By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Due to essential maintenance the subscribe/renew pages will be unavailable on Wednesday 26 October between 02:00- 08:00 BST/ 09:00 – 15:00 SGT/ 21:00- 03:00 EDT. Apologies for the inconvenience.
National Research Laboratory for Cardiovascular Stem Cells, Seoul, Korea
Department of Internal Medicine, Seoul National University College of Medicine, Seoul, Korea
Innovative Research Institute for Cell Therapy, Seoul National University Hospital, Seoul, Korea
National Research Laboratory for Cardiovascular Stem Cells, Department of Internal Medicine, Seoul National University College of Medicine, Innovative Research Institute for Cell Therapy, Seoul National University Hospital, 28 Yongon-dong, Chongno-gu, Seoul 110-744, Republic of Korea. Telephone: 82-2-2072-2226; Fax: 82-2-766-8904
National Research Laboratory for Cardiovascular Stem Cells, Seoul, Korea
Department of Internal Medicine, Seoul National University College of Medicine, Seoul, Korea
Innovative Research Institute for Cell Therapy, Seoul National University Hospital, Seoul, Korea
National Research Laboratory for Cardiovascular Stem Cells, Department of Internal Medicine, Seoul National University College of Medicine, Innovative Research Institute for Cell Therapy, Seoul National University Hospital, 28 Yongon-dong, Chongno-gu, Seoul 110-744, Republic of Korea. Telephone: 82-2-2072-2226; Fax: 82-2-766-8904;
Stem cell transplantation in acute myocardial infarction (AMI) has emerged as a promising therapeutic option. We evaluated the impact of AMI on mesenchymal stem cell (MSC) differentiation into cardiomyocyte lineage. Cord blood-derived human MSCs were exposed to in vitro conditions simulating in vivo environments of the beating heart with acute ischemia, as follows: (a) myocardial proteins or serum obtained from sham-operated rats, and (b) myocardial proteins or serum from AMI rats, with or without application of oscillating pressure. Expression of cardiac-specific markers on MSCs was greatly induced by the infarcted myocardial proteins, compared with the normal proteins. It was also induced by application of oscillating pressure to MSCs. Treatment of MSCs with infarcted myocardial proteins and oscillating pressure greatly augmented expression of cardiac-specific genes. Such expression was blocked by inhibitor of transforming growth factor β1 (TGF-β1) or bone morphogenetic protein-2 (BMP-2). In vitro cellular and electrophysiologic experiments showed that these differentiated MSCs expressing cardiomyocyte-specific markers were able to make a coupling with cardiomyocytes but not to selfbeat. The pathophysiologic significance of in vitro results was confirmed using the rat AMI model. The protein amount of TGF-β1 and BMP-2 in myocardium of AMI was significantly higher than that in normal myocardium. When MSCs were transplanted to the heart and analyzed 8 weeks later, they expressed cardiomyocyte-specific markers, leading to improved cardiac function. These in vitro and in vivo results suggest that infarct-related biological and physical factors in AMI induce commitment of MSCs to cardiomyocyte-like cells through TGF-β/BMP-2 pathways.
Disclosure of potential conflicts of interest is found at the end of this article.
Author contributions: S.-A.C. and E.J.L.: conception and design, collection and/or assembly of data, data analysis and interpretation, manuscript writing; H.-J.K.: conception and design, data analysis and interpretation; S.-Y.Z., H.-J.C., W.-K.H., and Y.-B.P.: conception and design, collection and/or assembly of data, data analysis and interpretation; J.-H.K., L.L., S.-W.Y., and J.-W.S.: provision of study material or patients, collection and/or assembly of data, data analysis and interpretation; C.-S.L., K.-H.K., and J.-Y.W.: provision of study material or patients, collection and/or assembly of data; K.-W.P.: data analysis and interpretation; S.-E.Y., W.I.O., and Y.S.Y.: provision of study material or patients; H.-S.K.: conception and design, data analysis and interpretation, manuscript writing, final approval of manuscript.
Stem cell transplantation has emerged as a promising treatment modality for myocardial infarction. Animal studies and several clinical trials reported favorable outcomes of stem cell therapy in acute myocardial infarction (AMI) [1, –3]. In the environment of myocardium with acute ischemia, enhancement of angiogenesis has been previously studied and is relatively well understood. Ischemic myocardium itself produces several stem cell-attracting and survival-promoting cytokines or transcription factors, such as vascular endothelial growth factor and stromal-cell-derived factor 1 , which augment homing and survival of stem cells . However, our knowledge about the effects of acute ischemia on stem cell differentiation is still limited. Furthermore, exogenous stimuli to cells transplanted into myocardium are different from those into other organs, because cells in myocardium are subject to oscillating pressure that is a unique component experienced only in myocardial environment, and those mechanical stimuli can also affect the fate of stem cells.
