Evolutionarily Conserved Promoter Region ContainingCArG*-Like Elements Is Crucial for Smooth Muscle

Myosin Heavy Chain Gene ExpressionA. Zilberman, V. Dave, J. Miano, E.N. Olson, M. Periasamy

Abstract—In recent years, significant progress has been made toward understanding skeletal muscle development. However,the mechanisms that regulate smooth muscle development and differentiation are presently unknown. To betterunderstand smooth muscle–specific gene expression, we have focused our studies on the smooth muscle myosin heavychain (SMHC) gene, a highly specific marker of differentiated smooth muscle cells. The goal of the present study was toisolate and characterize the mouse SMHC gene promoter, since the mouse promoter would be particularly suited for invivo promoter analyses in transgenic mice and would serve as a tool for targeting genes of interest into smooth muscle cells.We report here the isolation and characterization of the mouse SMHC promoter and its 59 flanking region. DNA sequenceanalysis of a 2.6-kb portion of the promoter identified several potential binding sites for known transcription factors.Transient transfection analysis of promoter deletion constructs in primary cultures of smooth muscle cells showed that theregion between 21208 and 21050 bp is critical for maximal SMHC promoter activity. A comparison of SMHC promotersequences from mouse, rat, and rabbit revealed the presence of a highly conserved region located between 2967 and21208 bp. This region includes three CArG/CArG*-like elements, two SP-1 binding sites, a NF-1–like element, anNkx2–5 binding site, and an Elk-1 binding site. Gel mobility shift assay and DNase I footprinting analyses show that allthree CArG/CArG*-like elements can form DNA-protein complexes with nuclear extract from vascular smooth musclecells. Protein binding to the CArG* elements can be competed out by either serum response element or by an authenticCArG element from the cardiac a-actin gene. Using a serum response factor (SRF) antibody, we demonstrate that SRFis part of the protein complex. In addition, we show that cotransfection with the SRF dominant-negative mutantexpression vector abolishes SMHC promoter activity, suggesting that SRF protein plays a critical role in SMHCgene regulation. (Circ Res. 1998;82:566-575.)

Key Words: smooth muscle cell n myosin heavy chain n gene expression

Smooth muscle cells have been the subject of intense studybecause abnormal growth and proliferation of SMCs are

involved in the pathogenesis of atherosclerosis and other formsof vascular disease. Various proximal signals that influencegrowth and differentiation of SMCs, including growth factors,their receptors, and the intracellular signaling pathways, havebeen extensively studied. However, the cellular and molecularbasis of smooth muscle differentiation remains poorly under-stood. More important, the transcription factors responsible forsmooth muscle commitment and differentiation are not yetdefined.

The differentiation of smooth muscle (cells) involves expres-sion of specific protein markers and acquisition of functionalproperties characteristic of a fully mature SMC phenotype.1 Todate, several smooth muscle–specific protein markers havebeen identified, including smooth muscle a-actin,2,3 g-actin,4

smooth muscle calponin,5 SM22a,6 h-caldesmon,7 myosinlight chains,8 and SMHCs (SM1 and SM2).9–14 Several of these

SMC markers, including smooth muscle a-actin, SM22a, andcalponin, are expressed in developing skeletal and cardiacmuscle tissues and then become restricted to SMCs in the adultstages.15–18 Thus, these markers and their genes provide uniquereagents to study how they become restricted to SMCs duringdevelopment. On the other hand, the expression of SMHCisoforms SM1 and SM2 is found only in smooth muscletissues.11–13 SM1 and SM2 are the products of a single myosinheavy chain gene generated by alternative RNA splicing, andtheir expression is developmentally regulated.10,11,14 The SM1isoform is expressed both in embryonic and adult stages,whereas the expression of the SM2 isoform is restricted to fullydifferentiated/mature SMCs. Recent analyses of SMHC geneexpression during mouse embryogenesis using in situ hybrid-ization demonstrated that SMHC gene expression is restrictedto SMCs and is not found in other cell types, including cardiacand skeletal muscle cells, at any stage of development.13 Theexpression of SMHC mRNA was shown to appear at 10.5 dpc

Received October 10, 1996; accepted December 31, 1997.From the Division of Cardiology and Cardiovascular Research Center (A.Z., V.D., M.P.), University of Cincinnati (Ohio); the Department of Physiology

(J.M.), Medical College of Wisconsin, Milwaukee; and the Department of Molecular Biology & Oncology (E.N.O.), The University of Texas SouthwesternMedical Center at Dallas (Tex).

Correspondence to Muthu Periasamy, PhD, Director of Molecular Cardiology, Division of Cardiology, University of Cincinnati, 231 Bethesda Ave, ML0542, Cincinnati, OH 45267.

© 1998 American Heart Association, Inc.

566 by guest on July 22, 2015http://circres.ahajournals.org/Downloaded from

in the dorsal aorta and at 11.5 dpc in the outflow tract and toremain confined to the SMC lineage as development pro-gressed, with peripheral blood vessels of the head, musculature,and intersomitic region initially displaying a positive signal at13.5 to 14.5 dpc and the esophagus, bladder, and uretershowing intense labeling at 17.5 dpc. These results establishedSMHC as a highly specific marker for the SMC lineage.

Although studies in skeletal and cardiac muscle cells haveidentified several transcriptional factors that play a critical rolein the formation of these cell types, the mechanisms regulatingsmooth muscle development and differentiation are poorlyunderstood. In recent years, skeletal muscle development hasbecome the paradigm for understanding tissue-specific geneactivation and cell differentiation because of the discovery ofmaster regulatory genes, namely, the MyoD family. However,the MyoD family of genes or related helix-loop-helix proteinsthat regulate skeletal muscle differentiation are not expressed inSMCs.19 This raises the possibility that other types of transcrip-tion factors are involved in smooth muscle myogenesis. Asecond class of transcription factors, namely, the MEF-2family, originally described in skeletal and cardiac muscle, isalso expressed in smooth muscle.20 However, the role ofMEF-2 proteins in SMC differentiation is yet to be established.

We believe that the cloning and identification of smoothmuscle–specific genes and their promoters will contribute tothe dissection of the molecular mechanisms controlling smoothmuscle myogenesis. The recent cloning and identification ofsmooth muscle–specific genes, including SMHC,21–24 a- andg-actins,25–28 SM22a,15,29 and calponin,17,18 provide uniquereagents toward understanding smooth muscle–specific geneexpression. Our laboratory has previously reported the isola-tion and characterization of the rabbit SMHC gene promot-er.21 The goal of the present study was to isolate and charac-terize the mouse SMHC, mSMHC, gene promoter. ThemSMHC gene promoter would be particularly suited for invivo promoter analyses using transgenic mice and would serveas a powerful tool for targeted expression of specific proteinsinto SMCs. To this effect, we have isolated an mSMHCgenomic clone that contains 9.5 kb of upstream promoterregion and characterized an '2.6 kb of the promoter region byDNA sequence analysis. While this work was in progress,Watanabe et al22 reported the characterization of 1.5 kb of the

mSMHC gene promoter and showed that 2188 bp is suffi-cient for high-level expression in SMCs. In the present study,we provide additional data demonstrating that a highly con-served region located between 21208 and 21050 bp is criticalfor the SMHC promoter activity. This region shows a highdegree of sequence similarity between mouse, rat, and rabbitSMHC promoters and includes two CArG*-like elements,one authentic CArG element, two SP-1 binding sites, and twoNF-1–like binding sites. Using GMSA and DNAse I footprint-ing, we demonstrate that CArG*-like elements bind specificSRF-containing protein complexes in SMCs. In the presentstudy, we provide evidence that SRFs play a critical role inregulating SMHC promoter activity.

