Cellular and genetic approaches to myocardial regeneration Tuyn, J. van

(1)

Cellular and genetic approaches to myocardial regeneration

Tuyn, J. van

Citation

Tuyn, J. van. (2008, January 9). Cellular and genetic approaches to myocardial

regeneration. Department of Cardiology and Department of Molecular Cell Biology (MCB), Faculty of Medicine, Leiden University Medical Center (LUMC), Leiden University.

Retrieved from https://hdl.handle.net/1887/12548

Version: Corrected Publisher’s Version

License: Licence agreement concerning inclusion of doctoral thesis in the Institutional Repository of the University of Leiden

Downloaded from: https://hdl.handle.net/1887/12548

Note: To cite this publication please use the final published version (if applicable).

(2)

Chapter

2

John van Tuyn Shoshan Knaän-Shanzer Marloes J.M. van de Watering Michelle de Graaf Arnoud van der Laarse Martin J. Schalij Ernst E. van der Wall Antoine A.F. de Vries Douwe E. Atsma

Cardiovascular Research 2005;67(2):245-255

Activation of cardiac and smooth muscle- speciﬁc genes in primary human cells after

forced expression of human myocardin

(3)

28 Abstract

Objective: Myocardin is a recently discovered transcriptional regulator of cardiac and smooth muscle development. Its ability to transactivate smooth muscle-speciﬁc genes has been ﬁrmly established in animal cells but its effect on heart muscle genes has been investigated less extensively and the consequences of ectopic myocardin expression in human cells are unknown.

Methods: In this study, primary human mesenchymal stem cells and foreskin ﬁbroblasts were transduced with human adenovirus vectors expressing the longest splice variant of the human myocardin gene (hAd5/F50.CMV.myocL) or with control vectors. One week later, the expression of muscle-restricted genes in these cells was analyzed by reverse transcription-polymerase chain reactions and immunoﬂuorescence microscopy.

Results: Forced expression of myocardin induced transcription of cardiac and smooth muscle genes in both cell types but did not lead to activation of skeletal muscle-speciﬁc genes. Double labeling experiments using monoclonal antibodies directed against striated (i.e. sarcomeric

α

^-actin

and sarcomeric

α

-actinin) and cardiac (i.e. natriuretic peptide precursor A) muscle-specific proteins together with a polyclonal antiserum specific for smooth muscle myosin heavy chain revealed that hAd5/F50.CMV.myocL- transduced cells co-express heart and smooth muscle-specific genes.

Conclusions: These data indicate that the myocardin protein is a strong inducer of both smooth and cardiac muscle genes, but that additional factors are necessary to fully commit cells to either cardiac or smooth muscle cell fates.

(4)

29

Chapter 2Activation of cardiac and smooth muscle-speciﬁc genes in primary human cells after forced expression of human myocardin

Introduction

The differentiation of cardiac, skeletal, and smooth muscle cells is characterized by the transactivation of overlapping but distinct sets of muscle-restricted genes.

Cardiovascular and neuromuscular diseases are often associated with alterations in the differentiation state of muscle cells. Accordingly, studying the factors that orchestrate the genetic program of cardiac and smooth muscle cells is important for understanding cardiac disease and to guide the development of novel treatment modalities aiming at the formation of new blood vessels and cardiomyocytes in infarcted myocardium.

Myocardin is a recently discovered transactivator of the ubiquitous transcription factor (TF) serum response factor¹, which regulates the expression of many growth- related and muscle-restricted genes by binding to CC(A/T)₆GG and closely related nucleotide sequence motifs (also known as CArG boxes) in their promoters². In situ hybridization, Northern blot, and reverse transcription-polymerase chain reaction (RT-PCR) analyses have shown that myocardin expression is (largely) restricted to cardiac and smooth muscle tissue both during embryogenesis and in adults^1,

3-6. The repression of muscle protein-coding genes in the heart anlage, but not in the developing somites of Xenopus laevis embryos after microinjection in 8-cell blastomeres of mRNA encoding a dominant negative form of myocardin, yielded further evidence for a speciﬁc role of myocardin in cardiac muscle development¹. Moreover, expression of a recombinant gene coding for a dominant negative version of myocardin in murine teratocarcinoma cells blocked their in vitro differentiation into cardiomyocytes⁷. Dominant negative myocardin mutants also inhibited the activity of the smooth muscle

α

-actin (ASMA), smooth muscle myosin heavy chain (smMHC), and transgelin (SM22) promoters but did not affect transcription of the smoothelin and aortic carboxypeptidase-like protein genes in various in vitro assay systems1, 4, 6, 8, 9. Failing human and porcine myocardium contained significantly more myocardin mRNA than heart muscle from healthy individuals⁵. Targeted degradation of myocardin transcripts in rat aortic smooth muscle cells by RNA interference technology inhibited expression of recombinant luciferase genes placed under the control of smooth muscle gene-specific promoters^{4, 6}. In addition, myocardin has been shown to activate promoters of both cardiac and smooth muscle genes in transfection assays based on the use of reporter gene constructs1, 3, 4, 6-12. However, transgenic mice lacking exons 8 and 9 of the myocardin gene displayed no obvious defects in cardiac morphogenesis but failed to form smooth muscle cells and died at embryonic day 10.5 with severe vascular abnormalities¹³. These results suggest that the smooth muscle developmental program critically depends on myocardin but that its function in heart muscle formation can be taken over by other TFs. Moreover, whereas forced expression of myocardin in non-human cells consistently induced the transcription of multiple smooth muscle genes, the activation of cardiomyocyte-specific genes was much less usual^{4, 6-12}.

Since until now the transcription-enhancing activity of myocardin has only been studied in animal cells, we investigated the effect of ectopically expressed human myocardin on the expression of muscle-speciﬁc genes in human mesenchymal stem cells (hMSCs) isolated from bone marrow (BM) and in human dermal

(5)

30

ﬁbroblasts (hDFs) derived from foreskin. hMSCs were chosen for this study because: (i) hMSCs are a readily obtainable source of autologous cells with the ability to generate bone, cartilage, fat, ligament, stroma, tendon, and under special conditions also muscle. (ii) hMSCs can be greatly expanded ex vivo using ordinary cell culture systems. (iii) Differentiation of hMSCs into cardiomyocytes has been reported both in vitro and in vivo, which makes these cells potentially useful for repairing heart muscle damage (see¹⁴ and references therein). The hDFs were included as previous in vitro studies on myocardin were mostly performed in ﬁbroblasts.

