| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 281, Issue 45, 34406-34420, November 10, 2006
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
-Tropomyosin Gene in the Three Muscle Cell Types*
12
1234
From the
Unité INSERM 441, Avenue du Haut-Lévêque, 33600 Pessac, France, Unité Mixte de Recherche CNRS 5164, Université Bordeaux 2, 146 rue Léo Saignat, 33076 Bordeaux, France, and ¶Unité Mixte de Recherche CNRS 8080, Université Paris-Sud, 91405 Orsay, France
Received for publication, March 10, 2006 , and in revised form, September 6, 2006.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
-tropomyosin (-Tm) gene in muscle cells. Promoter analysis revealed that a 284-bp proximal promoter region of the Xenopus laevis -Tm gene is sufficient for maximal activity in the three muscle cell types. The transcriptional activity of this promoter in the three muscle cell types depends on both distinct and common cis-regulatory sequences. We have identified a 30-bp conserved sequence unique to all vertebrate -Tm genes that contains an MCAT site that is critical for expression of the gene in all muscle cell types. This site can bind transcription enhancer factor-1 (TEF-1) present in muscle cells both in vitro and in vivo. In serum-deprived differentiated smooth muscle cells, TEF-1 was redistributed to the nucleus, and this correlated with increased activity of the -Tm promoter. Overexpression of TEF-1 mRNA in Xenopus embryonic cells led to activation of both the endogenous -Tm gene and the exogenous 284-bp promoter. Finally, we show that, in transgenic embryos and juveniles, an intact MCAT sequence is required for correct temporal and spatial expression of the 284-bp gene promoter. This study represents the first analysis of the transcriptional regulation of the -Tm gene in vivo and highlights a common TEF-1-dependent regulatory mechanism necessary for expression of the gene in the three muscle lineages. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
We have previously identified three ortholog tropomyosin genes, viz. -Tm, 
-Tm, and Tm-4 (also called 
-Tm), in the amphibian Xenopus laevis (6-8). We have shown that each of the Tm genes has a distinct temporal and spatial pattern of expression during early development and differentiation. The amphibian -Tm gene is structurally related to its mammalian orthologs. This gene contains one promoter and two sets of alternatively spliced exons and can produce both skeletal and smooth muscle isoforms (8, 9). The amphibian Tm-4 gene is identical to the avian ortholog, with two promoters and two sets of alternatively spliced exons. The distal promoter is used in cardiac muscle cells, whereas the internal one is used to produce the ubiquitously expressed non-muscle Tm isoform (7). The X. laevis -Tm gene is by far the most complex of all. Using RNA expression and genomic DNA analysis, we have shown that, in common with its avian and mammalian orthologs, the frog -Tm gene contains two promoters (internal and distal) and three sets of alternatively spliced exons. However, we have not found evidence of a brain-specific exon (10, 11). The low molecular mass non-muscle Tm transcripts are found throughout oogenesis and embryogenesis and in adult tissues. In contrast, the muscle -Tm transcripts exhibit a restricted expression pattern. Transcripts accumulate from the late neurula stage in both the somites and heart and later during development in smooth muscle tissues. Spatiotemporal expression of the gene is regulated through the two promoters. The internal promoter is ubiquitously active, whereas the distal promoter is active only in muscle cells. In the adult frog, the -Tm gene is expressed in the three muscle cell types, as are its avian and mammalian orthologs. Transcriptional regulation of the -Tm gene therefore provides a suitable model for the characterization of the cis-regulatory sequences involved in the transcriptional program that operates in all three muscle lineages.
Although the transcriptional regulation of several sarcomeric protein-coding genes has been studied in detail, very little is known about sequences that may regulate tropomyosin expression in the different muscle types. In Drosophila, myocyte enhancer factor-2 has been shown to be a positive regulator of expression of the Tm gene (12), but very few studies have addressed the transcriptional control of Tm genes in vertebrates. It has been shown that chicken -Tm gene transcription relies on several cis-sequences, including an E box, a CArG box, and a C box (13). In the rat, -Tm gene expression seems to be controlled by a 99-bp enhancer that differentially regulates both -Tm and a transcription unit (called N5), resulting in distinct tissue-specific expression patterns of the two (14). It is now established that multiple independent cis-regulatory regions, or modules, are required to direct the complete developmental pattern of expression of individual muscle-specific genes (15). For instance, the MyoD family members that govern skeletal muscle differentiation and that constitute the paradigm for cell type-specific transcription factors function by interacting with cofactors (16). Moreover, like other developmental processes, muscle differentiation relies on a transcriptional circuit that is dependent on combinatorial associations of cell type-specific and widely expressed transcription factors. These associations interpret cell identity, extracellular signals, and positional information within the embryo (17). Indeed, some factors involved in muscle gene activation are not restricted to muscle cell types. Among them, ubiquitously expressed transcription enhancer factor-1 (TEF-1) has emerged as a transcription factor implicated in the specific activation of several muscle genes (18). Originally described as a factor binding to non-muscle-specific cis-elements of the SV40 enhancer, TEF-1 has also been found to bind to the MCAT (5'-CATTCCT-3') regulatory sequence (19, 20). There are four TEF-1 genes in vertebrates that are expressed in a complex pattern and that can produce several isoforms (21-23). Notably, the MCAT sequence that binds TEF-1 is present in the regulatory regions of several mammalian and avian cardiac and skeletal muscle-specific genes, including those encoding cardiac muscle troponin T and skeletal muscle -actin, - and -myosin heavy chains, and -acetylcholine receptor (24-36). The MCAT sequence has also been implicated in regulation of the mammalian smooth muscle -actin gene in a complex mechanism that involves several single-strand DNA-binding proteins (37, 38).