Materials and MethodsScreening of the Mouse Genomic LibraryTo screen the mouse genomic library, a PCR probe corresponding tothe mSMHC gene promoter was generated using two primers.OligoA 59GGTGGATTAGCAGGAGGACACCGGATG-39 corre-sponds to the mSMHC 59 untranslated region sequence; oligoB59-GACTTCCTTTTATGGCCTG-39 corresponds to a highly con-served promoter region found in the rabbit and rat SMHC genepromoters (21072 to 21053 bp in the rabbit promoter). PCRamplification of the mouse genomic DNA with the above primersyielded a 1.1-kb DNA fragment; DNA sequence analysis of thisfragment showed that it contained 63 bp of the 59 untranslated regionplus '1.0 kb of the 59 flanking region of the mSMHC promoter, witha canonical TATA box at 228 bp. The 1.1-kb genomic fragment wasused to screen a mouse (Svj 129) genomic library constructed in the lDash II vector containing 1.13106 independent clones with insertsizes ranging from 17 to 21 kb. Phage screening yielded three positiveclones, mSMHC-614, mSMHC-723, and mSMHC-813. Genomicclones were mapped by Southern blotting analysis.30 DNA sequencingwas performed according to the procedure of Sanger et al31 and byautomated DNA sequencing.32 Nucleotide sequence data were ana-lyzed using MacDNAsis software (Hitachi).

Construction of the mSMHC GenePromoter–CAT Chimeric ConstructsA 2.8-kb promoter fragment containing 2565 bp of the 59 flankingregion, 90 bp of exon I, and 140 bp of intron I was excised from theclone lmSMHC-813 using XbaI endonuclease. The fragment wasligated in a 59 to 39 orientation into a unique XbaI site of pBLCAT6expression vector,33 and orientation was confirmed by DNA sequenc-ing. The 140 bp of the intron sequence was removed from the 2.8-kbCAT construct using BglII endonuclease. The resulting '2.6-kbmSMHC-CAT construct was used for producing additional SMHCgene promoter–CAT deletion constructs. The SMHC promoterconstructs (p1831CAT, p1208CAT, p1128CAT, p1050CAT,p366CAT, and p121CAT) were made using native restriction sites inthe 59 flanking region (StuI, SacII, EcoRI, XhoI, BglII, and PstI,respectively).

Cell CultureSMCs from rat thoracic aorta were isolated and cultured as describedin Katoh et al.21 Briefly, rat thoracic aortas were excised, washed inHBSS, cleaned from adhering fat and connective tissue, and openedlongitudinally. After preincubation of the vessels for 20 minutes at37°C in a 5% CO2 incubator in the presence of 1 mg/mL collagenase(219 U/mg, Worthington) and 0.2 mg/mL elastase (4.2 U/mg,Worthington) in HBSS, the adventitia and endothelial cells werecarefully removed. The resulting aortic pieces were minced into 1- to2-mm2 sections and incubated at 37°C (5% CO2) in fresh collagenase-elastase solution for an additional 80 to 100 minutes. The dissociatedcells were separated from undigested tissue by filtration through an85-mm stainless-steel screen, and fetal calf serum was added to a finalconcentration of 30%. After the isolated cells were collected by

Selected Abbreviations and Acronyms

b-Gal 5 b-galactosidaseCAT 5 chloramphenicol acetyltransferase

dpc 5 day(s) post coitumGMSA 5 gel mobility shift assay

mSMHC 5 mouse SMHCMSV 5 murine sarcoma virusNF-1 5 nuclear factor-1PCR 5 polymerase chain reaction

SM1, SM2 5 SMHC isoforms 1 and 2SM22a 5 smooth muscle–specific protein

SMC 5 smooth muscle cellSMHC 5 smooth muscle myosin heavy chain

SP-1 5 stimulatory protein-1SRE 5 serum response elementSRF 5 serum response factor

Zilberman et al 567

by guest on July 22, 2015http://circres.ahajournals.org/Downloaded from

sedimentation at 1500 rpm for 6 minutes and resuspended in medium199 containing 10% fetal calf serum and 1% antibiotic-antimycotic (allreagents from Life Technologies, Inc), cells were seeded at 13104/cm2. Cells after the first or the second passage have been used for invitro transfection studies and for preparation of nuclear extracts. Sol8,34 a mouse soleus muscle cell line, was maintained in DMEM (LifeTechnologies, Inc) supplemented with 10% fetal calf serum. Sol 8myoblasts were induced to differentiate by switching to a mediumcontaining 5% horse serum (Life Technologies, Inc). NIH 3T3, amouse fibroblast cell line, was maintained in DMEM supplementedwith 10% fetal bovine serum.

DNA Transfections and CAT AssaysmSMHC promoter constructs (10 mg) containing CAT reporter geneswere cotransfected with MSV-b-Gal (5 mg) into cultured rat aorticSMCs, Sol 8, and NIH 3T3 cells by the calcium phosphate copre-cipitation method.35 Cells were incubated with DNA for 5 hours, thenwashed twice with PBS, and glycerol-shocked for 1 to 2 minutes.Fresh growth medium containing 10% FBS was added after twowashes with PBS. Cultured rat aortic SMCs and NIH 3T3 cells wereharvested 48 to 72 hours after transfection. Sol 8 myogenesis wasinduced 12 to 16 hours later with the replacement of 5% horse serum.Sol 8 cells were also harvested 48 to 72 hours after transfection.Harvested cells were washed twice with PBS, resuspended in 150 mLof STE buffer (40 mmol/L Tris-HCl, pH 7.4, 1 mmol/L EDTA, and150 mmol/L NaCl), and lysed through three freeze-thaw cycles.b-Gal activity was determined for each sample.36 CAT activity wasassayed according to established procedures and normalized for trans-fection efficiency using values obtained from b-Gal expression. Datarepresent the average of three or more independent transfectionexperiments run in duplicate. The construct pSV2CAT was used as apositive control.37

For cotransfection experiments, the SRF dominant-negative ex-pression vector pSRFpm1 (0.1, 1, and 5 mg) was transfected togetherwith p1208CAT and MSV-b-Gal constructs into cultured rat aorticSMCs.