RT-PCR analyses of total RNA extracted from cultures of these two primary human cell types showed that myocardin activates genes encoding cardiomyocyte-specific TFs, sarcomeric components, and pore-forming proteins as well as smooth muscle cell-specific contractile proteins and thin filament regulatory proteins but does not induce expression of skeletal muscle-restricted genes. Immunofluorescence microscopy (IFM) revealed that myocardin initiated muscle-specific gene expression in 5 to 15% of the hMSCs. These particular cells stained positive for cardiac as well as smooth muscle cell markers. Our findings underscore that myocardin is an important activator of the smooth muscle genetic program and show for the first time its ability to transactivate multiple endogenous heart muscle-specific genes in vitro. Furthermore, the acquisition of both heart and smooth muscle characteristics by the myocardin-responsive hMSCs suggests that other factors are necessary to fully commit these cells to either smooth or cardiac muscle cell fates.

Methods

Primary human cells and human cell lines

hMSCs were purified from BM samples of adult patients as described¹⁴. All human materials were obtained after informed consent of the donors and experiments with the human specimens were carried out according to the official guidelines of the Leiden University Medical Center. Both hMSCs and hDFs (VH10 cells¹⁵) were cultured in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% fetal bovine serum (FBS), 100 U/ml penicillin, and 100 μg/ml streptomycin. The generation and propagation of conditionally immortalized euploid myoblasts derived from a patient with Duchenne muscular dystrophy has been described elsewhere¹⁶. These cells are phenotypically indiscernible from their nonimmortalized predecessors in various assays and can form myotubes in vitro. All cells were cultured at 37°C in a humidified air-10% CO₂ atmosphere except for the skeletal myoblasts, which were kept at 5% CO₂.

Characterization of hMSCs

The surface antigen expression proﬁle of bone marrow (BM)-derived mesenchymal stem cells (hMSCs) was determined by ﬂow cytometry. Prior to analysis, the cell monolayers were trypsinized, the cells were resuspended in phosphate-buffered

(6)

31

saline (PBS) containing 0.5% bovine serum albumin (BSA; Sigma-Aldrich), divided into aliquots of 5×10⁴ cells, and incubated with the antibodies specified in Table 1. When the primary antibodies were not labelled, secondary fluorochrome- conjugated antibodies (phycoerythrin-conjugated rat anti-mouse IgG1 and fluorescein isothiocyanate-labeled goat anti-mouse IgG; Becton Dickinson) were added to the cells immediately after the addition of the primary antibodies. As controls for nonspecific binding, we used either fluorescently labeled, isotype- matched mouse monoclonal antibodies (MAbs) specific for the nonhuman antigen keyhole limpet hemocyanin (Becton Dickinson) or incubated the cells exclusively with the fluorescently labeled secondary antibodies. After a 30-min incubation at 4°C the cells were washed with PBS containing 0.5% BSA and analysed using a FACSort flow cytometer (Becton Dickinson) equipped with a 488-nm argon ion laser and a 635-nm red diode laser. Typically, 5,000 events were acquired per sample. Data were processed using CellQuest Software (Becton Dickinson). The surface antigen profile of the hMSCs matches previously published data¹⁴. In contrast to the hDFs, the hMSCs differentiated into adipocytes and osteoblasts after proper stimulation confirming their multipotent nature (data not shown).

RNA extraction

Total cellular RNA was extracted from tissues and cultured cells using TRIzol reagent (Invitrogen). Human atrial RNA was derived from pooled right atrial appendices excised during cardiac surgery of adult patients; human ventricular RNA was obtained from an individual who underwent partial left ventriculectomy.

Skeletal muscle RNA was extracted from a specimen of vastus lateralis muscle removed during orthopedic surgery. Arteries from human umbilical cords served as a source of vascular smooth muscle RNA.

Cloning of human myocardin cDNA

The genomic sequence for human myocardin was identiﬁed in a tblastn search of the GenBank database using murine myocardin (GenBank accession number:

AAK71683.2) as query. The gene for a protein with 78% amino acid sequence identity to murine myocardin was found on human chromosome 17. Human myocardin-speciﬁc cDNA was generated from 500 ng of atrial RNA using 10 fmol of primer 5’-AGTATGTATGGCTGTGTGCTTTCC-3’, 10 nmol of each dNTP, and 200 U of SuperScript II RNase H- reverse transcriptase (SRT; Invitrogen) in a 20-μl volume.

The cDNA was ampliﬁed by polymerase chain reaction (PCR) in a 50-μl volume using 1 μl of DNA template, 10 pmol of primers 5’-CGGGATCCGGCAGCCTATGACATCA- 3’ and 5’- CTCCATTGGTCTTCCATAGCACTG-3’, 10 nmol of each dNTP, and 2.5 U of Platinum Taq DNA polymerase High Fidelity (Invitrogen). A touchdown PCR program of 40 cycles with the annealing temperature declining from 62°C to 57°C was used. The 3.3-kb PCR product was digested with SmaI and ligated into the HpaI-digested Ad vector shuttle plasmid pAdApt¹⁷. In the resulting plasmid, designated pAdApt.myoc1, the myocardin open reading frame is preceded by the human cytomegalovirus immediate early gene promoter and followed by the simian virus 40 polyadenylation signal. Two independent clones of pAdApt.myoc1

(7)

32

were selected to determine the nucleotide sequence of human myocardin cDNA by the dideoxy chain termination method using an ABI PRISM 3700 DNA analyser and the ABI PRISM BigDye Terminator version 3.0 cycle-sequencing system (both from Applied Biosystems) together with the primers listed in Table 2. Nucleotide sequence analyses revealed that the largest splice variant (splice variant B⁵) of the human myocardin gene had been cloned. The assembled nucleotide sequence has been deposited in GenBank (http://www.ncbi.nih.gov/) under accession number AY764180.

First-generation adenovirus vectors

Fiber-modified human adenovirus serotype 5 (hAd5) vectors (hAd5/F50 vectors) encoding human myocardin (hAd5/F50.CMV.myocL) or without heterologous expression cassette (hAd5/F50.empty) were generated as described¹⁷. The vectors hAd5/F5.CMV.eGFP and hAd5/F50.CMV.eGFP, expressing the enhanced green fluorescent protein (eGFP) gene and carrying human adenovirus serotype 5 and 50 fiber domains, respectively, have been documented elsewhere¹⁷. To determine the optimal vector (doses) for hMSCs and hDFs, these cell types were transduced with 0, 12.5, 25, 50, 100, or 200 infectious units (IU) of hAd5/F5.CMV.eGFP or hAd5/F50.CMV.eGFP per cell and seeded in 2-cm² wells at a concentration of 10⁴ cell/cm². At 72 hours postinfection, the cells were subjected to flow cytometric analyses.

