Expression of the SM22alpha promoter in transgenic mice provides evidence for distinct transcriptional regulatory programs in vascular and visceral smooth muscle cells.

SM22alpha is a putative calcium-binding protein that is expressed in cardiac, smooth, and skeletal muscle lineages during mouse embryogenesis and in adult smooth muscle cells (SMC). To define the mechanisms that regulate smooth muscle-specific gene transcription, we isolated the SM22alpha gene and analyzed its 5'-flanking region for elements that direct smooth muscle expression in transgenic mice. Using a series of promoter deletions, a region of the SM22alpha promoter containing 445 base pairs of 5'-flanking sequence was found to be sufficient to direct the specific expression of a lacZ transgene in mouse embryos in the vascular smooth, cardiac, and skeletal muscle lineages in a temporospatial pattern similar to the endogenous SM22alpha gene. However, in contrast to the endogenous gene, transgene expression was not detected in venous, nor visceral SMCs. This SM22alpha-lacZ transgene was therefore able to distinguish between the transcriptional regulatory programs that control gene expression in vascular and visceral SMCs and revealed heretofore unrecognized differences between these SMC types. These results suggest that distinct transcriptional regulation programs control muscle gene expression in vascular and visceral SMCs.

A Mef2 gene that generates a muscle-specific isoform via alternative mRNA splicing.

Members of the myocyte-specific enhancer-binding factor 2 (MEF2) family of transcription factors bind a conserved A/T-rich sequence in the control regions of numerous muscle-specific genes. Mammalian MEF2 proteins have been shown previously to be encoded by three genes, Mef2, xMef2, and Mef2c, each of which gives rise to multiple alternatively spliced transcripts. We describe the cloning of a new member of the MEF2 family from mice, termed MEF2D, which shares extensive homology with other MEF2 proteins but is the product of a separate gene. MEF2D binds to and activates transcription through the MEF2 site and forms heterodimers with other members of the MEF2 family. Deletion mutations show that the carboxyl terminus of MEF2D is required for efficient transactivation. MEF2D transcripts are widely expressed, but alternative splicing of MEF2D transcripts gives rise to a muscle-specific isoform which is induced during myoblast differentiation. The mouse Mef2, Mef2c, and Mef2d genes map to chromosomes 7, 13, and 3, respectively. The complexity of the MEF2 family of regulatory proteins provides the potential for fine-tuning of transcriptional responses as a consequence of combinatorial interactions among multiple MEF2 isoforms encoded by the four Mef2 genes.

Retinoic acid-induced tissue transglutaminase and apoptosis in vascular smooth muscle cells.

Retinoids exert antiproliferative and prodifferentiating effects in vascular smooth muscle cells (SMCs) and reduce neointimal mass in balloon-injured blood vessels. The mechanisms through which retinoids carry out these effects are unknown but likely involve retinoid receptor-mediated changes in gene expression. Here we report the cloning, chromosomal mapping, and biological activity of the retinoid-response gene rat tissue transglutaminase (tTG). Northern blotting studies showed that tTG is rapidly and dose-dependently induced in a protein synthesis-independent manner after stimulation with the natural retinoid all-trans retinoic acid (atRA). The induction of tTG was selective for atRA and its stereoisomers 9-cis and 13-cis RA, because little or no elevation in mRNA expression was observed with a panel of growth factors. Western blotting and immunofluorescence confocal microscopy showed an accumulation of cytosolic tTG protein after atRA stimulation. Radiolabeled cross-linking studies revealed a corresponding elevation in in vitro tTG activity. The increase in tTG activity was reduced in the presence of 2 distinct inhibitors of tTG (monodansylcadaverine and cystamine). atRA-induced tTG mRNA and protein expression were followed by a significant elevation in SMC apoptosis. Such retinoid-induced programmed cell death could be partially inhibited with each tTG inhibitor and was completely blocked when both inhibitors were used simultaneously. These results establish a role for atRA in the sequential stimulation of tTG and apoptosis in cultured SMCs. atRA-mediated apoptosis in SMCs seems to require the participation of active tTG, suggesting a potential mechanistic link between this retinoid-inducible gene and programmed cell death.

The A10 cell line: a model for neonatal, neointimal, or differentiated vascular smooth muscle cells?