Gel Mobility Shift AssayNuclear extract from cultured SMCs were prepared according toGossett et al.38 The protein concentration was determined by theBradford assay.39 GMSAs were performed by incubating 50 fmol of 32Pend-labeled CArG*I, CArG*II, CArG 1, c-fos SRE, or a-cardiacactin CArG box consensus DNA fragments with nuclear extract (4mg) for 30 minutes at room temperature in a volume of 20 mL ofbinding buffer40 and 2 mg of poly(dI)-poly(dC). Competition exper-iments were performed with 100-, 150-, and 200-fold molar excess ofMCK MEF-2,37 cardiac a-actin CArG box41 or SP-142 consensus,unlabeled DNA fragment, or 200-fold molar excess of SRE.43 Thebinding reactions were immediately loaded onto 6% nondenaturingpolyacrylamide gel containing 0.53 Tris-borate-EDTA buffer andelectrophoresed at 160 V for 3 hours. The gel was dried andautoradiographed. The following oligonucleotides were used in gelshift assays as probes or competitors: CArG*II, 59-GGCTGCGCGGGACCATATTTAGTCAGGG-39; CArG*I, 59-GCGCCTGGCCTTTTTGGGTTGTCTCCCGC-39; CArG 1, 59-GACTTCCTTTTATGGCCTGAG-39; c-fos SRE, 59-GGATGTCCATATTAGGACATCT-39; a-actin CArG, 59-GCGAAGGGGACCAAATAAGGCAAGGTGGC-39; MCK MEF-2, 59-GATCGCTCTAAAAATAACCCTGTCG-39; and SP-1 oligo, 59-ATTCGATCGGGGCGGGGCGAGC-39.

The SRF antibodies used for GMSA supershift were raised againstbacterially expressed human SRF.43 For supershift experiments, 1 mLof SRF antibody was added to the reaction mixture and incubated foran additional 15 minutes at room temperature and electrophoresed asdescribed above. In vitro synthesis of SRF was performed using atroponin T–coupled reticulocyte lysate translation system (Promega)with the human SRF cDNA clone pT7DATG44 as a template.

DNAse I FootprintingDNAse I footprinting was performed on the 21208-bp to 21050-bppromoter region using SMC nuclear extract. The DNA fragment was

39 end-labeled using [a-32P]dATP (3000 Ci/mmol) and Klenowenzyme. Each DNAse I mapping assay contained 5 fmol ('10 000cpm) of the end-labeled fragment in 50 mL of 23 binding buffer.40

Rat aortic nuclear extract was added at increasing concentrations of100 and 200 mg, and the binding reaction was carried out for 10minutes on ice. DNAse I (50 mL, Worthington) at a concentration of5 mg/mL in 10 mmol/L Tris-HCl, pH 8.0, 10 mmol/L MgCl2, and1 mmol/L CaCl2 was added to the binding mixture and incubated for30 seconds at room temperature. The reaction was terminated by theaddition of 100 mL of the stop solution (200 mmol/L NaCl,30 mmol/L EDTA, and 1% SDS), phenolized, and ethanol-precipi-tated. The samples were heat-denatured in the loading dye containing80% formamide and loaded on to a 6% sequencing gel. The A1Gladders were generated by the Maxam-Gilbert chemical sequencingmethod.45

ResultsIsolation of the mSMHC Gene PromoterA mouse genomic library (Svj 129) was screened with a 1.1-kbPCR-amplified DNA fragment (described in “Materials andMethods”) containing 1 kb of the mSMHC gene promoterand part of the first exon (63 bp). Screening of 1.13106

independent plaques yielded three different genomic clones:lmSMHC-614, lmSMHC-723, and lmSMHC-813. Re-striction mapping and Southern blotting revealed that thegenomic clones lmSMHC-614 and lmSMHC-813 over-lapped and contained the promoter region. The genomic clonelmSMHC-614 carried an '18-kb insert, which included 9.5kb of 59 flanking sequence, exon 1, and '8 kb of the intron(Fig 1). The clone lmSMHC-813 contained only 2.6 kb of the59 flanking sequence but included exon 1, the entire firstintron ('18 kb), and exon 2. The first exon encodes a portionof the 59 untranslated region (91 nt), and the sequence ishomologous (90%) to the rat gene. Interestingly, the size of thefirst intron (18 kb) in the mouse SMHC gene is very similar tothat of the rabbit SMHC gene.46

Characterization of the SMHC Gene PromoterRegion by DNA Sequence Analysis andComparison of the Promoter Elements BetweenMouse, Rat, and Rabbit SMHC GenesThe mSMHC promoter was recently cloned and sequencedfrom 21526 to 147 bp.22 In the present study, we performedadditional sequence analysis up to 22565 bp of the mousepromoter. The nucleotide sequence of the mSMHC promoterregion (22565 bp upstream from the transcription initiationsite) is shown in Fig 2. The start site of transcription wasdetermined by an RNase protection assay and found to mapprecisely to the site reported by Watanabe et al22 (data notshown). Sequence analysis revealed a canonical TATA box

Figure 1. Organization of the mSMHC genomic clonelmSMHC-614 containing the 9.5-kb promoter region. A canoni-cal TATAA box is present 28 nt upstream from the transcriptionstart site. 59UT indicates the 59 untranslated region; S, SacI; X,XbaI; and B, BglII.

568 Mouse Myosin Heavy Chain Gene Promoter

by guest on July 22, 2015http://circres.ahajournals.org/Downloaded from

(59TATAA39) sequence at position 228 bp, a GATA boxmotif at position 243 bp, and two CArG boxes: one at 2977bp (CArG 1, 59CCTTTTATGG39) and one at 21320 bp(CArG 2, 59CCAAAATAGG39). Also, two CArG*-like ele-ments were found: one at 21102 bp (CArG*I, 59CCTTTTTGGG39) and one at 21179 bp (CArG*II, 59-CCATATTTAG-39). An MEF-2–like sequence was identified at 21374bp, and two A/T-rich elements were also found at 2563 bpand 22085 bp. Sequence comparison with rabbit SMHC21 andrat SMHC23 gene promoters revealed that within the first2125 bp of the proximal promoter, there was a high degree ofsequence homology: mouse compared with rat was 81%, andmouse compared with rabbit was 65%. When compared withthe rat promoter, the mouse promoter showed 65% sequence

homology within 21200 bp, whereas comparison with therabbit promoter showed 55% homology within 21800 bp.