In this study, we have examined the molecular mechanisms that regulate the X. laevis -Tm gene in cardiac, skeletal, and smooth muscle cells. For this analysis, we have used a genomic clone encompassing the 5'-region of the gene (10). We have identified a 284-bp promoter fragment whose activity in the three muscle types is dependent on the presence of multiple cis-regulatory elements. Among them, we found an MCAT sequence that is embedded in a highly conserved 30-bp module unique to all vertebrate -Tm genes. This MCAT sequence can bind TEF-1 protein in vitro and in vivo in differentiated muscle cells. In smooth muscle cells, TEF-1 can be redistributed to the nucleus when cells are cultured under serum-deprived conditions that maintain them in a differentiated state. We found that overexpression of TEF-1 mRNA in embryonic cells activates both endogenous and exogenous -Tm genes. In the later case, we have demonstrated that this activation occurs through the MCAT sequence. Finally, mutation of the MCAT sequence leads to deregulation of the activity of the 284-bp transgene in embryonic and juvenile tissues. Taken together, our data suggest that correct regulation of the vertebrate -Tm gene in all three muscle cell types depends on TEF-1.
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
Construction of Reporter Plasmids—Promoter fragments of the -Tm gene were generated from a previously isolated genomic clone (10). A BamHI/BamHI fragment (-1763/+40) was excised from the genomic clone and inserted into the BglII site of the pGL3-Basic vector (Promega Corp.) to generate the pGL1763LUC construct. A SacI/BamHI fragment (-284/+40) was excised from the genomic clone and subcloned into the SacI and BglII sites of the pGL3-Basic vector to generate the pGL284LUC construct. The pGL235LUC, pGL120LUC, and pGL78LUC constructs were generated by PCR from the genomic clone using oligonucleotide primers containing SacI and BglII linkers such that PCR products could be directionally cloned into the pGL3-Basic vector. Site-directed mutagenesis was performed using an in vitro site-directed mutagenesis system (Promega Corp.) following the manufacturer's instructions. The primer mutant sequences were as follows (the mutated sequences are underlined): E box-1, 5'-GAGGCTGGAAAAAGTTAGAGTTTGCACTTC-3'; E box-2, 5'-AAATGCAGTTCTAGAGTCGGCAGAAGC-3'; E box-3, 5'-CCTTATATAGTTCTAGACAGTCAAATGGTC-3'; E box-4, 5'-GCTCCCTTCCAGATCTAGACCTGGAAGTTT-3'; A/T-rich region, 5'-AGCCATCCTTAAGGGAAGCCAGCAAGTGG-3'; CArG-like element, 5'-AAGGCAAGGCTCCCAGGTACCTATTGGGTG-3'; MCAT, 5'-TTGGGTGTCCGGAGGTACCTGTGTGCCCC-3'; and GC box, 5'-AATGTGTGTGCTGATCTTCGCTGCCTTC-3'. All plasmids were confirmed by sequencing.
DNA Transfections—U8A4, PAC1, A7r5, and NIH-3T3 cells were seeded the day before transient transfection assays into 12-well plates at a density of 25,000 cells/cm2 when cultured in defined medium and at a density of 15,000 cells/cm2 when cultured in medium supplemented with 10% FCS. 400 ng of luciferase reporter gene constructs and 400 ng of -galactosidase control vector pHook2-LacZ (Invitrogen) were transfected per well using 2.4 µl of TransFast (Promega Corp.) in Opti-MEM (Invitrogen). After 1 h, 1 ml of culture medium was added, and cells were cultured for 48 h before analysis. C2/7 cells were grown in 12-well plates at a density of 5,000 cells/cm2 until reaching 50% confluence and then transfected as described above. Following transfection, cells were induced to differentiate over 48 h before cell extracts were prepared.
Cardiomyocytes were seeded into 6-well plates at a density of 20,000 cells/cm2 3 days before transfection. Cells were incubated for 5 h using 4 µl of FuGENE 6 transfection reagent (Roche Applied Science) and 1 µg of plasmid DNA. After being rinsed with DMEM, cells were cultured for 48 h before cell extracts were prepared. For the luciferase assays, cell extracts were prepared using reporter lysis buffer (Promega Corp.) following the manufacturer's instructions. Luciferase and -galactosidase activities were measured according to the manufacturer's instructions. All luciferase activity values were normalized to -galactosidase activity values. The figures were obtained from at least two independent experiments, with each construct tested in triplicate. The pGL3-Control (Promega Corp.) and pGL3-Basic vectors were used as controls.
Comparative Sequence Analysis—The nucleotide sequences of vertebrate -Tm gene promoters were analyzed using sequence similarity computer programs (Infobiogen, Evry, France).
In Vitro Transcription and Translation—Chicken TEF-1 cDNA was kindly provided by Dr. C. Ordahl and subcloned into the pCS2+ vector (23). The resulting plasmid and the X. laevis serum response factor (SRF) cDNA cloned into the pSP64T vector (43) were used for in vitro transcription. Synthetic capped mRNAs were made using the mMESSAGE mMA-CHINE kit (Ambion, Inc.) according to the manufacturer's instructions. The products were purified on mini-quick RNA columns (Roche Applied Science). 200-400 ng of mRNA were translated in vitro in reticulocyte lysate (Promega Corp.) according to the manufacturer's instructions. The integrity and expected molecular masses of the proteins were assessed by resolving radiolabeled reaction products on a 10% polyacrylamide gel.
Western Blot Analysis—Total protein concentration in the samples was determined by the Bradford assay (Bio-Rad). 10 µg of nuclear proteins or 20 µg of total proteins were separated on a 10% SDS-polyacrylamide gel, followed by transfer onto Immobilon-P membrane (Millipore, Bedford, MA). The membranes were blocked for 1 h at room temperature in 5% nonfat dried milk made in 130 mM NaCl and 25 mM Tris (pH 7.5) containing 0.05% Tween 20 (TBST). After being washed with TBST, the membranes were incubated for 2 h in TBST supplemented with primary antibody. The immunoblots were then washed three times with TBST and incubated for 1 h with a 1:10,000 dilution of a horseradish peroxidase-conjugated anti-mouse or anti-rabbit Ig antibody in TBST. The blots were washed three times with TBST and developed using the ECL Western blot detection system (Amersham Biosciences). Signals were visualized on x-ray film after exposures of membranes of between 30 s and 5 min. The primary antibodies used were anti-tropomyosin monoclonal antibody TM311 (1:1000 dilution; Sigma) (44), anti-TEF-1 monoclonal antibody (1:1000 dilution; BD Transduction Laboratories), and anti-TEF-1A polyclonal antibody (generously provided by Dr. I. Farrance).
