CAT reporter activities are decreased during differentiation. C 2 C 12 myoblasts were transiently co-transfected with 5 g of p-667CAT or p-244CAT and RSV-␤gal expression plasmids. The cells were grown for 48 h in growth medium and then were transferred to differentiation medium for 1, 2, or 3 days. Cells were then harvested, and the levels of CAT and ␤-gal activity were determined. Activity (CAT/␤-gal) is expressed as fold decrease compared with the activity measured in myoblasts. The results correspond to two different experiments done in duplicate.
FIG. 6. Effect of growth factors on p-244CAT activity. C 2 C 12 myoblasts were transiently co-transfected with 5 g of p-244CAT and RSV-␤gal plasmids. The cells were grown for 48 h in growth medium and then changed for 2 days to growth medium (10% FCS) or differentiation medium (5% horse serum) containing 10 ng/ml bFGF or 2.5 ng/ml TGF-␤, or both. Then the cells were harvested and the levels of CAT and ␤-gal activity determined. The CAT/␤-gal activity obtained after 2 days in differentiation medium was set to 1, and the other activities are expressed relative to this value. The results correspond to two different experiments done in duplicate. NF-B. Vihinen et al. (51) have demonstrated that SP1-like transcription factors have an essential role in the regulation of the transcriptional activity of the syndecan-1 gene. This suggests that the down-regulation of syndecan-1 expression during myoblast terminal differentiation could be explained by variations in SP1 activity. There are several possible explanations for a decrease in SP1 activity. Changes in the level of expression of Sp1 during differentiation could occur, as was shown previously for different cells and tissues (52). An increase of SP1 phosphorylation which decreases its DNA binding activity, as demonstrated in terminal differentiation of liver (53), is also possible. Alterations of co-activators that are required by SP1 to modify gene expression have also been described (54). Finally, recent data demonstrated that SP1 and E2F, a cell cycle-regulated transcription factor, act synergistically to activate dihydrofolate reductase gene transcription in the absence of E2F DNA binding sites (55). E2F is sequestered by the retinoblastoma protein during differentiation (56), making SP1 possible less active and likely explaining the observed decrease in syndecan-1 expression during myoblast terminal differentiation.
The inhibitory effect of RA is interesting not only as an explanation for the decrease of syndecan-1 expression during myoblasts differentiation, but also because suppression of syndecan-1 expression has been shown to be associated with malignant conversion (57). Tumor necrosis factor-␣ is the only other factor described which restricts the expression of syndecan-1 (58). The inhibitory effect of RA on syndecan-1 expression could be explained by the presence of putative RA response elements in the promoter region. This sequence is sufficient to cause the inhibitory effect of RA on the murine Oct4 promoter (59). The syndecan-1 promoter region contains sequences that are similar, but not identical to RA response elements. These are located 55, 80, and 100 bp upstream of the TATA box and are included in the p-244CAT reporter.
The expression of perlecan during myogenesis is also downregulated (28). As indicated by Vihinen et al. (51), the promoter region of perlecan (60) resembles that of mouse syndecan-1. The upstream regions of these genes lack canonical TATA and CAAT boxes, but several SP1 transcription factor binding sites are present within the first 200 bp of the promoter. These observations suggest that similar regulatory mechanisms are involved in the regulation of expression of these cell surface macromolecules.
In summary, we have found that the expression of syndecan-1 is down-regulated during skeletal muscle differentiation. This phenomenon is modulated by growth factors and RA but is E-box-and myogenin-independent. The regulatory sequences responsible for this modulation are contained within a 277-bp segment of the gene which contains consensus binding sites for several transcriptions factor such as SP1, NF-B, and RAbinding proteins. These results will help toward understanding the complex regulation during development and differentiation of this key integral membrane heparan sulfate proteoglycan. FIG. 7. Effect of retinoic acid on p-244CAT expression. C 2 C 12 myoblasts were transiently co-transfected with 5 g of p-244CAT and RSV-␤gal plasmids. The cells were grown for 48 h in growth medium and then for 2 days in the medium indicated. Cells were then harvested, and the levels of CAT and ␤-gal activity were determined. The CAT/␤gal activity obtained after 2 days in differentiation medium was set to 1, and the other activities are expressed relative to this value. The results were obtained from two different experiments done in duplicate. Panel A, retinoic acid inhibits p-244CAT expression. Transfected cells were incubated for 2 days in differentiation medium (5% horse serum) containing the indicated RA concentrations. Panel B, retinoic acid abolishes the stimulatory effect of growth factors on p-244CAT expression. Transfected cells were incubated for 2 days in growth medium (10% FCS), differentiation medium (5% horse serum), or differentiation medium containing 10 ng/ml bFGF and 2.5 ng/ml TGF-␤ in the presence of 10 Ϫ6 M RA.