OBJECTIVES: The A10 cell line was derived from the thoracic aorta of embryonic rat and is a commonly used model of vascular smooth muscle cells (VSMC). Despite its wide use this cell line has not been well characterized. This is especially important in light of recent evidence of phenotypically distinct cell populations isolated from rat vascular tissue. Therefore, the present study was undertaken to confirm the VSMC nature of A10 cells and to investigate whether these cells particularly resemble adult, neonatal, or neointimal rat VSMC. METHODS: A variety of defining characteristics were used that included immunofluorescent analysis for smooth muscle alpha-actin, smooth and non-muscle myosin heavy chains, desmin and vimentin; Western analysis for smooth muscle and non-muscle myosin heavy chains; mRNA analysis for smooth muscle myosin heavy chain, calponin, SM22 alpha, tropoelastin and PDGF-B peptide; and functional assays of cell migration, proliferation and agonist induced intracellular Ca transients. RESULTS: A10 cells expressed smooth muscle alpha-actin, SM22 alpha, smooth muscle calponin and vimentin, characteristic of in vivo rat VSMCs; however they also resembled de-differentiated smooth muscle cells in that they expressed non-muscle myosin rather than smooth muscle myosin heavy chain. A10 cells resembled cultured rat neonatal smooth muscle cells ("pup cells") in that they had an epithelioid shape and lacked functional PDGF-alpha receptors: however they did not express PDGF-B mRNA or proliferate in low serum containing medium as do neonatal cells. A10 cells had several characteristics in common with neointimal cells including the expression of alpha-actin, vimentin, and non-muscle myosin and the lack of expression of PDGF-B mRNA as well as the ability to migrate in response to PDGF-BB. CONCLUSION: In conclusion, A10 cells are nondifferentiated VSMC that differ from neonatal but bear significant resemblance to neointimal cells.

Loss of serum response factor induces microRNA-mediated apoptosis in intestinal smooth muscle cells.

Serum response factor (SRF) is a transcription factor known to mediate phenotypic plasticity in smooth muscle cells (SMCs). Despite the critical role of this protein in mediating intestinal injury response, little is known about the mechanism through which SRF alters SMC behavior. Here, we provide compelling evidence for the involvement of SRF-dependent microRNAs (miRNAs) in the regulation of SMC apoptosis. We generated SMC-restricted Srf inducible knockout (KO) mice and observed both severe degeneration of SMCs and a significant decrease in the expression of apoptosis-associated miRNAs. The absence of these miRNAs was associated with overexpression of apoptotic proteins, and we observed a high level of SMC death and myopathy in the intestinal muscle layers. These data provide a compelling new model that implicates SMC degeneration via anti-apoptotic miRNA deficiency caused by lack of SRF in gastrointestinal motility disorders.

Expression, genomic structure and high resolution mapping to 19p13.2 of the human smooth muscle cell calponin gene.

Smooth muscle cells (SMC) express a battery of cell-restricted differentiation genes, many of which are down-regulated during the course of vascular disease. Here, we present the mRNA expression, genomic structure and chromosomal mapping of the gene encoding human smooth muscle cell calponin (SMCC). Human SMCC transcripts are restricted to tissues and cells of SMC origin and, in the latter case, appear to be uniquely controlled in two distinct human SMC lines of uterine and aortic origin. Restriction mapping. Southern blot and PCR analysis of a 70-kb human bacterial artificial chromosome (BAC) revealed a genomic structure (seven exons spanning > 11 kb) very similar to that reported for the mouse SMCC gene. Using a variety of human-rodent somatic cell hybrid and radiation hybrid mapping panels, the human SMCC gene was mapped to a genomic interval of less than 1.32 Mb in 19p13.2. These results provide new information concerning the regulation of SMCC gene expression and demonstrate the utility of two human SMC lines for the further characterization of this gene's expression control. The identification of a BAC harboring the entire human SMCC locus represents an important reagent for future analysis of SMCC regulatory sequences. Finally, the localization of SMCC to a defined genomic interval will facilitate an analysis of its potential as a candidate gene for disease phenotypes mapping to 19p13.2.

Radiation Hybrid (RH) Mapping of Human Smooth Muscle-Restricted Genes.