Electrophoretic Mobility Shift Assay (EMSA)—Nuclear extracts were prepared from cells according to established methods (45), and protein concentration was determined by the Bradford assay. All extracts were snap-frozen in liquid nitrogen and stored at -80 °C. 500 pmol of each sense and antisense oligonucleotide were annealed at 95 °C for 5 min in 50 µl of 50 mM NaCl, 10 mM Tris-HCl (pH 7.5), and 1 mM EDTA and then cooled to room temperature. 25 pmol of double-stranded oligonucleotides were end-labeled using [
-32P]ATP (6000 Ci/mmol) in the presence of T4 polynucleotide kinase. Labeled double-stranded oligonucleotides were recovered after purification on a Qiagen column in 10 mM Tris-HCl (pH 8) and 1 mM EDTA.
For EMSAs, 5-10 µg of nuclear extracts or 1-2 µl of in vitro translated TEF-1 and SRF were incubated for 20 min in binding buffer (10 mM Tris-HCl (pH 7.5), 5 mM HEPES, 1 mM EDTA, 100 mM KCl, 1 mM dithiothreitol, 1 mg/ml bovine serum albumin, and 10% glycerol) containing 0.20 µg of poly(dI-dC) and 4 x 104 cpm end-labeled probe. After a 20-min incubation at room temperature, the samples were subjected to electrophoresis on 5% nondenaturing polyacrylamide gels in 0.5x buffer containing 44.5 mM Tris-HCl (pH 8.3), 44.5 mM boric acid, and 1 mM EDTA for 2 h at 150 V. Gels were dried and exposed to films with an intensifying screen at -80 °C for 24-48 h. For supershift analysis, 750 ng of anti-TEF-1 monoclonal antibody were preincubated with nuclear extracts for 10 min before addition of the probe.
The nucleotide sequences of the probes and competitor DNA were as follows (only upper strands are shown): MCAT, 5'-GTGTCCGGAGGAATGTGTGTGCCC-3'; MCAT mutant, 5'-GTGTCCGGAGGTACCTGTGTGCCC-3'; CArG-like element, 5'-AAGGCAAGGCTCCCAAAAAAGTATTGGGTGT-3';c-fos SRE, 5'-CAGGATGTCCATATTAGGACATCTGCGT-3' (46); CArG1 smooth muscle myosin heavy chain (SM-MHC), 5'-GACTTCCTTTTATGGCCTGA-3' (47); CArG4 desmin, 5'-AACACCCATATATGTAAAAT-3' (48); MCAT troponin T (TnT), 5'-CGTGTTGCATTCCTCTCTGGATC-3' (32); and CArG-MCAT, 5'-GGCTCCCAAAAAAGTATTGGGTGTCCGGAGGAATGTGTGTGCCC-3'. Underlined nucleotides correspond to core-binding sites, and boldface nucleotides represent mutated nucleotides.
Identification of proteins in gel shift complexes was performed as described (24, 49). The mobility shift reactions were scaled up 5-fold in a final volume of 20 µl. After electrophoresis for 3 h, the gel was soaked twice for a total of 3 min in 2% SDS, 62.5 mM Tris, and 25 mM dithiothreitol and dried. The protein·DNA complexes were then identified by a 2-h autoradiography, and the bands of interest were excised and loaded onto a 10% SDS-polyacrylamide gel. Immunoblot analysis was carried out as described above with anti-TEF-1 and anti-TEF-1A antibodies.
Chromatin Immunoprecipitation (ChIP)—ChIP assays were performed on PAC1 smooth muscle cells cultured in 1 or 10% serum and on C2C12 myoblasts and myotubes according to a published protocol (50). Briefly, 2-5 x 106 cells were used in the assays. The DNA and protein were cross-linked in situ by treatment with 1% (v/v) formaldehyde for 10 min at room temperature. Following chromatin sonication, the lysates were precipitated overnight using either 5 µg of mouse anti-TEF-1 monoclonal antibody or 5 µg of rabbit anti-TEF-1 polyclonal antibody (AnaSpec, Inc., San Jose CA). Protein A-agarose beads were added to purify immune complexes. Cross-linking was reversed by heating the samples overnight at 65 °C. RNA was degraded with RNase A for 4 h at 65°C, and proteins were degraded by proteinase K treatment for 2 h at 45 °C. The DNA was further purified by phenol extraction and recovered in 10 mM Tris-HCl (pH 8) and 1 mM EDTA.
After DNA purification, samples were subjected to PCR with the following primers designed for the -Tm gene promoter (14): upper primer, 5'-GTCATATCCACCGTCGACTGG-3'; and lower primer, 5'-CGTGCTCCCTGATATGTACTTTCC-3'. The amplified PCR product was 197 bp. For the smooth muscle -actin gene promoter (51), the following primers were used: upper primer, 5'-CTCAAGCCCTTAGCTAATGG-3'; and lower primer, 5'-AACTCTCTAATCTGGGTGGC-3'. The amplified PCR product was 246 bp. PCR was performed for 40 cycles, and 10 µl of PCR products were analyzed by 2% agarose gel electrophoresis.
Immunofluorescence—Cells grown on coverslips were fixed for 10 min at 4 °C in methanol and then incubated for 2 h at 37 °C with anti-TEF-1A or anti-SRF (catalog no. sc-335 X, Santa Cruz Biotechnology, Inc.) primary antibody at a 1:400 dilution. Cells were incubated for 1 h at room temperature with fluorescein isothiocyanate-conjugated secondary antibody. Nuclear staining was performed with iodide propidium. Negative controls were performed without primary antibody. Immunofluorescence staining was examined using a Nikon microscope.