In addition, we identified a highly conserved region locatedbetween 2947 bp and 21208 bp in all three SMHC promot-ers (Fig 3). This region contains CArG*II at 21179 bp(59-CCATATTTAG-39), CArG*I at 21102 bp (59-CCTTTTTGGG-39), and one authentic CArG 1 box at 2977bp (59-CCTTTTATGG-39).47 The two CArG*-like boxeswere found in mouse, rat, and rabbit SMHC promoters andwere identical in their nucleotide sequence (Fig 3). TheseCArG*-like elements occurred at the same position andshowed similar spatial organization in all three promoters. Inaddition, three conserved SP-1 binding sites (59-GGGAGG-39at 21195 and 21069 bp and 59-CCCGCCC-39 at 21086 bp)and two NF-1–like binding sites48 (at 21147 and 21002 bp)were found in this region. Although the overall sequenceconservation between different SMHC gene promoters ispoor, the 21208- to 2947-bp region shows 90% of sequencehomology between mouse and rat and 70% between mouseand rabbit promoters (Fig 3).

Functional Characterization of the SMHC GenePromoter in Primary Vascular Smooth MuscleCultures and Its Tissue SpecificityTo characterize important cis elements regulating SMHC geneexpression, the SMHC-promoter deletions were linked to theCAT reporter vector, pBLCAT6 (Fig 4). Transient transfec-tion analyses into primary cultures of rat aortic SMCs revealedthat the promoter-CAT construct mSMHC-1208 was themost active (30-fold over pBLCAT6) and therefore was treatedas 100% activity for comparison with other constructs. Inclu-sion of more upstream sequences (p1831CAT and p2565CAT)decreased promoter activity (Fig 5). The shortest promoterconstruct, p121CAT, which contained the TATA box andtwo SP-1 binding sites, produced only 4% of the maximalCAT activity, whereas the p366CAT produced 35% of CATactivity (Fig 5).

Figure 2. Nucleotide sequence analysis of the mSMHC genepromoter. The putative cis-acting elements are underlined andlabeled. Asterisk denotes a mismatch with the consensussequence. The TATAAA box and 59 untranslated region (11 to101 bp) are underlined. E1 to E13 represent E boxes. The21208- to 2967-bp sequence, which is highly conservedbetween mouse, rat, and rabbit, is boxed.

Figure 3. Comparison of the mouse promoter with rat and rab-bit. The 21208- to 2967-bp region in the mouse promoter iscompared with rat and rabbit. The conserved cis-regulatory ele-ments are boxed. Spacing (dots) is introduced to align theregions of maximal homology.

Zilberman et al 569

by guest on July 22, 2015http://circres.ahajournals.org/Downloaded from

To investigate the functional importance of the highlyconserved promoter region (21208 to 2947 bp), two addi-tional deletion constructs were produced. The constructp1128CAT was created by deleting the CArG*II region. Inaddition, we created the p1050CAT construct, in which boththe CArG*II and CArG*I regions were deleted. Interestingly,when transfected into primary vascular SMCs, the p1128CATconstruct produced 57% of maximal CAT activity (Fig 5),whereas the deletion of both CArG*II and CArG*I resulted ina drastic reduction of the reporter activity (only 13% ofmaximal activity). These results suggest that both CArG*I andCArG*II elements are important for maximal promoteractivity.

To ascertain whether the mSMHC promoter is smoothmuscle specific, the promoter deletion CAT constructs weretransiently transfected into NIH 3T3 fibroblasts and Sol 8myotubes. As shown in Fig 5, none of the promoter constructsproduced significant CAT activity. However, the controlplasmid, PSV2CAT, gave a high level of CAT activity (30-foldover background) in these cell types (data not shown). Theseresults support our previous observations that SMHC geneexpression is restricted to SMCs.21

DNase I Footprinting Reveals Protection ofCArG*I and CArG*II-Like ElementsOur transient transfection analysis revealed that deletion of the160 bp of the promoter sequence between 21208 and 21050bp decreased CAT activity from 100% to 13%. An inspectionof the deleted sequence revealed the presence of two CArG-like boxes (CArG*I sequence [59-CCTTTTTGGG-39] andCArG*II sequence [59-CCATATTTAG-39]), one consensus-inverted Elk-1 site49 (59CAGGAAT-39), three SP-1 sites (twoidentical 59-GGGAGG-39 sites and one 59-CCCGCC-39 site),and one NF-1–like site (59-TGGTATGCCAC-39). To deter-mine the nature of protein binding sites within the 21208- to21050-bp promoter region, DNase I footprinting was per-formed using SMC nuclear extract (Fig 6). DNase I footprint-ing of the sense strand showed strong protection of theCArG*II sequence (footprint IV). In addition, footprintingwas also observed with the CArG* I binding site (footprint I).A significantly large area of protection was also observed atfootprint III (59-CACGCTGGAATTCCTG-39). A close ex-amination of this region reveals a putative protein bindingpalindromic sequence, 59-TGGAATTCCT-39. Interestingly,this protected region contains a consensus-inverted Elk-1 site,59-CAGGAAT-39.49 Footprint II indicated protection of 59-TTTCGAGAATTGCGCC-39. A transcription factor bind-

Figure 5. Transient transfection analyses ofSMHC promoter–CAT constructs in culturedsmooth muscle (VSMC), in NIH 3T3 fibro-blast cells, and in the skeletal muscle cellline (Sol 8). Cells were transfected with 5mg of MSV-b-Gal and 15 mg of SMHC pro-moter–CAT expression plasmids. The bargraph represents the relative CAT activitiescompared with the promoter constructp1208CAT, which is assigned a value of100% activity. The values of five separateexperiments were averaged, and eacherror bar represents the standard error ofthe mean.

Figure 4. Schematic representation of the SMHC promoter–CAT reporter constructs. The relative positions of the putativecis-activating regulatory elements, the TATA box, and the 59untranslated region (59UT) are indicated. Each SMHC-CAT con-struct is named on the basis of the number of nucleotidesupstream from the transcription initiation site.

570 Mouse Myosin Heavy Chain Gene Promoter

by guest on July 22, 2015http://circres.ahajournals.org/Downloaded from

ing-site search using the program TFSEARCH (version 1.3)showed maximum homology between the Nkx2-5 bindingsite50 (59-GATAATTG-39) and footprint II with a G instead ofa T (shown in bold).

CArG*-Like Elements Form Specific ProteinComplexes With Vascular Smooth MuscleNuclear ExtractTo determine precisely the nature of protein binding in thehighly conserved region (from 21208 to 2947 bp), weperformed GMSAs using 32P-labeled oligonucleotides corre-sponding to CArG*I, CArG*II, and CArG 1 sequences (Fig7). An oligonucleotide containing SRE derived from thehuman c-fos promoter was used as a positive control for SRFprotein binding and for competition analysis. As shown in Fig7, CArG*I, CArG*II, CArG 1, and c-fos SRE form asimilar-sized DNA-protein complex. However, CArG*I(lanes 1 to 4) and CArG*II (lanes 5 to 8) sequences consistentlyproduced more abundant complexes compared with the CArG1 element (lanes 9 to 12) and the c-fos SRE (lanes 13 to 15).Interestingly, 200-fold molar excess of c-fos SRE oligo wasable to compete out protein binding to CArG*I (lane 4) and toCArG 1 (lane 12) sequences. In contrast, SRE oligo was unableto compete out protein binding to the CArG*II sequence (lane8). This result was unexpected.

