In vitro system for differentiating pluripotent neural crest cells into smooth muscle cells.

The change in vascular smooth muscle cells (SMC) from a differentiated to a dedifferentiated state is the critical phenotypic response that promotes occlusive arteriosclerotic disease. Despite its importance, research into molecular mechanisms regulating smooth muscle differentiation has been hindered by the lack of an in vitro cell differentiation system. We identified culture conditions that promote efficient differentiation of Monc-1 pluripotent neural crest cells into SMC. Exclusive Monc-1 to SMC differentiation was indicated by cellular morphology and time-dependent induction of the SMC markers smooth muscle alpha-actin, smooth muscle myosin heavy chain, calponin, SM22alpha, and APEG-1. The activity of the SM22alpha promoter was low in Monc-1 cells. Differentiation of these cells into SMC caused a 20-30-fold increase in the activity of the wild-type SM22alpha promoter and that of a hybrid promoter containing three copies of the CArG element. By gel mobility shift analysis, we identified new DNA-protein complexes in nuclear extracts prepared from differentiated Monc-1 cells. One of the new complexes contained serum response factor. This Monc-1 to SMC model should facilitate the identification of nodal regulators of smooth muscle development and differentiation.

Arteriosclerosis and its complications (heart attack and stroke) are the major causes of death in developed and developing countries (1), and the change in vascular smooth muscle cells (SMC) 1 from a differentiated to a dedifferentiated state is the critical phenotypic response that promotes occlusive arteriosclerotic disease (1)(2)(3). Despite its importance, we know little about the genes that regulate SMC differentiation (2). This is in contrast with our more thorough understanding of transcription factors involved in the differentiation of skeletal muscle cells (4 -7). A major reason for the rapid progress in our understanding of skeletal as opposed to smooth muscle biology is the availability of an in vitro system for studying differentiation of skeletal muscle cells. The ability to induce myogenic differentiation by the use of demethylating agents and serum starvation in 10T1/2 cells and C2C12 cells, respectively, has been critical to the study of skeletal muscle development and differentiation (8,9). Indeed, the establishment of immortalized myogenic cell lines may have been the single most important development in the field of skeletal myogenesis (10). Until the neural crest cell to SMC differentiation system described here, a way of rapidly and uniformly inducing precursor cells to differentiate into SMC had been lacking.
Pluripotent neural crest cells can differentiate into neurons, glia, chondrocytes, melanocytes, and SMC (11)(12)(13). Arterial SMC of the chick ascending and thoracic aorta are of a neural crest origin (14 -16), and various members of the transforming growth factor-␤ superfamily can instructively promote differentiation of primary cultured neural crest cells into neuronal cells or SMC (12). Unfortunately, our ability to work with neural crest cells in primary culture has been limited by the difficulty of obtaining quantities sufficient for biochemical and genetic analysis. This problem was solved recently by the generation of an immortalized neural crest cell line, Monc-1, by retroviral transfection of mouse neural crest cells with the v-myc gene (17,18).
We hypothesized that Monc-1 cells could be used to develop an in vitro SMC differentiation system. We describe in this report the culture conditions under which Monc-1 cells can be differentiated efficiently into SMC. Exclusive Monc-1 to SMC differentiation was indicated by cellular appearance and induction of the SMC markers smooth muscle ␣-actin, smooth muscle myosin heavy chain, calponin, SM22␣, and APEG-1. Also, a 20 -30-fold increase in the activity of the SMC-specific promoter SM22␣ coincided with the formation of new DNA-protein complexes during differentiation.

EXPERIMENTAL PROCEDURES
Cell Culture and Reagents-The Monc-1 cell line was kindly provided by David Anderson (Pasadena, CA). Monc-1 cells were cultured in the undifferentiated state on fibronectin-coated plates in an L-15 CO 2based medium supplemented with chick embryo extract, hereafter referred to as complete medium, as described by Stemple and Anderson (11). Differentiation down the neuronal and glial pathways was performed on plates coated sequentially with poly-D-lysine (0.5 mg/ml) and fibronectin (0.25 mg/ml) in complete medium supplemented with 10% fetal bovine serum (HyClone, Logan, UT) plus 5 mM forskolin (Sigma) as described (17). SMC differentiation was induced by application of M199 (Life Technologies, Inc.) supplemented with 10% fetal bovine serum (HyClone), penicillin (100 units/ml), streptomycin (100 g/ml), and 25 mM HEPES (pH 7.4), hereafter referred to as SMC differentiation medium (SMDM).