DNA and RNA Injections—Eggs were obtained from X. laevis females, cultured, and microinjected as described previously (52). Embryonic stages were determined according to Nieuwkoop and Faber (53). Embryos were injected at the two-cell stage in each blastomere with the pGL285LUC and pGLMCATLUC reporter constructs (45 pg/blastomere) along with synthetic TEF-1 mRNAs (100 pg/blastomere). Three pools of three embryos were collected at gastrula stage 12, homogenized in passive lysis buffer (Promega Corp.), and assayed for luciferase activity. Data were normalized per embryo. For reverse transcription (RT)-PCR analysis, 500 pg of TEF-1 mRNAs were injected into each blastomere of two-cell stage embryos. Animal caps were then excised at stage 8.5 and cultured until stage 11. RNA was extracted from 10 animal caps and analyzed by RT-PCR as described previously (10) using the following primers: elongation factor-1, 5'-CAGATTGGTGCTGGATATGC-3 (upper primer) and 5'-ACTGCCTTGATGACTCCTAG-3' (lower primer); and -Tm, 5'-AGATGTCAAACTGGACAAGG-3' (upper primer) and 5'-CATCTGCAGCCTTCTCAGCC-3' (lower primer). PCR products were resolved on agarose gel.
Transgenesis—The PstI/BamHI fragments from the pGL284LUC and pGLMCATLUC plasmids were subcloned into the green fluorescent protein (GFP) vector kindly provided by Dr. T. Mohun to create the pGL284GFP and pGLMCATGFP constructs, respectively. -Tm promoter-GFP reporter sequences were excised from the backbone plasmid using NotI (New England Biolabs). The transgene was then used to generate transgenic Xenopus embryos following previously described protocols (54, 55). Reporter gene expression was detected by fluorescence in live embryos using an Olympus SZX12 loupe coupled to a Diagnostic Instruments Model 3.2.0 charge-coupled display camera.
RT-PCR Analysis of Transgenic Animals—RNA was extracted from juvenile tissues and analyzed by RT-PCR as described previously (10) using the following primers: striated muscle -Tm, 5'-CATTGAGGGTGATCTGGAAC-3' (upper primer) and 5'-ATAGAGGTGACCATTGGAAGC-3' (lower primer); smooth muscle -Tm, 5'-ATCTCCCAACGGAAAAGCAG-3' (upper primer) and 5'-TCTGGTCCAACATCTGATGC-3' (lower primer); ornithine decarboxylase, 5'-GTCAATGATGGAGTGTATGGATC-3' (upper primer) and 5'-TCCATTCCGCTCTCCTGACCAC-3' (lower primer); and GFP, 5'-GTGAAGGTGATGCAACATACG-3' (upper primer) and 5'-TCAAGAAGGACCATGTGGTG-3') (lower primer).
Whole Mount in Situ Hybridization—Whole mount in situ hybridization was performed using the XTM2 probe as described previously (56).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
-Tm Gene That Confers Maximal Activity in the Three Muscle Cell Types—We have shown previously that the 5'-region of the X. laevis -Tm gene is structurally related to its mammalian and avian orthologs; it contains two promoters flanking a pair of alternatively spliced exons, 2a and 2b, with exon 2a being a smooth muscle-specific exon (10). mRNAs encoding muscle tropomyosins originate from the distal promoter, whereas mRNAs encoding low molecular mass non-muscle tropomyosins originate from the internal one. As a first step toward the identification of sequences required for maximal promoter activity in muscle cells, a series of deletion constructs was created from a genomic fragment spanning 1763 bp upstream of the transcription initiation site. These different constructs were tested in skeletal, cardiac, and smooth muscle cells as well as in fibroblasts. The C2/7 cell line was used as a model for skeletal muscle differentiation. Transcriptional activities were tested both at the myoblast stage and after differentiation, when myotubes have formed. Transcriptional activities in cardiac and smooth muscle cell types were assessed in transfected neonatal rat primary cardiomyocytes and in the U8A4 cell line, respectively. We have demonstrated previously that U8A4 cells retain transcriptional and post-transcriptional potencies while maintaining a differentiated phenotype when expressing the smooth muscle -Tm isoform (40). We have also shown that U8A4 cells can be induced to dedifferentiate when the NFAT (nuclear factor of activated T cells) signaling pathway is blocked (57).
|
Because smooth muscle cell lines can behave very differently with respect to promoter activity, we tested the pGL284LUC construct in PAC1 and A7r5 smooth muscle cells. These cell lines have been considered to exhibit a differentiated phenotype (39). We found that the pGL284LUC construct was transcriptionally active in both cell lines, showing an efficiency equivalent to that observed in U8A4 cells. We obtained a similar result using the mouse SM22 promoter (data not shown). In contrast, in NIH-3T3 mouse fibroblasts, the 284-bp promoter yielded no activity (data not shown), suggesting that this region is regulated in a muscle-specific manner. Together, these results indicate that the 284-bp proximal region of the -Tm promoter can confer maximal activity in the three muscle cell types.