In addition, the cardiac a-actin CArG element (shown tobind SRF) was used as a competitor (Fig 8). Interestingly, thecardiac actin CArG box was able to compete protein bindingto the CArG*II element when added in the range of 100- to200-fold molar excess (Fig 8, lanes 3 to 5). Similarly, unlabeledCArG*II-like oligonucleotide added in 100- to 200-foldmolar excess could compete protein binding to the cardiac

Figure 6. DNase I footprint analysis of the mouse SMHC pro-moter region, 21208 to 21050 bp with SMC nuclear extract. A,A Maxam and Gilbert45 A1G ladder of the same region is shownin lane 1 (A/G). Lanes 2 and 3 correspond to increasing concen-trations of the nuclear extract (NE) in the presence of DNase I.Lane 4 represents DNA fragment incubated with DNase I (D)alone. Open boxes indicate protected regions. FP indicatesfootprints. B, Nucleotide sequences of protected regions areunderlined. Putative regulatory elements are shown in bold italic.

Figure 7. Binding of SMC nuclear proteins to the CArG*I-like,CArG*II-like, and CArG 1 elements from the SMHC promoter.Nuclear extracts from SMCs (4 mg per lane) were incubated withthe 32P-labeled oligonucleotides (see “Materials and Methods”).Lanes 1, 5, 9, and 13 represent free probes. Lanes 2, 6, 10, and14 show protein complexes formed with the different CArG ele-ments. H indicates competition with the 200 molar excess of thehomologous oligonucleotides, displayed in lanes 3, 7, 11, and15. S indicates competition with 200 molar excess of the SREoligonucleotides, shown in lanes 4, 8, and 12.

Zilberman et al 571

by guest on July 22, 2015http://circres.ahajournals.org/Downloaded from

a-actin CArG box (Fig 8, lanes 8 to 10). However, the cardiacactin CArG box produced a protein complex that is distinctfrom that of the CArG*II-like element in vascular SMCs (Fig8, lanes 1 and 6). These results suggest that the CArG elementsfrom different promoters may bind different SRF-containingprotein complexes.

It was previously shown that the 59-CC(A/T)6AG-39 se-quence present in Xenopus MyoDb promoter does bindMEF-2 protein.51 However, in our experiments the consensusoligo for MCK MEF-2 (100-fold and 200-fold molar excess)did not compete CArG*II protein binding, suggesting that thissequence does not bind MEF-2 protein in vascular SMCs. TheSP-1 consensus DNA fragment also failed to compete proteinbinding.

SRF Protein Is Part of the DNA-ProteinComplex Formed With CArG*I orCArG*II ElementsTo determine whether SRF is part of the protein complex thatbinds to the CArG*I or CArG*II element, we performedGMSA with an SRF antibody.43 For comparative analysis, we

included c-fos SRE, which is demonstrated to bind SRFprotein. Our results show (Fig 9) that addition of SRFantibody produced a supershift with CArG*I, CArG*II, andc-fos SRE. However, the antibody did not supershift the entirecomplex. This is probably due to the low concentration ofSRF antibody used in this assay, or, possibly, other DNAbinding proteins that are part of this complex interfere with theantibody’s access to SRF protein. Furthermore, we found thatin vitro–translated SRF protein binds to both CArG*I andCArG*II elements (data not shown).

Cotransfection of Dominant-Negative SRF toVascular SMCs Decreased SMHCPromoter ActivityTo determine whether the smooth muscle myosin promoterwas a direct target for SRF activity, the SRFpm1 mutantexpression vector was cotransfected with the p1208CATSMHC promoter construct into primary vascular smoothmuscle cultures as described in “Materials and Methods.” TheSRFpm1 mutant was previously demonstrated to block theendogenous SRF DNA-binding activity by forming het-erodimers with any available wild-type SRF.52 In response toSRFpm1 overexpression, the SMHC promoter activity wasdecreased significantly.

At concentrations 1 mg and higher, the SRF mutant vectorabolished 85% of SMHC promoter activity (Fig 10). Thisresult demonstrates that SRF proteins are critical for theSMHC promoter activity.

Figure 8. Cardiac a-actin CArG box competes protein bindingto the CArG*II-like element. Lane 1 shows CArG*II-like oligo pro-tein complexes; lane 2 includes a 100 molar excess of coldCArG*II-like (C II) oligo; lanes 3 to 5 show competition with 100,150, and 200 molar excess of cold cardiac a-actin (aAc) oligo;lanes 11 and 12 contain competition with 100 and 200 molarexcess, respectively, of cold MEF-2 oligo; and lane 13 repre-sents competition with 200-fold excess of SP-1 oligo. Lane 6shows shifted complex formed with cardiac a-actin CArG oligo;lane 7 includes a 100 molar excess of cold a-actin CArG (C a)oligo; lanes 8, 9, and 10 show competition with 100-, 150-, and200-fold excess, respectively, of the cold CArG*II-like oligo.

Figure 9. SRF protein binds to the CArG*I and CArG*II-like ele-ments. Nuclear extracts from SMCs (4 mg per lane) were incu-bated with the 32P-labeled oligonucleotides (see “Materials andMethods”). Lanes 1, 5, and 9 represent free probes. Lanes 2, 6,and 10 show protein complexes formed with the differentCArG*-like elements and c-fos SRE. Self competitions at 1503molar excess are displayed in lanes 3, 7, and 11. Supershiftedprotein complexes are shown in lanes 4, 8, and 12.

572 Mouse Myosin Heavy Chain Gene Promoter

by guest on July 22, 2015http://circres.ahajournals.org/Downloaded from

DiscussionThe goal of the present study was to isolate and characterizethe mouse SMHC (mSMHC) gene promoter. We report herethe isolation of the 9.5-kb mSMHC gene promoter and itsfunctional characterization in primary cultures of SMCs. ThemSMHC promoter is active only in SMCs but not in Sol 8 orNIH 3T3 cells, which is consistent with our earlier findings forthe rabbit SMHC promoter.21 Our transient transfection anal-ysis using the mSMHC promoter deletion constructs identifieda region between 21208 and 21050 bp in the SMHCpromoter as highly important for its expression in SMCs. Thisregion includes two CArG-like elements (CArG*II [59-CCATATTTAG-39] and CArG*I [59-CCTTTTTGGG-39]),two SP-1 binding sites (both 59-GGGAGG-39), and an NF-1–like element (59-TGGTATGCCAC-39). The two CArG*-like elements do not conform to the authentic CArG box,CC(T/A)6GG. Interestingly, this region is highly conservedbetween mouse, rat, and rabbit promoters, suggesting itsimportance for SMHC gene expression. In the present study,we demonstrate that CArG*I-like and CArG*II-like elementsform specific DNA-protein complexes that include SRF pro-tein as part of the complex. In addition, we demonstrate thatexpression of mutant SRF protein abolishes SMHC promoteractivity in SMCs. These results provide an important first steptoward further elucidating SMHC gene regulation both invitro and in vivo using transgenic mice.