Recent molecular genetic studies in cardiac and skeletal muscle have revealed mutations in a battery of sarcomeric muscle-restricted genes that appear to be associated with various myopathies (1,2). In sharp contrast, no mutations in smooth muscle cell (SMC)-restricted genes have been linked to a SMC disease phenotype, although a review of the literature indicates that many SMC diseases with a presumed genetic basis are present in human populations (3-13). An important first step in linking a disease phenotype to a mutation within a specific gene is the accurate physical mapping of the candidate gene to a specific chromosomal region within the context of other genetic markers, such as highly polymorphic microsatellite markers now routinely used for recombination-based linkage analysis of families segregating a particular disease phenotype. Several methods exist for the physical mapping of genes, including fluorescent in situ hybridization (FISH) (14) and interspecific mouse back-crossing (15). FISH analysis is relatively fast, but often requires large genomic clones and does not afford the high-resolution mapping required to link a gene locus to a disease phenotype. Interspecific mouse back-crossing can be quite powerful with respect to resolution, but studies are necessarily limited to the mouse genome. Thus, a broadly applicable, fast and simple method of gene mapping would be desirable to aid investigators in localizing potential candidate disease genes, especially those pertaining to SMC-associated diseases.

A comparative molecular analysis of four rat smooth muscle cell lines.

Transcriptional regulation of smooth muscle cell (SMC) differentiation is a rapidly growing area of interest that has relevance for understanding intimal disease. Despite the wealth of data accumulating in vitro, however, no study has compared the cell-specific marker profile, transfectability, promoter activity, and growth characteristics among several SMC culture systems. Accordingly, we performed a comprehensive analysis of the marker profile, growth properties, transfectability, and SMC promoter activity in four rat SMC lines (A7r5, adult and pup aortic, and PAC1). Despite alterations in chromosomal number and structure, A7r5, adult aortic, and PAC1 cells express all SMC markers studied including SM alpha-actin, SM calponin, SM22, tropoelastin, and to a lesser extent, SM myosin heavy chain (SMMHC). In contrast, pup aortic cells express very low or undetectable levels of all the above markers except tropoelastin. Adult aortic, pup, and PAC1 cells display similar growth curves and levels of proto-oncogene transcripts, whereas those in the A7r5 line are comparatively less. All cell lines studied except pup cells show expression of SMC differentiation genes during active growth, indicating that growth and differentiation are not mutually exclusive in cultured smooth muscle. Transfection studies reveal dramatic differences in DNA uptake and SMC-restricted promoter activity between cell lines. Collectively, these results provide detailed information relating to SMC molecular biology in culture that should facilitate the selection of a cell line for studying the transcriptional regulatory mechanisms underlying SMC differentiation.

Expression of the smooth muscle cell calponin gene marks the early cardiac and smooth muscle cell lineages during mouse embryogenesis.

Although several genes are considered markers for vascular smooth muscle cell (SMC) differentiation, few have been rigorously tested for SMC specificity in mammals, particularly during development where considerable overlap exists between different muscle gene programs. Here we describe the temporospatial expression pattern of the SMC calponin gene (formerly h1 or basic calponin) during mouse embryogenesis and in adult mouse tissues and cell lines. Whereas SMC calponin mRNA expression is restricted exclusively to SMCs in adult tissues, during early embryogenesis, SMC calponin transcripts are expressed throughout the developing cardiac tube as well as in differentiating SMCs. Transcription of the SMC calponin gene initiates at two closely juxtaposed sites in the absence of a consensus TATAA or initiator element. Transient transfection assays in cultured SMC demonstrated that high level SMC calponin promoter activity required no more than 549 nucleotides of 5 sequence. In contrast to the strict cell type-specificity of SMC calponin mRNA expression, the SMC calponin promoter showed activity in several cell lines that do not express the endogenous SMC calponin gene. These results demonstrate that SMC calponin responds to cardiac and smooth muscle gene regulatory programs and suggest that the cardiac and smooth muscle cell lineages may share a common gene regulatory program early in embryogenesis, which diverges as the heart matures. The finding that the isolated SMC calponin promoter is active in a wider range of cells than the endogenous SMC calponin gene also suggests that long-range repression or higher order regulatory mechanism(s) are involved in cell-specific regulation of SMC calponin expression.

A novel retinoid-response gene set in vascular smooth muscle cells.