An MCAT Sequence Is Essential for Full -Tm Promoter Activity in the Three Muscle Cell Types—Sequence analysis of the 284-bp promoter revealed the presence of several consensus DNA sequences known to be required for the binding of muscle-specific transcription factors (Fig. 2A). Among them are four E box motifs (hereafter referred to as E box-1-4 from the proximal to distal positions), a GC-rich region, an A/T-rich region, a slightly imperfect CArG box (CArG-like element), and one MCAT sequence in an antisense orientation. To evaluate the functional importance of the different cis-sequences in mediating the transcriptional activity of the 284-bp promoter, each sequence was mutated and tested in transient transfection assays in the three muscle cell types (Fig. 2B). In these experiments, the activity of the wild-type 284-bp promoter was set at 100%. Mutation of E box-4 in the context of the promoter had no effect on its activity in cardiac muscle cells, but rather stimulated its activity in smooth and skeletal muscle cells. Mutation of E box-3 had no effect on the activity of the promoter in skeletal or cardiac muscle cells, but resulted in a 50% reduction of activity in smooth muscle cells. Conversely, mutation of E box-2 had a more drastic effect on promoter activity in cardiac and skeletal muscle cells, but no significant effect in smooth muscle cells. Finally, E box-1 mutation led to 80% reduced promoter activity in skeletal muscle cells and a 40-50% reduction in cardiac and smooth muscle cells. Mutation of the GC box led to an 
70% reduction of promoter activity in the three muscle cell types. Mutation of the A/T-rich region had a significant effect only in skeletal muscle cells, reducing the activity of the promoter by 80%. Mutation of the CArG-like element had no effect on promoter activity in smooth and cardiac muscle cells, but induced a 60% reduction of promoter activity in skeletal muscle cells. These results suggest that the -Tm gene is controlled by both distinct and common cis-sequences in different muscle cell types.
The most striking results were obtained by mutation of the MCAT sequence. This had a dramatic effect on the transcriptional activity of the promoter in all three muscle cell types. The MCAT mutation led to an 80% reduction of activity in smooth (U8A4 and PAC1) and cardiac muscle cells and an almost complete abolition of promoter activity in skeletal muscle cells. Together, these results indicate that, although multiple cis-regulatory elements are involved in the transcriptional activity of the 284-bp -Tm gene promoter in muscle cells, the MCAT sequence appears to be an important positive regulator in all muscle cell types.
|
-Tm Genes—In studying genes (such the -Tm gene) that show conservation in both function and expression domains across species, DNA sequence comparison can be useful in identifying conserved regulatory regions. We therefore compared the 284-bp amphibian promoter with known vertebrate -Tm gene promoters. Although the overall sequence conservation between the different -Tm gene promoters is poor, a highly conserved 30-bp region encompasses the MCAT sequence and the flanking CArG-like sequence in all promoters examined (Fig. 3). There are only two or three nucleotide substitutions over this region in the amphibian, avian, and mammalian sequences. These substitutions are found in the 13-nucleotide linker sequence between the CArG-like element and the MCAT sequence. This 30-bp sequence is unique to the -Tm genes and is not found either in the promoter regions of the other tropomyosin gene family members or in other gene promoter regions. Moreover, the location of this sequence is conserved with respect to the transcription initiation sites that have been mapped in the Xenopus, quail, and rat genes. A similar spatial organization of the CArG-like element and the MCAT sequence is also found in all promoters. Together, these findings indicate that the 30-bp sequence has been highly conserved during evolution and suggest its potential importance in the regulation of -Tm gene expression.
The MCAT Sequence Binds TEF-1 in Vitro—The MCAT sequence has been described as the recognition site for TEF-1 protein in non-muscle, cardiac, and smooth muscle cells (18). To determine whether the MCAT sequence found in the promoter region of the -Tm gene could function as a TEF-1-binding site, we tested the ability of double-stranded oligonucleotides encompassing the MCAT sequence to bind TEF-1 in a gel shift assay. As shown in Fig. 4A (lane 1), TEF-1 translated in vitro bound very efficiently to the MCAT sequence. The retarded complex could be competed by an excess of the unlabeled probe (an oligonucleotide corresponding to the MCAT consensus sequence of the chicken TnT promoter), but not by a mutated version of the MCAT sequence (Fig. 4A, lanes 2-5). With nuclear extracts from cardiomyocytes, smooth muscle cells, or myotubes, we observed three major complexes named C1, C2, and C3 using the MCAT probe (Fig. 4, B-D, lanes 1). These complexes were competed by the unlabeled MCAT probe or the chicken cardiac muscle TnT gene MCAT sequence (Fig. 4, B-D, lanes 2 and 3, respectively), but not by a mutated MCAT site (lanes 4). Our results are in agreement with previous data showing the formation of three distinct complexes varying in intensity with nuclear extracts from striated and smooth muscle cells (24, 38, 49).
|
We next performed EMSA with nuclear extracts from non-muscle cell myoblasts (Fig. 4E, lanes 1-3) and NIH-3T3 cells (lanes 4-6). The three complexes specific for the MCAT probe were formed, and only the lower complex, C3, was supershifted by anti-TEF-1 antibody (Fig. 4E, lanes 3 and 6). Previous studies have reported the inability of some anti-TEF-1 antibodies to work in supershift analysis, indicating that some MCAT-binding activities may not be related to TEF-1 (38, 49). To test this possibility more directly, we investigated whether each of the three protein·MCAT DNA complexes contains polypeptides antigenically related to TEF-1. Briefly, protein·MCAT DNA complexes made with cardiomyocyte and U8A4 smooth muscle cell nuclear extracts were separated by EMSA, and the position of each complex was localized by autoradiography (Fig. 4F, lanes 1 and 2). Each complex was cut from the gel, and its constituent proteins were subjected to SDS-PAGE and Western blot analysis with anti-TEF-1 antiserum. As shown in Fig. 4G, all three complexes contained a 52-kDa TEF-1 polypeptide, but the intensity of the TEF-1 signal increased from complex C1 (lanes 1 and 4) to complex C2 (lanes 2 and 5) to complex C3 (lanes 3 and 6). This suggests that complexes C1 and C2 contain protein antigenically related to TEF-1. Because the intensity of complexes C1 and C2 was higher than that of complex C3, we hypothesized that the TEF-1-related polypeptide contained in complexes C1 and C2 is either much more prevalent than the TEF-1 in complex C3 or present at a lower level but with higher affinity for the MCAT sequence.