In a recent study, Watanabe et al22 reported that 2188 bp ofthe mouse SMHC gene promoter (which includes two sets ofCCTCCC elements located at 289 bp and 261 bp) issufficient for high-level expression in SMCs. These elementswere shown to bind SP-1. However, this region does notprovide tissue-specific expression when linked to a heterolo-gous promoter. In contrast, our results show that the proximal2121 bp of the SMHC promoter that includes the CCTCCCelements was the least active (4% of maximal activity) inprimary cultures of SMCs. Our transient transfection analysisdemonstrates that the 21208-bp promoter construct gave themaximal activity in SMCs. Our data also indicate that deletionof '160 bp between 21208 and 21050 bp drastically de-creased the SMHC promoter activity (to 13% of maximalactivity) in primary SMCs. The incremental loss of activityobserved for the constructs p1128CAT and p1050CAT indi-cates that more than one positively acting cis element is presentwithin this region. Deletions in the corresponding region ofthe rat23,53 and rabbit24 SMHC promoter also produced asignificant decrease in promoter activity, suggesting that ele-ments located in this region play an important role in SMHCgene expression, whereas studies of Watanabe et al22 showedthat deletions between 21226 bp and 2188 bp of the 59flanking sequence of the mSMHC gene did not affect thepromoter activity. The discrepancy between these two findingsmay be attributed to differences in the cell culture systemsused in these studies. It is well known that when SMCs aregrown in culture, they readily dedifferentiate and often down-regulate SMHC gene expression. Therefore, depending on thedifferentiated state of SMCs, the SMHC promoter activitymay vary.

The most striking feature of the SMHC gene promoter isthe high degree of sequence similarity in the 21208- to

2947-bp region between mouse, rat, and rabbit. In particular,a set of three CArG/CArG*-like elements is preserved in thisregion (Fig 3) in mouse, rat, and rabbit promoters.22–24,53

In a recent study describing the rat SMHC promoter,Madsen et al53 reported that CArG 1 and CArG*I elements (intheir nomenclature CArG 1 and CArG 2, respectively) func-tioned as positive-acting cis elements. Their study also dem-onstrated that CArG 1 and CArG*I (CArG 2)–like elementsformed DNA-protein complexes that contained a factor anti-genically related to SRF. However, their study did not find theCArG*II (CArG 3) element as important for maximal SMHCpromoter activity, and they failed to detect any protein bindingto the CArG*II-like (CArG 3) element. In the present study,we found that the CArG*II-like element was equally impor-tant for SMHC promoter activity.

In the present study, we show that deletion of this elementreduces SMHC promoter activity significantly (Fig 5, con-struct p1128CAT), whereas deletion of both CArG*II andCArG*I decreases promoter activity to a low level. We alsoshow that all three CArG elements can form similar DNA-protein complexes with SMC nuclear extract. Interestingly,protein binding to the CArG*II-like element can be competedout by the cardiac a-actin CArG box, but not by c-fos SRE.Using SRF antibody, we have demonstrated that SRF is part ofthe protein complex produced by both CArG*II and CArG*Ielements. These data suggest that the CArG*II-like element isequally important for SMHC promoter activity.

In the present study, we also found that SRF pure proteincan bind directly to each of these CArG*-like elements.However, our GMSAs with SMC nuclear extract indicate thatSRF is not the only protein binding to the CArG*II-like orCArG*I-like element. On the basis of the abundance of theDNA-protein complex formed with individual CArG ele-ments, it appears that additional proteins are binding to thesesites. We interpret these data to suggest that protein binding to

Figure 10. Cotransfection of dominant-negative SRF in vascularSMCs decreased SMHC promoter activity. The SMHC promoterconstruct p1208CAT (10 mg), which contains three conservedCArG elements, was transfected together with the dominant-negative pSRFpm1 expression vector (0.1, 1, and 5 mg DNA)into primary vascular SMCs as described in “Materials andMethods.” The SRF dominate-negative transfections were car-ried out in duplicates, and values of two separate experimentswere averaged. Values represent the activity of the p1208CATconstruct divided by the pBLCAT6 promoterless vector.

Zilberman et al 573

by guest on July 22, 2015http://circres.ahajournals.org/Downloaded from

a CArG* element may be influenced by the core and flankingnucleotides of a CArG element. Future studies will focus onhow the flanking nucleotides determine protein binding to theCArG element.

Our cotransfection analyses with dominant-negative SRFexpression vector provide additional evidence for the role ofSRF proteins in the regulation of the SMHC promoter. Thestudies by Madsen et al53 on the rat promoter and our studieson the murine SMHC promoter strongly suggest that CArG/CArG*-like elements and CArG-SRF interaction are criticalfor the tissue-specific expression of the SMHC gene.

To date, transcription factors that are unique to SMCs havenot been identified, and smooth muscle–specific gene expres-sion remains underexplored. However, the recent isolation andcharacterization of the smooth muscle–specific gene promotersfor a-smooth muscle actin,25–27,54 g-smooth muscle actin,28

SM22a,15,29,55 calponin,17,18 and SMHC provide unique oppor-tunities for dissecting the various cis- and trans-regulatoryfactors involved in SMC-specific gene expression. Studiesusing the above-mentioned promoters have identified a num-ber of known cis elements, including the CArG box, E box,GATA-binding site, AP2, SP-1, CACCC boxes, and A/T-rich element. In particular, CArG elements that bind SRF havebeen shown to be important for smooth muscle a- and g-actingene expression.25,26,54 Similarly, CArG box elements are foundin the SM22a proximal promoter and have proved to beindispensable for its regulation in vascular SMCs.29,55,56 Thesestudies suggest that CArG elements play important roles indirecting the expression of a subset of smooth muscle–specificgenes. However, the precise role of SRF transcription factorsin smooth muscle–specific gene expression remains to beexplored. Future experiments will be directed toward identi-fying the role of SRF and its associated factors in thetranscriptional regulation of SMHC gene expression.

AcknowledgmentsThis study was supported by grant R01 HL-38355 from the NationalInstitutes of Health. We would like to thank Dr Robert J. Schwartz,Baylor College of Medicine, for providing pSRFpm1 expressionvector and for his helpful suggestions; Dr Narasimhaswamy S.Belaguli, Baylor College of Medicine, for providing SRF protein andantibody; and Sandy Nagel for her expert secretarial assistance.