Table 1. Antibodies used for the immunophenotypic characterization of hMSCs

antigen source clone (iso)type label species

CD1a BD SK9 IgG2b - mouse

CD10 CLB CLB-CALLA/1, 4F9 IgG2a PE mouse

CD14 CLB CLB-mon/I, 8G3 IgG2a PE mouse

CD19 BD 4G7 IgG1 PE mouse

CD34 BD 8G12 IgG1 FITC mouse

CD44 BD G44-26 (C26) IgG2b PE mouse

CD45 CLB CLB-T200/1, 15D9 IgG1 PE mouse

CD46 BD E4.3 IgG2a FITC mouse

CD56 CLB B159 IgG1 PE mouse

CD61 BD RUU-PL7F12 IgG1 FITC mouse

CD62P CLB CLB-tromb/6, C2 IgG1 PE mouse

CD71 CLB RVS10 IgG1 PE mouse

CD90 BD 5E10 IgG1 PE mouse

CD105 DC SN6h IgG1 - mouse

CD106 BD 51-10C9 IgG1 PE mouse

CXCR-4 (fusin) BD 12G5 IgG2a PE mouse

HLA-ABC BD G46-2.6 IgG1 FITC mouse

HLA-DR CLB CLB-HLA-Dr, 1E5 IgG1 FITC mouse

VEGFR-2 (Flk-1) SC A-3 IgG1 - mouse

keyhole limpet hemocyanin BD X40 IgG1 PE/FITC mouse

keyhole limpet hemocyanin BD X39 IgG2a PE mouse

mouse IgG1 BD X56 IgG1 PE rat

mouse IgG BD - - FITC goat

The immunoglobulins directed against keyhole limpet hemocyanin were used as isotype- matched control antibodies. BD, Becton Dickinson; CLB, Sanquin; DC, DakoCytomation; SC, Santa Cruz; PE, phycoerythrin; FITC, ﬂuorescein isothiocyanate. All antibody preparations were used at the concentrations recommended by the suppliers.

(8)

33

Reverse transcription-polymerase chain reaction analysis

hMSCs and hDFs were either mock-infected or infected with 100 IU/cell of hAd5/

F50.CMV.myocL or hAd5/F50.empty and plated at a density of 4×10³ cells/cm². At 24 hours postinfection, the inoculum was replaced by fresh medium. Total cellular RNA was extracted from the cells 7 days after transduction. Next, cDNA was synthesized in 50-μl volumes using 2 μg of RNA, 0.25 μg of random hexanucleotides, 25 nmol of each dNTP, and 500 U of SuperScript II RNase H- reverse transcriptase (Invitrogen). To detect cardiac, skeletal, and smooth muscle-specific mRNAs, 1 μl of cDNA was subjected to PCR using 2.5 U of SuperTaq (HT Biotechnology) and 10 pmol of the primer pairs listed in Table 3. The amplification scheme consisted of a 2-minute incubation at 94°C, followed by 24 to 40 cycles of 15 seconds at 94°C (denaturation step), 30 seconds at 60 to 64°C (annealing step), and 30 seconds at 72°C (extension step; see Table 3). The forward primer in each set always targeted another exon than the reverse primer. As internal controls for the quantity and quality of the RNA specimens, reverse transcription-PCR (RT-PCR) amplifications targeting transcripts of the housekeeping gene glyceraldehyde-3- phosphate dehydrogenase (GAPDH) were performed in parallel. Total RNA samples derived from human atrium, ventricle, umbilical cord smooth muscle, and vastus lateralis skeletal muscle, and from a human skeletal myoblast cell line before and after myotube formation were also subjected to RT-PCR to provide positive controls and to determine the specificity of the various marker genes assayed.

PCRs carried out with water instead of cDNA served as negative controls. The PCR products were separated in agarose gels containing 1 μg/ml ethidium bromide and visualized with the aid of a Gel Doc 2000 digital imaging system and Quantity One software (both from BioRad).

Immunoﬂuorescence microscopy

For immunostaining, hMSCs were transduced with hAd5/F50.CMV.myocL or hAd5/

F50.empty and plated on glass coverslips coated with fibronectin (Sigma-Aldrich) at a density of 3×10³ cells/cm². Seven days later, the cells were rinsed three times with PBS and fixed with 10% (v/v) phosphate-buffered formaldehyde solution (J.T. Baker) at room temperature for 30 minutes. After fixation, the cells were washed thrice with PBS containing 10 mM glycine (PBSG) and permeabilized with 1% Triton X-100 (Sigma) in PBS for 5 minutes at ambient temperature.

Following three washes with PBSG aspeciﬁc antibody binding sites were blocked by incubating with 5% FBS in PBS (PBSF) at 4°C for at least 30 minutes. Cells

Table 2. Primers used for nucleotide sequence analysis of human myocardin cDNA ctccattggtcttccatagcactg

cgggatccggcagcctatgacatca atccacgctgttttga

gctgcaataaacaagtt acgccactgagcaatacc tcataggatggaggctgt ggtccattccaactgctcagatgaag cagtggcgttgaagaagagttt

Primer sequences are given in the 5’ to 3’ direction.

(9)

34

were labeled for 2 to 16 hours at 4°C with mouse monoclonal antibodies (MAbs) directed against sarcomeric

α

-actin, sarcomeric

α

-actinin, sarcoplasmic reticulum Ca(2+)-ATPase 2a (SERCA2a), or atrial natriuretic peptide precursor A (ANF) and polyclonal antisera specific for slow skeletal troponin I (ssTnI), smMHC¹⁸, or GATA4. All antibody preparations were diluted 100-fold in PBSF before application to the cells, except for the anti-ANF and anti-GATA4 antibodies, which were used at a 10-fold dilution. For more information about these antibodies see Table 4. Excess primary antibody was removed by three washes with PBSG and the cells were stained at 4°C for 1 hour with highly cross-absorbed affinity-purified AlexaFluor488-conjugated goat anti-mouse IgG (1:100; Molecular Probes), AlexaFluor555-conjugated rabbit anti-goat IgG (1:100; Molecular Probes), or AlexaFluor568-conjugated goat anti-rabbit IgG (1:100; Molecular Probes). The coverslips were washed thrice with PBSG followed by a 5-minute staining of the nuclei with 10 μg/ml Hoechst 33342 (Molecular Probes) in PBS and four more washes with PBSG. The coverslips were mounted onto glass slides in Vectashield (Vector Laboratories). Double labeling experiments were performed in a similar manner except that the coverslips were simultaneously incubated with the rabbit anti-smMHC antiserum in combination with the sarcomeric

α

-actin-, sarcomeric

α

^-

actinin-, SERCA2a-, or ANF-speciﬁc MAbs. To visualize the primary antibodies, the cells were incubated with a mixture of the aforementioned secondary antibodies.

Slides were inspected using an Olympus IX51 inverse ﬂuorescence microscope, recorded using a Peltier-cooled digital camera (ColorView II), and processed using AnalySIS software (Soft Imaging Systems). Images were merged in Adobe Photoshop version 5.0.