The MCAT Sequence Binds TEF-1 in Vivo—To ascertain the presence of TEF-1 family members in the nuclear extracts we used in EMSA, we performed Western blot analyses with muscle and non-muscle extracts. The results of immunoblot analysis with anti-TEF-1 antibody showed that all nuclear extracts contained a 52-kDa polypeptide antigenically related to TEF-1 (Fig. 4H, lanes 1-5).
|
|
-actin has been shown to contain MCAT elements active in smooth muscle myocytes (38), this promoter was used as a control. PAC1 smooth muscle cells cultured in either 1 or 10% serum, C2C12 myoblasts, and myotubes were fixed directly with formaldehyde. Cross-linked chromatin was immunoprecipitated with anti-TEF-1 antibody. The precipitated chromatin DNA was then purified and subjected to PCR analysis for enrichment of the target sequences. As shown in Fig. 5, anti-TEF-1 antibody specifically enriched the MCAT-containing region of the -Tm promoter in differentiated smooth muscle cells (1% serum) and in myotubes (compare lanes 6 and 12 with lanes 5 and 11). In PAC1 smooth muscle cells cultured in 10% serum or in myoblasts, there was no binding of TEF-1 to the -Tm gene promoter region (Fig. 5, compare lanes 3 and 9 with lanes 2 and 8). Therefore, the binding of TEF-1 to the MCAT sequence seems to be specific to the differentiation status of the cells. The smooth muscle -actin gene region that contains MCAT sequences was enriched in samples immunoprecipitated with anti-TEF-1 antibody (Fig. 5, lanes 3 and 6). These MCAT elements have been shown to contribute to cell type-specific regulation of the smooth muscle -actin gene (38). We have shown here for the first time that this correlates with the in vivo binding of TEF-1 to that region. An unexpected finding of our ChIP assays is that, in myotubes, anti-TEF-1 antibody bound to MCAT sequences of the smooth muscle -actin gene promoter (Fig. 5, compare lanes 8 and 9 with lanes 11 and 12). However, a previous study reported the immunodetection of smooth muscle -actin in C2C12 myoblast cells within hours of exposure to differentiation medium (58).
The CArG-like Element of the 30-bp Module Does Not Bind SRF Efficiently—CArG boxes are 10-bp elements with the sequence CC(A/T)6GG that are present in promoter regions of immediate-early growth response genes. CArG boxes are also found in many muscle-specific gene promoters, where they regulate transcriptional activity by binding SRF (59). Several studies have described divergent CArG boxes (named CArG-like elements) that have a single base mismatch compared with the original sequence but still bind SRF, although with lower affinity. Some CArG-like elements present a C or G substitution in the (A/T)6 core, such as the CArG-like element (5'-CCTTTTTGGG-3') found in the mouse SM-MHC gene promoter (47). Other elements exhibit a slight change at the 5'-or3'-end of the sequence, such as the CArG4 element (5'-CCATATATGT-3') of the mouse desmin gene promoter (48). The CArG-like element (5'-CCAAAAAAGT-3') located 13-bp upstream of the MCAT site within the 30-bp module of the -Tm gene promoter belongs to this type of element. We performed a gel shift assay to determine whether this CArG-like element can bind SRF. A double-stranded oligonucleotide containing the CArG-like sequence and 8- and 12-bp of 5'- and 3'-flanking sequence, respectively, was labeled and incubated with muscle cell nuclear extracts. Our first attempts failed to detect any binding between nuclear extracts and the CArG-like element (data not shown). We then turned to in vitro translated X. laevis SRF protein and failed to detect any binding even when using increasing amounts of protein or after extensive exposure of the autoradiogram (data not shown). To control whether in vitro translated SRF was able to bind to a bona fide CArG box and to the CArG-like sequence, we used the CArG1 box sequence (5'-CCTTTTATGG-3') of the mouse SM-MHC gene and the CArG4 element (5'-CCATATATGT-3') of the mouse desmin gene as probes (Fig. 6, A and B). In vitro translated SRF bound the CArG1 box sequence efficiently, resulting in a complex that could be competed by an excess of the unlabeled probe, the CArG box of the c-fos gene, or the CArG-like element of the desmin gene (Fig. 6A, lanes 2, 4, and 5). The CArG-like element of the -Tm promoter could compete the complex, albeit not completely, even at a 200-fold excess of the competitor (Fig. 6A, lane 3). SRF formed a complex with the CArG4 desmin probe that was competed by the unlabeled probe and the CArG boxes of SM-MHC and the c-fos gene (Fig. 6B, lanes 2-5 and 8-13). The CArG-like element of the -Tm gene could not compete the complex totally even when in large excess (Fig. 6B, lanes 5-7). Together, these results demonstrate that the CArG-like element of the 30-bp module of the -Tm gene promoter cannot bind SRF or does so with very low affinity. This is similar to previous findings showing, for example, that the CArG-like element of the cardiac muscle TnT gene does not compete efficiently when analyzed by competition footprinting (33).
Previous reports have suggested the importance of the CArG box and MCAT sites within the avian TnT and mammalian skeletal muscle actin gene promoters (30, 60). It has also been shown that SRF and TEF-1 can interact through a direct and stable interaction (60). To investigate whether SRF can help in the binding of TEF-1 to the MCAT sequence, we performed EMSA using a double-stranded oligonucleotide of 44 bp covering the 30-bp module containing the CArG-like and MCAT sequences with 9 and 5 bp of 5'- and 3'-flanking sequence, respectively. We used a fixed amount of in vitro translated TEF-1 and increased the amount of in vitro translated SRF. In vitro translated SRF did not bind to the probe, whereas TEF-1 did (Fig. 6C, lanes 1, 2, and 6). When increasing amounts of SRF were added to the reaction, an increase in the TEF-1·MCAT complex was observed (Fig. 6C, lanes 3-5). Together, these results suggest that, at least in vitro, SRF can help TEF-1 bind to its sequence or prevent its dissociation from its binding site.