RNA Extraction, Northern Analysis, and Reverse Transcription Po-lymerase Chain Reaction (PCR)-Total RNA from cultured cells was prepared by guanidinium isothiocyanate extraction and centrifugation through cesium chloride (19). RNA from mouse aorta was prepared by the RNAzol B method (Tel-Test, Friendswood, TX) according to the manufacturer's instructions. Total RNA was fractionated on a 1.3% formaldehyde-agarose gel and transferred to nitrocellulose filters, which were hybridized with the appropriate, randomly primed, 32 Plabeled probe (19,20). The smooth muscle ␣-actin was provided by J. Lessard (Cincinnati, OH); the angiotensin type II receptor (AT2) cDNA was provided by A. D. Strosberg (Paris, France). The calponin cDNA was isolated from a mouse aortic cDNA library. mRNAs for the smooth muscle myosin heavy chain isoforms SM1 and SM2 were detected by reverse transcription PCR with primers designed from the mouse SM1 and SM2 cDNAs (GenBank™ accession numbers D85923 and D85924, respectively). The forward primer 5Ј AGGAAA-CACCAAGGTCAAGCA 3Ј and the reverse primer 5Ј GGGACTGTAC-CACAGGTTAG 3Ј were used to amplify a 324-base pair SM1 fragment and a 363-base pair (alternatively spliced) SM2 fragment. To control for efficiency of reverse transcription, an aliquot of template cDNA was analyzed by PCR with a forward primer (5Ј TGAAGGTCGGTGTGAA-CGGATTTGGC 3Ј) and a reverse primer (5Ј CATGTAGGCCATGAGG-TCCACCAC 3Ј) designed from the mouse glyceraldehyde-3-phosphate dehydrogenase cDNA sequence.
Immunocytochemistry and Western Analysis-Monc-1 cells were grown on glass slides coated with fibronectin or fibronectin plus poly-D-lysine in the appropriate medium (see "Cell Culture and Reagents"). Immunostaining for smooth muscle ␣-actin, calponin, glial fibrillary acidic protein, and peripherin was performed as described (12,21). Proteins from undifferentiated and differentiated Monc-1 cells and mouse aortas were prepared according to standard procedures (19) with minor modifications. Proteins were resolved on 5% sodium dodecyl sulfate-polyacrylamide gels (22), transferred electrophoretically to nitrocellulose membranes (Schleicher and Schuell), and incubated with a rabbit anti-smooth muscle myosin heavy chain antibody (23) (kindly provided by Ute Groschel-Stewart of London, UK) diluted 1:5000, followed by incubation with a horseradish peroxidase-conjugated goat anti-rabbit antibody diluted 1:4000. Membranes were processed with an enhanced chemiluminescence reagent (Pierce) and exposed to film.
Transfection and Luciferase Assays-A 1.4-kilobase fragment of the SM22␣ promoter was obtained by PCR with mouse genomic DNA and the following primers: forward 5Ј CAGTGGCTGGAAAGCAAGAGC 3Ј and reverse 5Ј GGGCTGGGGCAGACGGGC 3Ј. The promoter fragment was subcloned into the pGL2-Basic vector (Promega, Madison, WI). Generation of the multimerized CArG and CArG mutant constructs in the pGL2-Basic vector will be described elsewhere. 2 Monc-1 cells were transfected transiently by electroporation as described (24). Cell extracts were prepared 48 -72 h after transfection, and luciferase and ␤-galactosidase assays were performed as described (25,26). Each construct was transfected at least six times. Data for each construct are presented as the mean Ϯ S.E.