Table 3. Primer sequences and PCR conditions used for RT-PCR analysis

product (size in bp) Ta cycles forward primer (5’ to 3’) reverse primer (5’ to 3’) GAPDH (302) 62 25 agccacatcgctcagacacc gtactcagcgccagcatcg myocardin (1141, 997) 63 35 ggactgctctggcaacccagtgc catctgctgactccgggtcatttgc cTnI (250) 61 35 ccctgcaccagccccaatcaga cgaagcccagcccggccaact cTnT (150) 61 32 ggcagcggaagaggatgctgaa gaggcaccaagttgggcatgaacga Mlc2a (381) 61 35 aaggtgaagtgtcccagagg acagagtttattgaggtgcccc Mlc2v (382) 60 40 ggtgctgaaggctgattacgtt tattggaacatggcctctggat αMHC (413) 61 35 gtcattgctgaaaccgagaatg agacatcgcactgactgagaac βMHC (118) 55 35 gatcaccaacaacccctacg gaagcccagcacatcaaaag dHand (258) 60 35 ccgacgtgaaagaggagaag ggatgattccaaatgcaagg eHand (342) 60 40 ggagacgcactgagagcatt cggctcactggtttaactcc GATA4 (475) 61 35 gacgggtcactatctgtgcaac agacatcgcactgactgagaac Mef2C (300) 60 35 ctgggaaaccccaacctatt gctgcctggtggaataagaa Nkx2.5 (279) 60 40 tctatccacgtgcctacagc agatcttgacctgcgtggac ANF (406) 63 35 gaaccagaggggagagacagag ccctcagcttgctttttaggag Cx43 (232) 60 35 aggcgtgaggaaagtaccaa acaccttccctccagcagtt SERCA2a (186) 60 25 ggtgctgaaaatctccttgc atcagtcatgcacagggttg ASMA (965) 60 30 ccagctatgtgtgaagaagagg gtgatctccttctgcattcggt CNN1 (671) 60 30 gagtgtgcagacggaacttcagcc gtctgtgcccagcttggggtc SM22 (928) 60 30 cgcgaagtgcagtccaaaatcg gggctggttcttcttcaatgggc smMHC (479) 60 24 cagatccgagctcgccat ccgagtagatgggcaggtgt fsTnI (201) 60 38 cctgaagcaggtcaagaagg tctgggtgcatctccctagt skMHC (167) 60 28 gaaattacttctggcaaaatacagg gcagatgccagttttccagt ssTnI (310) 64 35 aacccaagatcactgcctcccg cgcttgaacttcccacggaggt bp, base pairs; Ta, annealing temperature. For an explanation of the abbreviations of the marker genes, see the main text of the paper.

(10)

35

Cell viability assay

To compare the viability of untreated hMSCs with that of hAd5/F50.empty- and hAd5/F50.CMV.myocL-transduced hMSCs, unﬁxed populations of these cells were consecutively incubated with propidium iodide (Becton Dickinson) and Hoechst 33342 and subjected to ﬂow cytometric analysis, essentially as described by Darzynkiewicz et al.¹⁹.

Results

Expression of human myocardin

The currently available data indicate that there are at least three alternatively spliced variants of human myocardin, tentatively designated myocardin-A, myocardin-B, and myocardin-C⁵. The two largest splice variants of myocardin were detected in human atrial, ventricular, and smooth muscle tissue by RT-PCR using forward and reverse primers targeting exons 10 and 13, respectively (Figure 1). A similar result was obtained using RNA extracted from human skeletal muscle as template although the relative yield of the myocardin-speciﬁc PCR fragments was much lower than for the other muscle tissues (Figure 1). The preponderance of the largest splice variant of myocardin (i.e. myocardin-B) in cardiac and smooth muscle tissue from human adults (Figure 1) ﬁts with previous observations in the mouse⁷. It is possible that the myocardin transcripts in the skeletal muscle specimen are derived from vascular smooth muscle cells. We therefore also investigated the presence of myocardin mRNA in human skeletal myoblasts¹⁶ before and after differentiation into myotubes. No myocardin expression was detected in these samples, corroborating earlier claims that skeletal muscle cells do not express myocardin1, 3, 6, 10.

Table 4. Antibodies used for the detection of cardiac and smooth muscle proteins

antigen vendor clone (iso)type label species

sarcomeric α-actin SA 5C5 IgM - mouse

sarcomeric α-actinin SA EA-53 IgG1 - mouse

SERCA2a AB 2 A7-A1 IgG2a - mouse

ANF CH 23/1 IgG1 - mouse

ssTnI SC - - - rabbit

smMHC - - - - rabbit

GATA4 SC - - - goat

rabbit IgG MP - - Alexa568 goat

mouse IgG MP - - Alexa488 goat

goat IgG MP - - Alexa555 rabbit

SA, Sigma-Aldrich; AB, Afﬁnity BioReagents; CH, Chemicon; SC, Santa Cruz; MP, Molecular Probes. The smMHC-speciﬁc monoclonal antibody was a gift of R.S. Adelstein¹⁸.

(11)

36

Immunophenotypic characterization of hMSCs

Adherent cells derived from the BM of several adult human donors were expanded ex vivo and characterized by immunostaining and ﬂow cytometry. In agreement with previous reports¹⁴, the cells in each sample that we analysed expressed very high levels of the hyaluronate receptor (CD44), the major T-cell antigen (Thy-1;

CD90), endoglin (CD105), the vascular cell adhesion molecule 1 (VCAM-1; CD106), and human leucocyte class I antigens (HLA-ABC). The cells also expressed low levels of the transferrin receptor (CD71), P-selectin (CD62P), β3 integrin (CD61), the neural cell adhesion molecule (NCAM; CD56), and the membrane cofactor protein of the complement system (CD46). Signiﬁcantly, the BM cells expanded in culture did not express the hematopoietic markers CD45, CD34, CD19, CD14, and the vascular endothelial growth factor receptor 2 (VEGFR-2; Flk-1) and also stained negative for CD1a, CD10, human leucocyte class II subtype DR antigens (HLA-DR), and the CXC motif chemokine receptor 4 (CXCR-4; fusin).

Efﬁcient transduction of hMSCs and hDFs by hAd5F50 vectors

We tested the wild-type and fiber-modified hAd5 vectors hAd5/F5.CMV.eGFP and hAd5/F50.CMV.eGFP for their ability to transduce hMSCs and hDFs. Cells were infected with different vector doses and analysed by flow cytometry. hAd5/F50.