|
-Tm Gene Promoter in Differentiated Smooth Muscle Cells—Unlike skeletal or cardiac muscle cells that are terminally differentiated, smooth muscle cells retain a remarkable plasticity and can undergo reversible changes in phenotype in response to changes in local environment cues, such as exposure to growth factors (61). These changes in smooth muscle cell phenotype are referred to as phenotypic modulation. For example, after injury, mature cells exhibiting a contractile phenotype switch to a synthetic state characterized by an increase in the rate of proliferation, migration, and synthesis of the extracellular matrix. We have shown previously that, when cultured in the absence of serum, U8A4 cells retain a differentiated phenotype and express several smooth muscle markers, including the smooth muscle-specific -Tm isoform (40). To determine whether expression of the -Tm gene promoter is sensitive to the phenotypic modulation of smooth muscle cells, we transfected the pGL284LUC DNA construct, either wild-type or mutated at the MCAT sequence (pGLMCATLUC), in U8A4 cells cultured in the presence or absence of 10% serum. In these experiments, the activity of the pGL3-Control vector was set at 100%. As expected, the pGL284LUC construct showed a high level of expression in U8A4 cells when cultured in the absence of serum (Fig. 7A). However, when the cells were cultured in 10% serum, the pGL284LUC construct showed a 70% decrease in activity. pGLMCATLUC showed a very low level of activity whether the cells were cultured in the presence or absence of serum. Thus, the TEF-1 transcription-promoting activity of the -Tm gene is selectively increased in serum-deprived myocytes. We also observed a similar increase in activation of the pGL284LUC DNA construct in the PAC1 smooth muscle cell line when cultured in low serum conditions (1% serum) compared with high serum conditions (10% serum) (data not shown). Together with the ChIP data, these results demonstrate the absolute requirement for binding of TEF-1 to the MCAT sequence for transcriptional activity of the -Tm gene in smooth muscle cells.
The Subcellular Localization of TEF-1 in Smooth Muscle Cells Is Dependent on Serum—It has been demonstrated that SRF-dependent gene expression in smooth muscle cells can be regulated through nuclear translocation of SRF (62, 63). To test whether cytoplasmic redistribution of TEF-1 could account for the diminished TEF-1 transcription-promoting activities observed when cells were cultured in high serum, we analyzed TEF-1 localization by immunofluorescence staining in U8A4 cells cultured in the presence or absence of serum. As a control experiment, we also analyzed the cellular distribution of SRF under the same conditions. TEF-1 immunoreactivity was restricted to the nuclei of U8A4 cells cultured in the absence of serum (Fig. 7C, panel a), whereas in serum-fed cells, TEF-1 was partially redistributed to the cytoplasm (panel c). In contrast and as reported previously (62), SRF protein appeared to be strictly localized to the cytoplasm of cells in the absence of serum, but was found in the nuclei of cells cultured in the presence of serum (Fig. 7C, panels e and g). We observed similar results using PAC1 smooth muscle cells (data not shown). To determine whether TEF-1 content is modified in cells in response to culture conditions, we performed Western blot analysis. As shown in Fig. 7B, TEF-1 protein content was not modified in U8A4 smooth muscle cells cultured in the presence and absence of serum. As a control, there was no variation of the ubiquitously expressed Tm proteins. Together, these data indicate that the full activity of the -Tm gene in smooth muscle cells does not correlate with new synthesis of TEF-1 in the cytoplasm, but with TEF-1 nuclear localization.
|
-Tm Gene in Embryonic Cells—We next investigated whether TEF-1 can activate the -Tm gene in Xenopus embryonic cells. For this purpose, we took advantage of the animal cap assay, which has been widely used for testing gene activation by transcription factors. For instance, we have shown that, in this model, the skeletal muscle myosin light chain-1/3 and -Tm genes are induced by the MyoD family of transcription factors (10, 52). As shown in Fig. 8A, animal cap explants from TEF-1 mRNA-microinjected embryos showed a higher level of expression of muscle -Tm mRNA compared with uninjected control embryos. This suggests that TEF-1 can activate the endogenous -Tm gene in embryonic cells. To determine whether TEF-1 could activate the pGL284LUC DNA, we injected two-cell stage embryos with the DNA construct together with TEF-1 mRNA and then measured the luciferase activity in gastrula stage embryos. Embryos injected with TEF-1 mRNA showed a 3.5-fold increase in luciferase activity compared with embryos injected with the pGL284LUC DNA alone (Fig. 8B). The luciferase activity observed in embryos injected with pGLMCATLUC DNA was equivalent to the activity resulting from injection of the wild-type pGL284LUC DNA. Together, these data suggest that TEF-1 can activate the -Tm gene in embryonic cells through the MCAT site found in the minimal promoter.
The MCAT Sequence Is Required for Correct Expression of the -Tm Transgene in Vivo—We have shown previously, by in situ hybridization and RT-PCR analyses, that the -Tm gene is expressed in the somites, the heart, and later in the smooth muscle tissues of the embryo (10). Whole mount in situ hybridization with a striated muscle -Tm probe confirmed expression of the gene in the somites of late neurula stage and tadpole embryos (Fig. 9A, panels a-c) and in the tad pole embryonic heart (panel c).
Because a relatively small upstream region of the X. laevis -Tm gene promoter is sufficient to direct high levels of cardiac, skeletal, and smooth muscle-specific gene expression in vitro, we investigated whether this same region functions in an equivalent manner in the in vivo environment of the X. laevis embryo and juvenile. In preliminary transient expression experiments using microinjection of DNA luciferase constructs, maximal reporter gene expression was observed in embryos injected with the pGL284LUC construct. In these experiments, transcription of the transgene was correctly initiated (data not shown).