References1. Katoh Y, Periasamy M. Growth and differentiation of smooth muscle cells

during vascular development. Trends Cardiovasc Med. 1996;6:100–106.2. Kocher O, Gabbiani G. Expression of actin mRNA in rat aortic smooth

muscle cells during development, experimental intimal thickening, andculture. Differentiation. 1986;32:245–251.

3. Ruzicka D, Schwartz RJ. Sequential activation of alpha actin genes duringavian cardiogenesis: vascular smooth muscle a-actin gene transcripts markthe onset of cardiac myocyte differentiation. J Cell Biol. 1988;107:2575–2586.

4. Sawtell NM, Lessard JL. Cellular distribution of smooth muscle actinduring mammalian embryogenesis: expression of the a-vascular but not theg-enteric isoform in differentiating striated myocytes. J Cell Biol. 1989;109:2929–2937.

5. Gimona M, Sparrow MP, Strasser P, Herzog M, Small JV. Calponin andSM22 isoforms in avian and mammalian smooth muscle: absence of phos-phorylation in vivo. Eur J Biochem. 1992;205:1067–1075.

6. Shahanan CM, Weissberg PL, Metcalfe JC. Isolation of gene markers ofdifferentiated and proliferating vascular smooth muscle cells. Circ Res.1993;73:193–204.

7. Frid MG, Shekhonin BV, Koteliansky VE, Glukhova MG. Phenotypicchanges of human smooth muscle cells during development: late expressionof heavy caldesmon and calponin. Dev Biol. 1992;153:185–193.

8. Kawashima M, Nabeshima Y, Obinata T, Fujii-Kuriyama Y. A commonmyosin light chain in chicken embryonic skeletal, cardiac, and smoothmuscle and in brain continuously express from embryo to adult. J BiolChem. 1987;262:14408–14414.

9. Nagai R, Larson DM, Periasamy M. Characterization of a mammaliansmooth muscle myosin heavy chain gene cDNA clone and its expression invarious smooth muscle cell types. Proc Natl Acad Sci U S A. 1988;85:1047–1051.

10. Nagai R, Kuro-o M, Babij P, Periasamy M. Identification of two types ofsmooth muscle myosin heavy chain isoforms: cDNA cloning and immu-noblot analysis. J Biol Chem. 1989;264:9734–9737.

11. Kuro-o M, Nagai R, Katoh H, Tsuchimochi H, Yazaki Y, Ohkubo A,Takaku F. Developmentally regulated expression of vascular smoothmuscle myosin heavy chain isoforms. J Biol Chem. 1989;264:18272–18275.

12. Borrione AC, Zanellato AMC, Giuriato L, Scannapieco G, Pauletto P,Sartore S. Myosin heavy chain isoforms in adult and developing rabbitvascular smooth muscle. Eur J Biochem. 1989;183:413–417.

13. Miano JM, Cserjesi P, Ligon KL, Periasamy M, Olson EN. Smooth musclemyosin heavy chain exclusively marks the smooth muscle lineage duringmouse embryogenesis. Circ Res. 1994;75:803–812.

14. Babij P, Periasamy M. Myosin heavy chain diversity in smooth muscle cellsis produced by differential RNA processing. J Mol Biol. 1989;210:673–679.

15. Solway J, Seltzer J, Samaha FF, Kim S, Alger LE, Nui Q, Morrisey EE, IpHS, Parmacek MS. Structure and expression of a smooth muscle cell-specific gene, SM22a. J Biol Chem. 1995;270:13460–13469.

16. Li L, Miano J, Cserjesi P, Olson EN. SM22a, a marker of adult smoothmuscle, is expressed in multiple myogenic lineages during embryogenesis.Circ Res. 1996;78:188–195.

17. Samaha FF, Ip HS, Morrisey EE, Seltzer J, Tang Z, Solway J, Parmacek S.Developmental pattern of expression and genomic organization of thecalponin-h1 gene. J Biol Chem. 1996;271:395–403.

18. Miano JM, Olson EN. Expression of the smooth muscle cell calponin genemarks the early cardiac and smooth muscle cell lineages during mouseembryogenesis. J Biol Chem. 1996;271:7095–7103.

19. Olson EN. MyoD family: a paradigm for development. Genes Dev. 1990;4:1454–1461.

20. Firulli AB, Miano JM, Bi W, Johnson AD, Casscells W, Olson EN,Schwarz JJ. Myocyte enhancer binding factor-2 expression and activity invascular smooth muscle cells: association with the activated phenotype. CircRes. 1996;78:196–204.

21. Katoh Y, Loukianov E, Kopras E, Zilberman A, Periasamy M. Identifi-cation of functional promoter elements in the rabbit smooth muscle myosinheavy chain gene. J Biol Chem. 1994;269:30538–30545.

22. Watanabe M, Sakomura Y, Kurabayashi M, Manabe I, Aikawa M, Kuro-oM, Suzuki T, Yazaki Y, Nagai R. Structure and characterization of the59-flanking region of the mouse smooth muscle myosin heavy chain(SM1/2) gene. Circ Res. 1996;78:978–989.

23. White SL, Low RB. Identification of promoter elements involved incell-specific regulation of rat smooth muscle myosin heavy chain tran-scription. J Biol Chem. 1996;271:15008–15017.

24. Kallmeier RC, Somasundaram C, Babij P. A novel smooth muscle-specificenhancer regulates transcription of the smooth muscle myosin heavy chaingene in vascular smooth muscle cells. J Biol Chem. 1995;270:30949–30959.

25. Carroll SL, Bergman DJ, Schwartz RJ. A 29-nucleotide DNA segmentcontaining an evolutionarily conserved motif is required in cis for cell-type-restricted repression of the chicken a-smooth muscle actin gene corepromoter. Mol Cell Biol. 1988;8:241–250.

26. Blank RS, McQuinn TC, Yin KC, Thompson MM, Takeyasu K,Schwartz RJ, Owens GL. Elements of the smooth muscle alpha-actinpromoter required in cis for transcriptional activation in smooth muscle:evidence for cell type-specific regulation. J Biol Chem. 1992;267:984–989.

27. Foster DN, Min B, Foster LK, Stoflet ES, Sun S, Getz MJ, Strauch AR.Positive and negative cis-acting regulatory elements mediate expression ofthe mouse vascular smooth muscle alpha-actin gene. J Biol Chem. 1992;267:11995–12003.

28. Quin J, Kumar A, Szucsik J, Lessard JM. Tissue and developmental specificexpression of murine smooth muscle g actin (SMGA) fusion genes intransgenic mice. Dev Dyn. 1996;207:135–144.

29. Li L, Miano JM, Mercer B, Olson EN. Expression of the SM 22a promoterin transgenic mice provides evidence for distinct transcriptional regulatoryprogram in vascular and visceral smooth muscle cells. J Cell Biol. 1996;132:849–859.

574 Mouse Myosin Heavy Chain Gene Promoter

by guest on July 22, 2015http://circres.ahajournals.org/Downloaded from

30. Maniatis T, Fritsch EF, Sambrook P. Molecular Cloning: A LaboratoryManual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press;1989.