Electrophoretic Mobility Shift Analysis-Nuclear extracts were prepared according to the method of Ritzenthaler et al. (27) with minor modifications. Electrophoretic mobility shift analysis was performed as described (28,29). In brief, double-stranded oligonucleotide probes synthesized according to the sequence of the SM22␣ CArG element 5Ј TCGAGACTTGGTGTCTTTCCCCAAATATGGAGCCTGTGTGGAGTG 3Ј were radiolabeled as described (21). The reaction mixture was incubated at room temperature for 20 min and analyzed by 5% native polyacrylamide gel electrophoresis in 0.25 ϫ TBE buffer (22 mM Tris base, 22 mM boric acid, and 0.5 mM EDTA). A 250-fold excess of specific or nonspecific oligonucleotide was used for competition experiments. For supershift experiments, 1 l of antibody to serum response factor (sc-335x, Santa Cruz Biotechnology, Santa Cruz, CA) or anti-YY1 antibody (sc-281x, Santa Cruz) was incubated with nuclear extracts and probes.

Identification of Culture Conditions That Promote Monc-1 to SMC Differentiation-Monc-1 cells on fibronectin-coated plates
can be maintained in the undifferentiated state in an L-15 CO 2 -based medium supplemented with chick embryo extract (11,17). After confirming that Monc-1 cells incubated in this complete medium expressed the low-affinity nerve growth factor receptor (data not shown), a marker for undifferentiated neural crest cells (11,12,17), we looked for culture conditions that would allow us to differentiate the Monc-1 cells down the smooth muscle lineage.
Although several media had little effect on the Monc-1 phenotype, culturing the cells in M199 (Life Technologies, Inc.) supplemented with 10% fetal bovine serum (HyClone), penicillin (100 units/ml), streptomycin (100 g/ml), and 25 mM HEPES (pH 7.4) induced a dramatic morphologic change. We studied two different lots of fetal bovine serum and observed no difference in their ability to induce Monc-1 cell differentiation. Within 24 h of placement in this SMDM, the cells began to assume a flat, fusiform appearance and the size of the cytoplasm increased. By 4 days of culture in SMDM nearly 100% of the cells had assumed this form (Fig. 1A, bottom). In comparison with undifferentiated Monc-1 cells (Fig. 1A, top), cells cultured in SMDM grew much more slowly. At confluence the differentiated cells took on the "hill and valley" appearance characteristic of differentiated SMC (not shown). Immunostaining with antibodies specific to smooth muscle ␣-actin (Fig.  1B) and calponin (Fig. 1C), two well established markers of the smooth muscle lineage (2), revealed robust expression of these genes in Monc-1 cells cultured in SMDM for 4 days but not in controls. As expected, differentiation of Monc-1 cells into glial and neuronal cells (17,18) was accompanied by immunoreactivity to glial acidic fibrillary protein (Fig. 1D) and peripherin (not shown). In contrast, SMC differentiated from Monc-1 cells did not stain for glial acidic fibrillary protein or peripherin (not shown).

Differentiation of Monc-1 Cells into Smooth Muscle Cells 5994
8 days). RNA from mouse aorta was used as a positive control for differentiated SMC. As a negative control, Monc-1 cells were induced to differentiate down the neuronal and glial pathways (17,18) in a parallel experiment. Smooth muscle ␣-actin and calponin mRNA expression increased as early as 2 days after placement in SMDM (Fig. 2). Neither message was detected after differentiation down the neuronal pathway.