We then used a permanent transgenesis assay (54, 55) with a GFP reporter gene placed under the control of the 284-bp promoter. The pGL284GFP gene was expressed in the embryo in a correct temporal and spatial pattern (Fig. 9B, panels a, b, e, and f). Expression of the transgene was not detected at the early neurula stage (Fig. 9A, panel b), but was first detected at the late neurula stage within the somites (data not shown). Later during development, GFP expression became obvious within the somites and embryonic heart at the tadpole stage (Fig. 9B, panel f). Therefore, the 284-bp promoter can recapitulate expression of the -Tm endogenous gene in somites and embryonic heart (Fig. 9A, panels a-c). To determine whether the MCAT-binding site located in the 284-bp promoter is required for correct expression of the reporter gene in transgenic embryos, we made a pGLMCATGFP construct in which the MCAT sequence was mutated and showed no binding of in vitro translated TEF-1 (see Fig. 4A). The resulting pGLMCATGFP transgene was expressed in the early neurula stage embryo, with no restriction in cell specificity (Fig. 9C, panel d) (data not shown). Later during development, the transgene was expressed in muscle and non-muscle cells of the embryo (panel h), a pattern that resembles the expression seen with transgenes driven by a ubiquitous promoter. Therefore, mutation of the MCAT sequence and the resulting impairment of TEF-1 binding led to a temporal and spatial deregulation of transgene expression. It is difficult to detect live GFP expression in smooth muscle tissues of the embryo because of the autofluorescence of these tissues. To determine whether the GFP transgene was expressed in smooth muscle tissues, we let the transgenic embryos develop further and undertook RT-PCR analysis of juvenile tissues. We analyzed the GFP mRNA content of striated muscles (skeletal and heart), smooth muscle (stomach and intestine), and non-muscle liver tissue. In parallel, we analyzed the levels of striated muscle and smooth muscle -Tm mRNA isoforms (10). As shown in Fig. 9D, pGL284GFP gene expression was restricted to all muscle tissues, but was absent in liver. In contrast, the pGLMCATGFP transgene, containing a mutated version of the MCAT sequence, was expressed in all tissues, albeit at a lower level. Together, these data suggest that, in vivo, an intact MCAT sequence is critical for a correct pattern of expression of the -Tm gene in the three muscle lineages.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
-Tm gene in different muscle cell types. Our previous work has shown that, like its mammalian orthologs, the X. laevis -Tm gene is expressed in skeletal, cardiac, and smooth muscle cells during development and in the adult (6, 10). As yet, there have been no studies describing the elements involved in activation of the gene in these different muscle types. To characterize such sequences, we have used, in a first approach, an in vitro cell culture system with skeletal, cardiac, and smooth muscle cells. From deletion analysis, we determined that a minimal promoter containing the first 284 bp upstream from the transcription start site is sufficient for maximal expression in all muscle cells, but is inactive in myoblasts and fibroblasts. We found that this promoter contains several known cis-elements that have been implicated in the regulation of numerous muscle genes, i.e. four E box motifs, a GC-rich region, an A/T-rich region, a CArG-like element, and one inverted MCAT sequence. Mutation analysis revealed that all these cis-sequences, with the exception of E box-4, mediate positive regulation of the -Tm promoter. The decrease in promoter activity resulting from the different mutations is dependent on the muscle cell type. Regarding E box sequences, we observed that E box-1 has a more pronounced positive effect in skeletal muscle, whereas E box-2 and E box-3, although having identical core sequences, have positive effects in skeletal and smooth muscle cells, respectively. These data suggest, as has been already established, the presence in muscle cells of distinct basic helix-loop-helix factors that can bind E box elements and the combinatorial interaction between those factors and more generally expressed factors (15). It has been previously reported that expression of the rat smooth muscle -actin promoter can be directed by E box elements through distinct basic helix-loop-helix factors in skeletal versus smooth muscle cells (64). Similarly, we may hypothesize that distinct basic helix-loop-helix factors could participate in activation of the gene in the different muscle lineages. Accordingly, when performing gel shift analysis with the distinct E box probes, we observed variations in the efficiency of competition between the different E box sequences, suggesting the existence between muscle tissues of proteins with different binding affinity (data not shown). The A/T-rich sequence of the 284-bp -Tm promoter resembles, although not completely, a myocyte enhancer factor-2 consensus site; and when mutated in the context of the promoter, a significant reduction of promoter activity was observed only in skeletal muscle. In contrast, mutation of the GC box impaired transcriptional activity in the three muscle cell types with a similar efficiency. The GC box constitutes a DNA motif present in the promoter of a very large number of genes that is recognized by members of the Sp1 family. Toutant et al. (13) have characterized an enhancer element in the chicken -Tm gene that contains a stretch of 7 Cs whose mutation results in a decrease in myotube-specific transcriptional activity. The GC sequence of the frog -Tm promoter also contains 7 Cs and could represent similar target sequence for Sp1 family members. Our in vitro analysis indicated that muscle-restricted activity of the -Tm gene is regulated by multiple positive regulatory elements that are active in different muscle cell types.
|
-Tm gene in muscle cells, and additional lines of evidence support this. Originally described as an essential element for the promoter activity of the chicken TnT gene in skeletal muscle, the MCAT element (5'-CATTCCT-3') has now been firmly established as an important regulatory element for the activation of several skeletal and cardiac muscle genes (18). The MCAT sequence has also been shown to be involved in regulation of the mammalian smooth muscle -actin gene (37, 38). Although the MCAT element of the -Tm promoter is in an inverse orientation (5'-AGGAATG-3'), it is present in all vertebrate -Tm gene promoters in this same orientation and at a conserved position relative to the transcription initiation site. Moreover, the MCAT sequence is embedded in a highly conserved 30-bp sequence that shows only 3 divergent nucleotides in all vertebrate promoters. Outside this region, there is no obvious sequence homology, suggesting a conserved role of this region in distant species.
We have therefore focused our work on the MCAT sequence because of its conservation and its critical involvement in the activity of the promoter in the three muscle types. The results from EMSA analysis demonstrated that the MCAT sequence of the -Tm promoter was bound by TEF-1 protein present in nuclear extracts of cardiac, skeletal, and smooth muscle cells. Wit