CMV.eGFP transduced both hMSCs and hDFs far more efﬁciently than hAd5/

F5.CMV.eGFP (Figure 2). A dose of 100 IU of hAd5/F50.CMV.eGFP per cell was found optimal as it resulted in high-level eGFP expression in virtually 100% of the

SK hDFs

hMSCs

U E M U E M

AT SMNC VE

1 23 4 5 6 7 8 9 10 11 12 13 14

myocardin-B

1 23 4 5 6 7 8 9 10 12 13 14

myocardin-A

F R

F 1141 bp R

997 bp

MB MT

GAPDH myocardin-B myocardin-A

Figure 1. RT-PCR analysis of myocardin expression. A, Schematic representation of myocardin splice variants. Myocardin-B consists of 14 exons and myocardin-A contains 13 exons. The position of the forward (F) and reverse (R) primers used for RT-PCR as well as the expected lengths of the PCR products are indicated. The myocardin-A-speciﬁc PCR fragment is 144 bp smaller than the ampliﬁcation product of myocardin-B due to the lack of exon 11. B, Left panels: RT-PCR analysis of myocardin expression in untreated (U) hMSCs and hDFs and in hAd5/F50.empty (E)- or hAd5/F50.CMV.myocL (M)-infected cells using total RNA extracted 7 days after vector addition. Right panels: myocardin expression in atrium (AT), smooth muscle (SM), ventricle (VE), skeletal myoblasts (MB), skeletal myotubes (MT), and skeletal muscle (SK). PCRs carried out with water instead of cDNA served as negative control (NC).

A

B

(12)

37

cells without noticeable toxicity.

On the basis of the aforementioned results, a hAd5/F50 vector expressing the largest splice variant of myocardin was generated. Transduction of hMSCs or hDFs with 100 IU/cell of this so-called hAd5/F50.CMV.myocL vector resulted in high- level expression of myocardin-B as determined by RT-PCR (Figure 1). Interestingly, relatively low amounts of myocardin-A and -B transcripts were also detected in untreated hMSCs and hDFs cells and in cells infected with the negative control vector hAd5/F50.empty (Figure 1). Analysis of nuclear morphology and propidium iodide and Hoechst 33342 uptake revealed that transduction of hMSCs with hAd5/

F50.CMV.myocL or hAd5/F50.empty did not result in increased cell death by either apoptosis or necrosis as compared to untreated hMSCs (Figure 4, 5, 6, and data not shown).

Forced myocardin expression activates heart muscle genes

We ﬁrst investigated the ability of myocardin to induce in hMSCs and hDFs the transcription of genes that are active in human cardiomyocytes (Figure 3A).

Mock- or hAd5/F50.empty-transduced hMSCs and hDFs did not express detectable amounts of the genes encoding cardiac troponin I (cTnI), cardiac troponin T (cTnT), the atrial and ventricular forms of myosin light chain 2 (Mlc2a and Mlc2v, respectively), cardiac

α

- and β-myosin heavy chain (

α

MHC and βMHC, respectively), eHand, GATA4, Nkx2.5, ANF, and SERCA2a. Transduction of hMSCs and hDFs with hAd5/F50.CMV.myocL, however, resulted in the activation of all

% eGFP-positive cells

F5 F50

hMSCs hDFs

IU / cell IU / cell

0 20 40 60 80 100

0 25 50 75 100 125 150 175 200

0 20 40 60 80 100

0 25 50 75 100 125 150 175 200

0 1000 2000 3000 4000 5000

0 25 50 75 100 125 150 175 200

0 1000 2000 3000 4000 5000

0 25 50 75 100 125 150 175 200

mean fluorescence intensity

F5 F50

Figure 2. Flow cytometric analysis of hMSCs and hDFs infected with hAd5/F5.CMV.eGFP (F5) or hAd5/F50.CMV.eGFP (F50). The percentage of eGFP-positive cells and the mean ﬂuorescence intensity (arbitrary units) of the cells are plotted against the vector dose.

Values are given as means ± standard deviation (n=4).

(13)

38

these genes with the exception of cTnI and Nkx2.5. GATA4 staining was observed both in the nucleus and the cytoplasm of hAd5/F50.CMV.myocL-transduced (Figure 5). The presence of GATA4 in the cytosol has been previously reported by Wang et al.²⁰.

Interestingly, we found that both hMSCs and hDFs naturally express the genes encoding dHand, Mef2C, and connexin 43 (Cx43). Cx43 expression in hMSCs and hDFs has been previously reported^{21, 22}.

Forced myocardin expression induces smooth muscle genes

To evaluate the induction of smooth muscle gene expression by myocardin, the RNA samples were analyzed using primer pairs speciﬁc for the ASMA, smooth muscle calponin (CNN1), SM22, and smMHC genes (Figure 3B). Each of these genes was expressed in arterial smooth muscle from umbilical cord. Low amounts of CNN1-, SM22-, and smMHC-speciﬁc transcripts were also detected in the myocardial samples, most likely due to the presence of blood vessels in these clinical specimens. No evidence for ASMA expression in heart muscle was obtained.

However, ASMA-encoding transcripts were clearly present in the three samples representative of skeletal muscle. The SM22 gene was expressed in cultured myoblasts and myotubes. No SM22-speciﬁc mRNA was found in vastus lateralis skeletal muscle consistent with the reported expression of SM22 in developing but not mature striated muscles²³. Analysis of the RNA extracted from the mock- treated hMSCs revealed that expression of the relatively widely expressed ASMA, CNN1, and SM22 genes but not of the highly speciﬁc smooth muscle marker gene smMHC²³ is intrinsic to these cells. Transduction of the hMSCs with hAd5/F50.

CMV.myocL led to the accumulation of smMHC-speciﬁc transcripts in these cells.

With the hDFs somewhat different results were obtained. Of the four selected smooth muscle marker genes, these cells naturally expressed only CNN1 and SM22. Forced myocardin expression in the hDFs resulted in the accumulation of ASMA and smMHC mRNA.

Forced myocardin expression does not activate skeletal muscle genes

Finally, we examined whether myocardin could also transactivate skeletal muscle genes in hMSCs and hDFs. To this end, RT-PCR amplifications were carried out with primers specific for the fast-twitch skeletal troponin I (fsTnI), skeletal myosin heavy chain 2a (skMHC), and slow-twitch skeletal troponin I (ssTnI) genes (Figure 3C). Whereas in humans the expression of the fsTnI and skMHC genes is restricted to skeletal muscle, the ssTnI gene is active not only in skeletal muscle but also in embryonic, fetal, and neonatal myocardium²⁴. Each of these genes was expressed in the vastus lateralis skeletal muscle, the myotubes, and the myoblasts. The RT- PCR analyses of the myocardial RNA samples showed that the right atrium and left ventricle contain ssTnI- but not fsTnI- or skMHC-encoding transcripts. The smooth muscle specimen, on the other hand, did not yield fsTnI-, skMHC-, or ssTnI- specific RT-PCR fragments. Infection of hMSCs and hDFs with hAd5/F50.CMV.

myocL induced the expression of ssTnI, but not of the fsTnI or skMHC genes.

(14)

39

Myocardin activates both cardiac and smooth muscle genes in single cells.