Another well studied marker of SMC is the SM22␣ gene (2, 28, 30 -32), which is expressed exclusively in vascular and visceral SMC in adult animals (28). As with the smooth muscle ␣-actin and calponin mRNAs, expression of the SM22␣ mRNA increased after incubation in SMDM (Fig. 2). APEG-1, a nuclear protein preferentially expressed in vascular SMC, was cloned recently by our laboratory (33). Because its expression is high in arterial SMC in vivo and undetectable (by passage 3) in dedifferentiated arterial SMC in culture (33), APEG-1 represents a sensitive marker for differentiated SMC. Expression of APEG-1 mRNA also increased during Monc-1 cell culture in SMDM (Fig. 2), indicating that these cells share properties of differentiated vascular SMC. The last marker we tested in this experiment, the AT2, is expressed in the vasculature and many other tissues during embryogenesis (34). Although AT2 expression is down-regulated in the adult vasculature, it continues in adult neuronal cells (34). As Fig. 2 shows, AT2 mRNA was expressed in undifferentiated Monc-1 cells (day 0). Although AT2 receptor mRNA was still visible after 2 days of incubation in SMDM, it disappeared after 5 and 8 days of incubation. AT2 mRNA expression increased as neuronal/glial cell differentiation progressed, as reported previously (34). Expression of these five genes in the Monc-1 cell to SMC differentiation system was similar to that in the aorta, where SMC are highly differentiated.

Induction of Smooth Muscle Myosin Heavy Chain Expression in Response to
Monc-1 Cell Differentiation-To determine whether smooth muscle myosin heavy chain, a specific marker of differentiated SMC, is expressed in Monc-1 cells, we performed reverse transcription PCR with a pair of primers that amplify 324-and 363-base pair fragments of the SM1 and SM2 isoforms, respectively. The SM1 and SM2 DNA fragments were both amplified from reverse transcribed mouse aorta RNA (Fig.  3A). SM1 was amplified from differentiated but not undifferentiated Monc-1 cell RNA. We then used primers for glyceraldehyde-3-phosphate dehydrogenase to amplify a specific band from all RNA samples. SM1 and SM2 were both detected in samples prepared from mouse aortas by high resolution Western analysis with an antibody to smooth muscle myosin heavy chain (23) (Fig. 3B). Although undifferentiated Monc-1 cells expressed only non-muscle myosin heavy chain (Fig. 3B, asterisk), differentiated Monc-1 cells expressed SM1. Taken together, these data indicate the presence of SM1 in SMC differentiated from Monc-1 cells.
Induction of SM22␣ Promoter Activity by Monc-1 to SMC Differentiation-The molecular mechanisms regulating expression of the SM22␣ gene are very well characterized. For example, we know that a cis-acting element, the CArG box (CC A/T 6 GG), is critical for expression of SM22␣ in vascular SMC, in vitro and in vivo (28,30,32). To see whether the same cisacting element is critical to induction of SM22␣ in Monc-1 cells differentiated down the smooth muscle lineage, we performed transient transfection assays with the SM22␣ promoter.
First we generated a luciferase reporter construct, Ϫ1.4 kilobase SM22␣, containing 1.4 kilobases of the SM22␣ promoter. Monc-1 cells were transfected by electroporation and cultured in complete medium or SMDM, and luciferase activity was measured in cell lysates 72 h after transfection. The Ϫ1.4 kilobase SM22␣ promoter was minimally active in undifferentiated Monc-1 cells (Fig. 4A). After differentiation into SMC, however, promoter activity increased by 20 -30-fold. We then generated two more reporter constructs to see whether the CArG element was critical to regulation of SM22␣ promoter activity after Monc-1 cell to SMC differentiation in vitro. The first, CArG (3X) (Fig. 4A), contained three SM22␣ CArG elements upstream of the luciferase reporter gene, and the second, mt CArG (3X) (Fig. 4A), contained three CArG elements in which the core sequence had been modified from 5Ј CCAAATATGG 3Ј to 5Ј CCACACATGG 3Ј. This mutated sequence cannot function as an enhancer in SMC. Like the activity of the native SM22␣ promoter, the activity of the multimerized CArG reporter construct increased dramatically (Fig.  4A) after Monc-1 cell to SMC differentiation, again by about 20-fold. The mutated CArG construct was not active after neural crest cell to SMC differentiation.