A modified suppression subtractive hybridization assay was performed to uncover genes induced by all-trans retinoic acid in cultured smooth muscle cells (SMC). Northern blotting studies confirmed the induction of 14 genes, many of which have heretofore been unrecognized as retinoid-inducible. Temporal expression and cycloheximide studies allowed us to categorize these genes as either immediate-early (LOX-1, endolyn, Stoned B/TFIIA alpha/beta-like factor, Src Suppressed C Kinase Substrate, and tissue transglutaminase) or delayed (cathepsin-L, ceruloplasmin, epithelin, importin alpha, alpha(8)-integrin, lactate dehydrogenase B, retinol dehydrogenase, spermidine/spermine N(1)-acetyltransferase, and VCAM-1) retinoid-response genes. A survey of rat tissues showed two of the genes (tissue transglutaminase and alpha(8)-integrin) to be highly restricted to vascular tissue. In situ hybridization verified expression of both tissue transglutaminase and alpha(8)-integrin to SMC in balloon-injured rat carotid artery. These findings unveil a new retinoid-response gene set that should be exploited to define molecular pathways involved in the antagonistic effects of retinoids on SMC growth and neointimal formation.

Expression and chromosomal mapping of the mouse smooth muscle calponin gene.

Smooth muscle calponin (Cnn1) is a multifunctional protein whose expression is tightly restricted to differentiated smooth muscle cell (SMC) lineages during embryonic and post-natal life. As such, Cnn1 represents an ideal locus from which to dissect out regulatory elements that control its expression and hence the mature SMC phenotype. Previous work has focused on the expression and chromosomal mapping of the rat and human Cnn1 orthologs. In this report, we describe a unique pattern of Cnn1 expression during the growth and differentiation of BC3H1 cells, a mouse cell line that has transcriptional characteristics of both smooth and skeletal muscle lineages. Actively growing BC3H1 cells exhibit Cnn1 mRNA expression, which is extinguished when these cells are induced to differentiate upon serum withdrawal. Replating differentiated BC3H1 cells restores steady-state Cnn1 mRNA levels. The down-regulation of Cnn1 mRNA during BC3H1 differentiation coincides with the induction of myogenin, a skeletal muscle transcription factor that is not present in SMC lineages. Results from cycloheximide and actinomycin D studies suggest the existence of a labile repressor protein(s) that destabilizes the pool of Cnn1 mRNA and/or silences transcription of the Cnn1 locus. Mapping of the mouse Cnn1 locus to Chr 9, which is homologous to human Cnn1 on 19p13.2 and rat Cnn1 on 8q, suggests no gross rearrangement of this locus in the BC3H1 cell line. These results are the first to show reversible expression of Cnn1 and demonstrate the utility of the BC3H1 muscle cell line as a model system for the further characterization of Cnn1 gene regulation.

Early proto-oncogene expression in rat aortic smooth muscle cells following endothelial removal.

To study the mechanism(s) of vascular smooth muscle cell proliferation in vivo, mRNA levels of c-fos, c-jun, and c-myc were determined by Northern blot analysis following vascular balloon de-endothelialization (BDE). Medial smooth muscle cells (SMC) were separated and studied by enzymatic digestion of the vessel wall. mRNA levels of c-fos and c-jun from aortic smooth muscle cells (SMC) were simultaneously induced within 30 minutes of BDE and declined to baseline by 1.5 hours, c-myc mRNA did not begin to increase until 1 hour after vascular injury. Levels of c-myc peaked at 2 hours and were sustained for an additional 4 hours before gradually declining. Smooth muscle cells derived from enzyme-treated control aortae that did not undergo BDE expressed c-fos and c-jun, but showed no evidence of c-myc message. In contrast, nonenzymatically treated, non-BDE whole aortae (containing both media and adventitia) demonstrated a prominent c-myc signal, but failed to express c-fos and c-jun. Corresponding examination of adventitia derived from enzyme-treated aortae showed this tissue to be a source of all three proto-oncogenes. The results of this study demonstrate the earliest in vivo molecular markers of vascular injury reported to date and implicate SMC proto-oncogene expression in the initiation of SMC proliferation. Furthermore these findings suggest two avenues for proto-oncogene induction, that are due to (1) vessel wall manipulation and (2) humoral stimulation.

EVEC, a novel epidermal growth factor-like repeat-containing protein upregulated in embryonic and diseased adult vasculature.