To conﬁrm the induction of cardiac and smooth muscle genes at the protein level, hAd5/F50.CMV.myocL-transduced hMSCs were analyzed by IFM using antibodies speciﬁc for sarcomeric

α

-actin, sarcomeric

α

-actinin, SERCA2a, ANF, ssTnI, smMHC (Figure 4), and GATA4 (Figure 5). As mentioned above and shown in Figure 3B, the smMHC protein is a highly speciﬁc marker for smooth muscle cells.

The sarcomeric

α

-actin and

α

-actinin proteins have been found in both skeletal and cardiac but not in smooth muscle²⁵. We (Figure 3C) and others (vide supra) have shown that the expression of the ssTnI gene is also conﬁned to striated muscle. In contrast, SERCA2a molecules are found in all three muscle types²⁶, while the ANF and GATA4 genes are expressed in cardiac but not in smooth or

E M E M

hMSCs hDFs AT SM NC VE MB MT SK

GAPDH

GATA4 cTnI

Mlc2a Mlc2v

ANF AMHC

SERCA2a Cx43 cTnT

smMHC ASMA CNN1 SM22

skMHC fsTnI ssTnI eHand dHand

Nkx2.5 BMHC

Mef2C

A

B

C

Figure 3. RT-PCR analysis of muscle marker genes in hMSCs, hDFs, and control tissues.

Total RNA extracted from mock (-)-, hAd5/F50.empty (E)-, or hAd5/F50.CMV.myocL (M)- transduced cells 7 days after (mock)-infecton was subjected to RT-PCR using cardiac (A), smooth (B), and skeletal (C) muscle gene-speciﬁc primers. The abbreviations of the marker genes are explained in the main text.

(15)

40

skeletal muscle (Figure 3). The immunostainings revealed that approximately 10%

of the infected hMSCs stained positive for each of these proteins, while none of the mock- (data not shown) or hAd5/F50.empty-transduced cells exhibited speciﬁc labeling with any of the antibodies (Figure 4). The number of ﬂuorescent cells was comparable at one, two, and three weeks after infection with Ad5/F50.CMV.

myocL. The hMSCs that expressed the aforementioned proteins did not develop well-organized sarcomeres and did not display spontaneous beating within the period of observation.

The results of the single antibody staining experiments raised the question whether cardiac and smooth muscle marker genes are expressed by the same or different cells in the hAd5/F50.CMV.myocL-treated hMSCs cultures. To investigate this issue, we performed double labeling experiments with MAbs directed against sarcomeric

α

-actin, sarcomeric

α

-actinin, SERCA2a, or ANF in combination with polyclonal antibodies recognizing smMHC (Figure 6). These experiments showed that the smMHC-speciﬁc antibodies labeled all hMSCs that stained positive for sarcomeric

α

-actin, sarcomeric

α

-actinin, SERCA2a, or ANF. We hence conclude that these cells had acquired characteristics of both cardiac and smooth muscle cells.

Discussion

In this paper, we have demonstrated that forced expression of human myocardin is sufficient to induce the expression of a broad range of cardiac and smooth muscle genes in human adult stem cells and primary human fibroblasts. These findings support the concept that myocardin plays an important role in cardiomyogenesis and vasculogenesis and are consistent with the cardiac and smooth muscle- restricted expression of myocardin^{1, 3-7}. Furthermore, IFM studies revealed that under our experimental conditions myocardin simultaneously activates (in single cells) both heart and smooth muscle genes. Accordingly, other factors are required to delimit the cardiac and smooth muscle transcriptomes.

While in our experiments myocardin acted as a potent inducer of both cardiomyocyte- and smooth muscle-speciﬁc genes, two previous papers only reported on the ability of myocardin to regulate smooth muscle gene expression^4,

6. Furthermore, in a detailed study carried out with murine embryonic ﬁbroblasts (i.e. 10T1/2 cells) ectopic myocardin expression caused the activation of smooth muscle genes as well as the skeletal

α

-actin gene but not of cardiomyocyte- speciﬁc genes¹¹. In yet another study, forced expression of myocardin in A404 cells, which are derived from mouse embryonic teratocarcinoma cells and readily differentiate into smooth muscle cells, resulted in the accumulation of cardiac and skeletal

α

-actin mRNAs besides smooth muscle gene-speciﬁc transcripts but did not lead to GATA4 expression⁹. There are several possible explanations why we found ample evidence for the activation of endogenous cardiac genes by myocardin whereas others did not: (i) In the studies of Du, et al.⁴ and Yoshida, et al.⁶ the activation of cardiac muscle genes was not investigated. (ii) We are the ﬁrst to test the effects of human myocardin in human cells. Species-related differences in myocardin activity may hence account for the observed differences in target gene

(16)

41 hAd5/F50.CMV.myocL hAd5/F50.empty

Figure 4. Immunoﬂuorescent staining of hMSCs transduced with hAd5/F50.CMV.myocL or hAd5/F50.empty. Blue, nuclei; green, sarcomeric α-actin (A), sarcomeric α-actinin (B), SERCA2a (C), or ANF (D); red, ssTnI (E) or smMHC (F).

A

B

C

D

E

F

(17)

42

speciﬁcity. (iii) In none of the previous studies the transactivating potential of the myocardin isoform encoded by splice variant B of myocardin⁵ was investigated.

Perhaps, the additional 48 amino acid residues encoded by this isoform make it a stronger inducer of cardiomyocyte-speciﬁc genes than the myocardin-A-encoded protein. Moreover, as far as we could deduce, in all previous studies focusing on the ability of myocardin to switch on endogenous genes, an amino-terminally truncated version of the protein has been used. The amino terminus of myocardin contains RPEL motifs. For MAL (also known as BSAC, MLK1, and MRTF-A), a protein closely related to myocardin, these motifs have been shown to control its intracellular distribution and therefore its transcriptional activity through their interaction with G actin²⁷. It is conceivable that the RPEL domains of myocardin are also involved in the regulation of its function. (iv) The developmental options of a cell are limited by the methylation pattern and chromatin structure of its DNA and by the identity, concentration, and activity of its transcriptional regulators.

Accordingly, the cell type in which the myocardin gene is expressed may have a large impact on the outcome of the experiments. Human adult stem cells and primary human ﬁbroblasts may have a epigenetic make-up that allows them to enter into a cardiomyocyte-speciﬁc gene expression pattern much easier than some other (immortal) cell types.

In Figure 4, 5 and 6, we showed that after infection of hMSCs with 100 IU of hAd5/F50.CMV.myocL per cell, a vector dose sufﬁcient to transduce the entire cell population (Figure 2), only approximately 10% expresses cardiac and smooth muscle genes. Although the reason of this apparent discrepancy is unclear, we have some preliminary data showing that the phenomenon is cell type-dependent (data not shown).