Induction of Serum Response Factor DNA Binding Activity by Monc-1 to SMC Differentiation-We hypothesized that specific trans-acting factors may be induced after Monc-1 differentiation that would bind to the CArG element and thereby regulate SM22␣ expression. We tested this hypothesis in electrophoretic mobility shift assays with the CArG (3X) oligonucleotide used as a probe. Five specific DNA-protein complexes (Fig. 4B) were revealed. Complexes 1 and 4 appeared in nuclear extracts from differentiated Monc-1 cells but not in those from undifferentiated cells. Complex 2 was present under both conditions but appeared to intensify after differentiation. Antibody supershift experiments showed that complex 4 (shifted to 5) contains a protein antigenically identical or related to serum  3. Expression of smooth muscle myosin heavy chain isoform SM1 in differentiated Monc-1 cells. A, total RNA was harvested from stripped mouse aortas and Monc-1 cells at 0 (Undiff) and 6 (Diff) days after placement in SMDM. Reverse transcription PCR was performed with primers designed from the mouse SM1 and SM2 cDNAs. To control for efficiency of reverse transcription, an aliquot was analyzed for glyceraldehyde-3-phosphate dehydrogenase (GAPDH). B, total cellular proteins isolated from stripped mouse aortas (10 g) and Monc-1 cells (25 g) treated as in A were analyzed by Western blotting with anti-smooth muscle myosin heavy chain antibody. Asterisk, nonmuscle myosin heavy chain.

Differentiation of Monc-1 Cells into Smooth Muscle Cells 5995
response factor, whereas complex 2 (shifted to 3) contains a protein related to YY1. Complex 1 was the only one visible solely in nuclear extracts from Monc-1 cells after differentiation. The protein contained in this novel complex is currently under investigation. A 0.4-kilobase region of the SM22␣ promoter, within which lie two CArG elements, has been shown to confer expression only in the arterial SMC of transgenic mice (28,30). Because mutation of the proximal CArG element eliminates all SM22␣ expression in transgenic animals, this element appears to be necessary and sufficient for directing high level expression restricted to the SMC lineage (28). We show here that the activity of the SM22␣ promoter and a multimerized reporter construct containing the proximal CArG element increased markedly in Monc-1 cells after differentiation down the smooth muscle lineage (Fig. 4A). These data suggest that specific factors induced by differentiation bind to the CArG element and activate the SM22␣ promoter. Indeed, three DNA-protein complexes (Fig. 4B) were seen in nuclear extract from differentiated but not undifferentiated Monc-1 cells by gel mobility shift analysis, and the intensity of a fourth complex increased (Fig.  4B) in extract from differentiated Monc-1 cells. Our analysis shows that these four complexes contain proteins identical or antigenically related to serum response factor and YY1, two factors known to synergistically activate the CArG box of SM22␣ and other genes (28). Together our studies of the SM22␣ promoter suggest that the genetic program normally instituted in SMC in vivo is recapitulated during Monc-1 cell to SMC differentiation in vitro.
Because the excessive proliferation of vascular SMC is critical to the development of occlusive vascular diseases such as heart attack and stroke, the identification of nodal regulators of SMC differentiation is essential to the design of strategies for treating vascular disease. These nodal genes have been difficult to identify because there has been no system for producing differentiated SMC in vitro. We believe that the Monc-1 cell to SMC differentiation system described here represents the equivalent of the 10T1/2 cell system used to study regulators of skeletal muscle differentiation in vitro (2). Further study in the Monc-1 cell to SMC system may permit identification of analogous regulators of vascular SMC differentiation.

FIG. 4. Induction of SM22␣ promoter after differentiation. A,
Monc-1 cells were transfected by electroporation with 12.5 g of the indicated reporter plasmid and 2.5 g of cytomegalovirus-␤-galactosidase to control for transfection efficiency. Cell extracts were assayed after 72 h in complete medium (Ctrl) or SMDM (Diff) (means Ϯ S.E., n ϭ 6, *, p Ͻ 0.05). B, electrophoretic mobility shift analysis was performed with the CArG element of the SM22␣ promoter in nuclear extracts prepared from undifferentiated Monc-1 cells (Control) and cells incubated in SMDM (Differentiated) for 8 days. A 250-fold excess of identical (I) and nonidentical (NI) oligonucleotides was included. Supershift analysis was done with antibodies to serum response factor (SRF) and YY1 (YY1).