A hallmark of vascular lesions is the phenotypic modulation of vascular smooth muscle cells (VSMCs) from a quiescent, contractile state to a more primitive, proliferative phenotype with a more fetal pattern of gene expression. Using subtraction hybridization to identify genes that may regulate this transition, we cloned a novel gene named EVEC, an acronym for its expression in the embryonic vasculature and the presence of Ca2+ binding epidermal growth factor-like repeats contained in the predicted protein structure. Although these repeats are characteristic of the extracellular matrix proteins, fibrillin, fibulin, and the latent transforming growth factor-beta binding proteins, EVEC most closely resembles the H411 and T16/S1-5 gene products, the latter of which are believed to regulate DNA synthesis in quiescent fibroblasts. Using in situ hybridization, we demonstrated that EVEC is expressed predominantly in the VSMCs of developing arteries in E11.5 through E16.5 mouse embryos. Lower levels of expression are also observed in endothelial cells, perichondrium, intestine, and mesenchyme of the face and kidney. EVEC mRNA expression is dramatically downregulated in adult arteries, except in the uterus, where cyclic angiogenesis continues; however, EVEC expression is reactivated in 2 independent rodent models of vascular injury. EVEC mRNA is observed in cellular elements of atherosclerotic plaques of LDL receptor-deficient, human apolipoprotein B transgenic mice and in VSMCs of the media and neointima of balloon-injured rat carotid arteries. These data suggest that EVEC may play an important role in the regulation of vascular growth and maturation during development and in lesions of injured vessels.

Localized adenovirus-mediated gene transfer into vascular smooth muscle in the hamster cheek pouch.

OBJECTIVE: Our purpose was to develop a method for adenovirus delivery to the hamster cheek pouch to experimentally target gene transfer in tissue used for microvascular studies. METHODS: Separate constructs were tested with transgenes for lacZ or green fluorescent protein (GFP) driven by three promoters: RSV, CMV, and SM22. With university approval, adenovirus was delivered in anesthetized (pentobarbital, 70 mg/kg) hamsters (n = 28) by using either a vascular systemic injection or tissue infiltration (interstitial space behind the pouch). During 3 days, animals receiving infiltration gained the expected weight, whereas those receiving vascular injection lost weight; no other behavior changes were noted. RESULTS: On day 3 postadenoviral delivery (infiltration), expression of lacZ (histology, beta-galactosidase) or GFP (fluorescence microscopy) was confirmed across the tissue (CMV and RSV promoters) and exclusively in vascular smooth muscle cells (specific SM22 promoter), without evidence of tissue inflammation. In vitro microvascular experiments verified normal responses in the cheek pouch of day 3 postadenoviral delivery animals. We tested local dilation to methacholine, adenosine, remote dilation to methacholine, adenosine, nitroprusside, and LM609 (alpha(v)beta3 integrin agonist), flow-dependent dilation, and flow recruitment. CONCLUSIONS: Thus, this method enables targeted, cell-specific gene transfer to one tissue important for microvascular studies, without significant systemic exposure and without adverse inflammation.

Cloning of a novel retinoid-inducible serine carboxypeptidase from vascular smooth muscle cells.

Retinoids block smooth muscle cell (SMC) proliferation and attenuate neointimal formation after vascular injury, presumably through retinoid receptor-mediated changes in gene expression. To identify target genes in SMC whose encoded proteins could contribute to such favorable biological effects, we performed a subtractive screen for retinoid-inducible genes in cultured SMC. Here, we report on the cloning and initial characterization of a novel retinoid-inducible serine carboxypeptidase (RISC). Expression of RISC is low in cultured SMC but progressively increases over a 5-day time-course treatment with all-trans-retinoic acid. A near full-length rat RISC cDNA was cloned and found to have a 452-amino acid open reading frame containing an amino-terminal signal sequence, followed by several conserved domains comprising the catalytic triad common to members of the serine carboxypeptidase family. In vitro transcription and translation experiments showed that the rat RISC cDNA generates an approximately 51-kDa protein. Confocal immunofluorescence microscopy of COS-7 cells transiently transfected with a RISC-His tag plasmid revealed cytosolic localization of the fusion protein. Western blotting studies using conditioned medium from transfected COS-7 cells suggest that RISC is a secreted protein. Tissue Northern blotting studies demonstrated robust expression of RISC in rat aorta, bladder, and kidney with much lower levels in all other tissues analyzed; high level RISC expression was also observed in human kidney. In situ hybridization verified the localization of RISC to medial SMC of the adult rat aorta. Interestingly, expression in kidney was restricted to proximal convoluted tubules; little or no expression was observed in glomerular cells, distal convoluted and collecting tubules, or medullary cells. Radiation hybrid mapping studies placed the rat RISC locus on chromosome 10q. These studies reveal a novel retinoid-inducible protease whose activity may be involved in vascular wall and kidney homeostasis.