To determine the extent to which myocardin induces the differentiation of hMSCs and hDFs into cardiac, skeletal, or smooth muscle cells, we analyzed the expression of marker genes for each of these muscle lineages. A limitation of this approach is the considerable overlap between the gene expression proﬁles of cardiac, skeletal, and smooth muscle cells. Most of the genes that have been used as smooth muscle markers, are for instance, also expressed in developing heart and skeletal muscle (see Figure 3A-C and²³) and many myoﬁbrillar proteins are present in the sarcomeres of both cardiac and skeletal muscle²⁵. Another confounding factor may be the effect of tissue culture on the cells. Fibroblasts

Figure 5. Immunoﬂuorescent staining of hMSCs transduced with hAd5/F50.CMV.myocL or hAd5/F50.empty using a GATA4-speciﬁc polyclonal antiserum. Blue, nuclei; red, GATA4.

hAd5/F50.CMV.myocL hAd5/F50.empty

(18)

43

Chapter 2Activation of cardiac and smooth muscle-speciﬁc genes in primary human cells after forced expression of human myocardin hAd5/F50.CMV.myocL

A-actin+smMHC A-actin smMHC

smMHC A-actinin

smMHC SERCA2a

A-actin+smMHC

A-actinin+smMHC A-actinin+smMHC

SERCA2a+smMHC SERCA2a+smMHC

smMHC

ANF ANF+smMHC ANF+smMHC

hAd5/F50.empty

Figure 6. Immunoﬂuorescent double staining of hMSCs transduced with hAd5/F50.CMV.

myocL or hAd5/F50.empty. Blue, nuclei; green, sarcomeric α-actin, sarcomeric α-actinin, SERCA2a, or ANF; red, smMHC.

(19)

44

and mesenchymal stem cells have been reported to adopt a myofibroblast- or smooth muscle cell-like phenotype during in vitro culture^28-31. This would explain the presence of ASMA-specific transcripts in hMSCs and CNN1 and SM22 mRNAs in hMSCs and hDFs (Figure 3B). The consequences of the adaptation of these cells to the in vitro culture conditions for their ability to form cardiac, skeletal, or smooth muscle cells are unknown. Nonetheless, smMHC can be considered as a highly specific marker for smooth muscle cells²³. Likewise, cTnI-, Mlc2a-, GATA4-, and ANF-specific transcripts have not been detected in either skeletal or smooth muscle. In addition, fsTnI and skMHC expression are truly confined to skeletal muscle. The fact that the latter two genes are not switched on in hAd5/F50.CMV.

myocL-transduced hMSCs and hDFs, while expression of Mlc2a, GATA4, ANF, and smMHC is clearly induced, leaves us with no other conclusion than that myocardin transactivates both cardiac and smooth muscle genes but not (genuine) skeletal muscle genes. This conclusion is at variance with that of Yoshida, et al.⁹ who interpreted the myocardin-dependent activation of the cardiac and skeletal

α

^-actin

genes in A404 cells as evidence for the inappropriate induction by myocardin of cardiac and skeletal CArG box-dependent genes in cultured smooth muscle cells.

However, since the skeletal

α

-actin gene is expressed in murine and rat hearts in a developmentally regulated fashion with a rapid decrease of the amount of skeletal

α

-actin-specific transcripts in this organ after birth²⁵, the activation by myocardin of an embryonic heart muscle expression program may provide an alternative explanation for the results of Yoshida, et al.⁹. This would also explain our finding that in hAd5/F50.CMV.myocL-transduced hMSCs and hDFs ssTnI instead of cTnI gets activated (Figure 3C). The ASMA, CNN1, and SM22 genes have been shown to be (transiently) expressed in the developing heart of mammals as well²³. The activation of these genes after infection of hDFs with hAd5/F50.CMV.myocL could therefore be another indication for the transactivation of early cardiac muscle genes by myocardin. The co-labeling of hAd5/F50.CMV.myocL-transduced hMSCs with antibodies directed against the most explicit smooth muscle marker smMHC and antibodies specific for sarcomeric

α

-actin, sarcomeric

α

-actinin, or ANF is, however, hard to reconcile with myocardin’s activity being restricted to genes expressed in embryonic cardiomyocytes. We hence favor the idea that myocardin is a transcriptional (co)activator of both cardiac and smooth muscle genes but does not regulate skeletal muscle-speciﬁc gene expression and propose that myocardin cooperates with other factors to establish either a heart of smooth muscle phenotype.

As shown in Figure 3A, we could not detect Nkx2.5-speciﬁc transcripts in hAd5/

F50.CMV.myocL-transduced hMSCs or hDFs, although the Nkx2.5 promoter is activated by myocardin in reporter gene assays¹ and the Nkx2.5 protein is involved in the regulation of myocardin expression⁷. Whether Nkx2.5 plays a role in limiting the transcription-enhancing activity of myocardin to heart muscle-speciﬁc genes remains to be investigated.

The ability of myocardin to selectively activate cardiomyocyte- and smooth muscle- speciﬁc genes may allow for its therapeutic application. There is currently a large interest in the use of hMSCs to replenish cardiomyocytes that have been lost during an ischemic insult and to improve myocardial vascularization in patients suffering from coronary artery disease¹⁴. Previous studies have shown that the spontaneous

(20)

45

differentiation of transplanted BM-derived stem cells into cardiomyocytes is inefﬁcient^{14, 32} and that differentiation into potentially undesirable cell types such as ﬁbroblasts occurs^{32, 33}. We envision that treatment of hMSCs with hAd5/F50.

CMV.myocL before transplantation may enhance their propensity to differentiate into cardiac and smooth muscle cells in vivo.

Acknowledgements

The authors are indebted to Binie Klein for supplying the VH10 cells, Didier Trono for making available the conditionally immortalized dystrophin-negative human myoblasts, and Pieter Doevendans, Robbert Klautz, and Rob Nelissen for providing left over surgical material. We also thank Robert Adelstein for donating the smMHC-specific rabbit antiserum and Marie-José Goumans for supplying the βMHC-specific primers. Finally, Crucell N.V. is gratefully acknowledged for sharing their adenovirus vector production system. Part of this work was financed by the Netherlands Organisation for Health Research and Development (ZonMw).

(21)

46

References

1. Wang DZ, Chang PS, Wang Z, Sutherland L, Richardson JA, Small E, Krieg PA, Olson EN. Activation of cardiac gene expression by myocardin, a transcriptional cofactor for serum response factor. Cell. 2001;105(7):851- 862.

2. Miano JM. Serum response factor: toggling between disparate programs of gene expression. J. Mol. Cell Cardiol. 2003;35(6):577-593.

3. Wang DZ, Li S, Hockemeyer D, Sutherland L, Wang Z, Schratt G, Richardson JA, Nordheim A, Olson EN. Potentiation of serum response factor activity by a family of myocardin- related transcription factors.

PNAS. 2002;99(23):14855-14860.

4. Du KL, Ip HS, Li J, Chen M, Dandre F, Yu W, Lu MM, Owens GK, Parmacek MS.

Myocardin is a critical serum response factor cofactor in the transcriptional program regulating smooth muscle cell differentiation. Mol. Cell Biol.