31. Sanger F, Nicklen S, Coulson AR. DNA sequencing with chain termi-nating inhibitors. Proc Natl Acad Sci U S A. 1977;74:5463–5467.

32. Prober J, Trainor G, Dam R, Hobbs F, Robertson C, Zagursky R,Cocuzza R, Jensen M, Baumeister K. A system for rapid DNA sequencingwith fluorescent chain terminating dideoxinucleotides. Science. 1987;238:336–341.

33. Boshart M, Kluppel M, Schmidt A, Schultz G, Luckow B. Reporterconstructs with low background utilizing the cat gene. Gene. 1992;110:129–130.

34. Mulle C, Benoit P, Pinset C, Roa M, Changeux J-P. Calcitonin gene-related peptide enhances the rate of desensitization of the nicotinic acetyl-choline receptor in cultured mouse muscle cells. Proc Natl Acad Sci U S A.1988;85:5728–5732.

35. Brent FR, Kingston R, Moore D, Siedman G, Smith J, Struhl K. CurrentProtocols in Molecular Biology. New York, NY: John Wiley & Sons Inc;1987.

36. Herbomel P, Bourachot B, Yaniv M. Two distinct enhancers with dif-ferent cell specificities coexist in the regulatory region of polyoma. Cell.1984;39:653–662.

37. Gorman CM, Moffat LF, Howard BH. Recombinant genomes whichexpress chloramphenicol acetyltransferase in mammalian cells. Mol CellBiol. 1982;2:1044–1051.

38. Gossett LA, Kelvin DJ, Sternberg EA, Olson EN. A new myocyte-specificenhancer-binding factor that recognizes a conserved element associatedwith multiple muscle-specific genes. Mol Cell Biol. 1989;9:5022–5033.

39. Bradford MM. A rapid and sensitive method for the quantitation ofmicrogram quantities of protein utilizing the principle of protein-dyebinding. Anal Biochem. 1976;72:248–254.

40. Pascal E, Tjian R. Different activation domains of SP1 govern formation ofmultimers and mediate transcriptional synergism. Genes Dev. 1991;5:1646–1656.

41. Sartorelli V, Webster KA, Kedes L. Muscle-specific expression of thecardiac alpha-actin gene requires MyoD1, CArG-box binding factor andSP1. Genes Dev. 1990;4:1811–1822.

42. Briggs MR, Kadonaga JT, Bell SP, Tjian R. Purification and biochemicalcharacterization of the promoter-specific transcriptional factor, SP1. Science.1986;234:47–52.

43. Manak JR, Prywes R. Phosphorylation of serum response factor by caseinkinase II: evidence against a role in growth factor regulation of c-fosexpression. Oncogene. 1993;8:703–711.

44. Norman C, Renswick M, Pollock R. Isolation and properties of cDNAclones encoding SRF, a transcriptional factor that binds to the c-fos serumresponse element. Cell. 1988;55:989–1003.

45. Maxam AM, Gilbert W. Sequencing end-labeled DNA with base specificchemical cleavages. Methods Enzymol. 1980;65:499–560.

46. Babij P, Kelly C, Periasamy M. Characterization of a mammalian smoothmuscle heavy chain gene: complete nucleotide and protein codingsequence and analysis of the 59 end of the gene. Proc Natl Acad Sci U S A.1991;88:10676–10680.

47. Boxer L, Prywes R, Roeder R, Kedes L. The sarcomeric actin CArG-binding factor is indistinguishable from the c-fos serum response factor. MolCell Biol. 1989;9:515–522.

48. Faist S, Meyer S. Compilation of vertebrate-encoded transcriptionalfactors. Nucleic Acids Res. 1992;20:3–26.

49. Triesman R, Marais R, Wynne J. Spatial flexibility in ternary complexesbetween SRF and its accessory proteins. EMBO J. 1992;11:4631–4640.

50. Chen CY, Schwartz RJ. Identification of novel DNA binding targets andregulatory domains of a murine tinman homeodomain factor, nkx 2–5.J Biol Chem. 1995;270:15628–15633.

51. Leibham D, Wong MW, Cheng TC, Schroeder S, Weil AP, Olson EN,Perry M. Binding of TFIID and MEF-2 to the TATA element tran-scription of the Xenopus MyoDA promoter. Mol Cell Biol. 1994;14:686–699.

52. Croissant JD, Kim JH, Eichele G, Goering L, Lough J, Prywes R, SchwartzRJ. Avian serum response factor expression restricted primarily to musclecell lineages is required for a-actin gene transcription. Dev Biol. 1996;177:250–264.

53. Madsen GS, Hershey JC, Hautman MB, White SL, Owens GK. Expressionof the smooth muscle myosin heavy chain gene is regulated by a negative-acting GC-rich element located between two positive acting serumresponse factor binding elements. J Biol Chem. 1997;272:6332–6340.

54. Shimizu RT, Blank RS, Jervis R, Lawrenz-Smith SC, Owens GK. Thesmooth muscle a-actin gene promoter is differentially regulated in smoothmuscle versus non-smooth muscle cells. J Biol Chem. 1995;270:7631–7643.

55. Moessler H, Mericskay M, Li Z, Nagl S, Paulin D, Small JV. The SM 22promoter directs tissue-specific expression in arterial but not in venous andvisceral smooth muscle cells in transgenic mice. Development. 1996;122:2415–2425.

56. Kim S, Ip HS, Lu MM, Clendenin C, Parmacek MS. A serum responsefactor dependent transcriptional regulatory program identifies distinctsmooth muscle cell sublineages. Mol Cell Biol. 1997;17:2266–2278.

Zilberman et al 575

by guest on July 22, 2015http://circres.ahajournals.org/Downloaded from

A. Zilberman, V. Dave, J. Miano, E. N. Olson and M. Periasamyfor Smooth Muscle Myosin Heavy Chain Gene Expression

Evolutionarily Conserved Promoter Region Containing CArG*-Like Elements Is Crucial

Print ISSN: 0009-7330. Online ISSN: 1524-4571 Copyright © 1998 American Heart Association, Inc. All rights reserved.is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation Research

doi: 10.1161/01.RES.82.5.5661998;82:566-575Circ Res. 

http://circres.ahajournals.org/content/82/5/566World Wide Web at:

The online version of this article, along with updated information and services, is located on the

  http://circres.ahajournals.org//subscriptions/

is online at: Circulation Research Information about subscribing to Subscriptions: 

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

  document. Permissions and Rights Question and Answer about this process is available in the

located, click Request Permissions in the middle column of the Web page under Services. Further informationEditorial Office. Once the online version of the published article for which permission is being requested is

can be obtained via RightsLink, a service of the Copyright Clearance Center, not theCirculation Researchin Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:

by guest on July 22, 2015http://circres.ahajournals.org/Downloaded from