Smooth muscle cell immediate-early gene and growth factor activation follows vascular injury. A putative in vivo mechanism for autocrine growth.

To understand the molecular events governing smooth muscle cell (SMC) proliferation in vivo, immediate-early gene (IEG) expression was assessed and related to growth factor ligand and receptor mRNA and SMC DNA synthesis after aortic injury. Balloon catheter injury evoked increases in SMC c-myc and thrombospondin (tsp) within 2 hours. The induction of these IEGs was followed by elevated transcripts to platelet-derived growth factor-A (PDGF-A), transforming growth factor-beta 1 (TGF-beta 1) and a basic fibroblast growth factor (bFGF) receptor. Whereas PDGF type-beta receptor mRNA was demonstrated in SMCs from control and balloon-injured aortas, no detectable signal was observed for the PDGF type-alpha receptor. To explore the potential linkage between IEG products and growth factor mRNA expression, cycloheximide was employed to block early protein synthesis after balloon injury. Induction of PDGF-A and TGF-beta 1 was attenuated by cycloheximide, but bFGF induction was unaffected. Moreover, cycloheximide superinduced IEGs and revealed PDGF-B transcripts, which were otherwise undetected. Seven days after aortic injury, a spontaneous increase in c-myc and tsp mRNA was noted. This IEG reactivation was followed 12 hours later by a twofold increase in SMC DNA synthesis. These findings corroborate an autocrine mode of SMC proliferation in vivo and suggest the IEG products may control such growth by stimulating growth factor genes.

Localization of Fos and Jun proteins in rat aortic smooth muscle cells after vascular injury.

The availability of specific reagents to measure gene activity has provided important tools and potential new directions for the study of smooth muscle cell (SMC) proliferation in vivo. In this report, we have measured steady-state mRNA levels of several fos and jun family members in aortic tissue by Northern blotting after vascular injury. In addition, protein products of these genes were analyzed by immunocytochemistry. Within 15 minutes of balloon injury, mRNA levels of c-fos, fosB, c-jun, junB, and junD were elevated severalfold. In contrast, fos-related antigen (fra-1) mRNA showed a delayed onset of expression. The expression kinetics of these immediate early genes was similar to those in cultured cells stimulated to undergo proliferation by growth factors, suggesting that such SMC gene activation in vivo reflects permeation of blood-derived growth factors into the vessel wall or intravascular release of preformed growth factors. Translation of fos and jun genes into immunoreactive products was demonstrated 2 hours after balloon injury with antisera to Fos and Jun proteins. Treating rats with cycloheximide abolished this immunoreactivity. The distribution of Fos and Jun products was concentrated in SMC nuclei at the luminal border of the rat aorta. Such focal expression may have consequences for the initiation of SMC DNA synthesis and migration after vascular injury. Furthermore, the expression of Fos and Jun proteins in SMC after vascular balloon injury may be used as an index of SMC activation under a variety of experimental settings.

SM22 alpha, a marker of adult smooth muscle, is expressed in multiple myogenic lineages during embryogenesis.

SM22 alpha is a calponin-related protein that is expressed specifically in adult smooth muscle. To begin to define the mechanisms that regulate the establishment of the smooth muscle lineage, we analyzed the expression pattern of the SM22 alpha gene during mouse embryogenesis. In situ hybridization demonstrated that SM22 alpha transcripts were first expressed in vascular smooth muscle cells at about embryonic day (E) 9.5 and thereafter continued to be expressed in all smooth muscle cells into adulthood. In contrast to its smooth muscle specificity in adult tissues, SM22 alpha was expressed transiently in the heart between E8.0 and E12.5 and in skeletal muscle cells in the myotomal compartment of the somites between E9.5 and E12.5. The expression of SM22 alpha in smooth muscle cells, as well as early cardiac and skeletal muscle cells, suggests that there may be commonalities between the regulatory programs that direct muscle-specific gene expression in these three myogenic cell types.