2003;23(7):2425-2437.

5. Torrado M, López E, Centeno A, Medrano C, Castro-Beiras A, Mikhailov AT. Myocardin mRNA is augmented in the failing myocardium: expression proﬁling in the porcine model and human dilated cardiomyopathy. J Mol Med. 2003;81(9):566-577.

6. Yoshida T, Sinha S, Dandr‚ F, Wamhoff BR, Hoofnagle MH, Kremer BE, Wang DZ, Olson EN, Owens GK. Myocardin is a key regulator of CArG- dependent transcription of multiple smooth muscle marker genes. Circ.

Res. 2003;92(8):856-864.

7. Ueyama T, Kasahara H, Ishiwata T, Nie Q, Izumo S. Myocardin expression is regulated by Nkx2.5, and its function is required for cardiomyogenesis.

Mol Cell Biol. 2003;23(24):9222-9232.

8. Du KL, Chen M, Li J, Lepore JJ, Mericko P, Parmacek MS. Megakaryoblastic leukemia factor-1 transduces cytoskeletal signals and induces smooth muscle cell differentiation from undifferentiated embryonic stem cells. J Biol Chem. 2004;279(17):17578-17586.

9. Yoshida T, Kawai-Kowase K, Owens GK. Forced expression of myocardin is not sufﬁcient for induction of smooth muscle differentiation in multipotential embryonic cells. Arterioscler Thromb Vasc Biol. 2004;24(9):1596-1601.

10. Chen J, Kitchen CM, Streb JW, Miano JM. Myocardin: a component of a molecular switch for smooth muscle differentiation. J. Mol. Cell Cardiol.

2002;34(10):1345-1356.

11. Wang Z, Wang DZ, Pipes GCT, Olson EN. Myocardin is a master regulator of smooth muscle gene expression. PNAS. 2003;100(12):7129-7134.

12. Wang Z, Wang DZ, Hockemeyer D, McAnally J, Nordheim A, Olson EN.

Myocardin and ternary complex factors compete for SRF to control smooth muscle gene expression. Nature. 2004;428(6979):185-189.

13. Li S, Wang DZ, Wang Z, Richardson JA, Olson EN. The serum response factor coactivator myocardin is required for vascular smooth muscle development. PNAS. 2003;100(16):9366-9370.

14. Pittenger MF, Martin BJ. Mesenchymal stem cells and their potential as cardiac therapeutics. Circ Res. 2004;95(1):9-20.

(22)

47

15. Klein B, Pastink A, Odijk H, Westerveld A, van der Eb AJ. Transformation and immortalization of diploid xeroderma pigmentosum ﬁbroblasts. Exp Cell Res. 1990;191(2):256-262.

16. Cudré-Mauroux C, Occhiodoro T, Konig S, Salmon P, Bernheim L, Trono D.

Lentivector-mediated transfer of Bmi-1 and telomerase in muscle satellite cells yields a Duchenne myoblast cell line with long-term genotypic and phenotypic stability. Hum Gene Ther. 2003;14(16):1525-1533.

17. Knaän-Shanzer S, van der Velde I, Havenga MJE, Lemckert AA, de Vries AAF, Valerio D. Highly efﬁcient targeted transduction of undifferentiated human hematopoietic cells by adenoviral vectors displaying ﬁber knobs of subgroup B. Hum Gene Ther. 2001;12(16):1989-2005.

18. Kawamoto S, Adelstein RS. Characterization of myosin heavy chains in cultured aorta smooth muscle cells. A comparative study. J Biol Chem.

1987;262(15):7282-7288.

19. Darzynkiewicz Z, Bruno S, Del Bino G, Gorczyca W, Hotz MA, Lassota P, Traganos F. Features of apoptotic cells measured by ﬂow cytometry.

Cytometry. 1992;13(8):795-808.

20. Wang J, Feng X-h, Schwartz RJ. SUMO-1 modiﬁcation activated GATA4- dependent cardiogenic gene activity. J Biol Chem. 2004;279(47):49091- 49098.

21. Krenacs T, Rosendaal M. Connexin43 gap junctions in normal, regenerating, and cultured mouse bone marrow and in human leukemias: their possible involvement in blood formation. Am J Pathol. 1998;152(4):993-1004.

22. Guo H, Acevedo P, Parsa FD, Bertram JS. Gap-junctional protein connexin 43 is expressed in dermis and epidermis of human skin: differential modulation by retinoids. J Invest Dermatol. 1992;99(4):460-467.

23. Miano JM. Mammalian smooth muscle differentiation: origins, markers and transcriptional control. Results Probl Cell Differ. 2002;38:39-59.

24. Murphy AM. Contractile protein phenotypic variation during development.

Cardiovasc Res. 1996;31 Spec No:E25-33.

25. Schiaffino S, Reggiani C. Molecular diversity of myofibrillar proteins: gene regulation and functional significance. Physiol Rev. 1996;76(2):371-423.

26. Lytton J, Zarain-Herzberg A, Periasamy M, MacLennan DH. Molecular cloning of the mammalian smooth muscle sarco(endo)plasmic reticulum Ca2+-ATPase. J Biol Chem. 1989;264(12):7059-7065.

27. Miralles F, Posern G, Zaromytidou AI, Treisman R. Actin dynamics control SRF activity by regulation of its coactivator MAL. Cell. 2003;113(3):329- 342.

28. Silva WA, Jr., Covas DT, Panepucci RA, Proto-Siqueira R, Siuﬁ JL, Zanette DL, Santos AR, Zago MA. The proﬁle of gene expression of human marrow mesenchymal stem cells. Stem Cells. 2003;21(6):661-669.

29. Masur SK, Dewal HS, Dinh TT, Erenburg I, Petridou S. Myoﬁbroblasts differentiate from ﬁbroblasts when plated at low density. PNAS.

1996;93(9):4219-4223.

30. Cai D, Marty-Roix R, Hsu HP, Spector M. Lapine and canine bone marrow stromal cells contain smooth muscle actin and contract a collagen- glycosaminoglycan matrix. Tissue Eng. 2001;7(6):829-841.

(23)

48

31. Kinner B, Zaleskas JM, Spector M. Regulation of smooth muscle actin expression and contraction in adult human mesenchymal stem cells. Exp Cell Res. 2002;278(1):72-83.

32. Wang JS, Shum-Tim D, Chedrawy E, Chiu RC. The coronary delivery of marrow stromal cells for myocardial regeneration: pathophysiologic and therapeutic implications. J Thorac Cardiovasc Surg. 2001;122(4):699- 705.

33. Wu GD, Nolta JA, Jin YS, Barr ML, Yu H, Starnes VA, Cramer DV. Migration of mesenchymal stem cells to heart allografts during chronic rejection.

Transplantation. 2003;75(5):679-685.

(24)

49

(25)