We hypothesized that the myocardial environmental milieu in AMI would enhance the differentiation of mesenchymal stem cells (MSCs) into cardiomyocyte-like cells. In the present study, we evaluated the effects of infarcted myocardial proteins obtained from rat AMI and oscillating pressure change, which would mimic the myocardial environment of AMI settings, on the differentiation of cord blood-derived human MSCs into cardiomyocyte-like cells.
Materials and Methods
Isolation and Culture of Cord Blood-Derived Human Mesenchymal Stem Cells
The study protocol was approved by the institutional review board of Seoul National University Hospital. MSCs from human umbilical cord blood were provided by Medipost Inc. (Kyung-Gi do, Korea, http://www.medi-post.co.kr). MSCs were isolated, with informed consent, from human umbilical cord blood of women who had a delivery as previously described . In brief, mononuclear cells were isolated by centrifugation in a Ficoll-Hypaque gradient (density, 1.077 g/cm3; Sigma-Aldrich, St. Louis, http://www.sigmaaldrich.com); washed; and suspended in α-minimum essential medium (Gibco-BRL, Carlsbad, CA, http://www.gibcobrl.com) supplemented with 10% fetal bovine serum (FBS; HyClone, Logan, UT, http://www.hyclone.com), 10,000 units/ml penicillin G sodium, 10,000 μg/ml streptomycin sulfate, and 25 μg/ml amphotericin B (Invitrogen, Carlsbad, CA, http://www.invitrogen.com). Cultures were maintained in a humidified 37°C cell incubator with 5% CO2 with a change of culture medium twice a week. Three weeks later, a monolayer of fibroblast-like adherent cells was trypsinized (0.25% trypsin; HyClone), washed, resuspended in culture medium, and subcultured at a concentration of 5 × 106 cells per cm2.
These MSCs have self-renewal capability (Fig. 1A, 1B) and potencies to differentiate into multilineage. Flow cytometric analysis of MSCs isolated from human umbilical cord blood was performed to detect surface antigens. MSCs were uniformly positive for CD29 (β1-integrin), CD44 (hyaluronate receptor), SH2, SH3, CD90, and CD166 [7, –9]; these findings were identical to those in bone marrow-derived MSCs. In contrast, MSCs were negative for other markers of the hematopoietic lineage, including CD14, CD34, CD45, CD51–61, anti-human-specific leukocyte antigen (HLA)-DR, and CD106 (Fig. 1C). MSCs of passages 5–8 from three different donors were used for these experiments.
Rat Coronary Artery Ligation Model and Myocardial Protein Extraction
The Institutional Animal Care and Use Committee of Seoul National University approved all animal protocols, and all animals were housed in Laboratory Animal Care Animal Facility of the Clinical Research Institute of Seoul National University. Female Fisher-344 rats underwent surgery to induce AMI by ligature of the left coronary arteries or sham operation as described previously in our laboratory . Ligation of the proximal left anterior descending artery generated an irreversible pale area on the surface of the middle and apical portion of left ventricle. At 72 hours after surgery, after confirmation of myocardial infarction with regional wall motion abnormality by echocardiography, hearts were totally excised and rinsed with phosphate-buffered saline (PBS). After removal of both atria and right ventricle, whole myocardium of left ventricle, including infarct and peri-infarct tissue, was frozen with liquid nitrogen and ground with a pestle. For myocardial harvest from the left ventricle of sham-operated rats, the same procedures were done. Myocardial proteins were extracted by buffer (0.075 M potassium acetate, 0.3 M NaCl, 0.1 M l-arginine basic salt, 0.01 M EDTA-HCl, 0.25% Triton X-100, pH 7.4). Rat serum was harvested simultaneously with myocardium.
Oscillating Pressure Chamber
An oscillating pressure chamber was constructed with modifications as previously described . Intrachamber pressure was regulated with a check valve system, and the pressure level was continuously monitored. A piston connected to the chamber allowed generation of oscillating hyperbaric intrachamber pressure from 10 to 150 mmHg at a frequency of 150 beats per minute. The pressure chamber was placed into a humidified 37°C cell incubator with 5% CO2. The interplay between valves permitted a continuous exchange of air between intrachamber and atmosphere.
Differentiation Induction by Exposure to Combination of Myocardial Proteins and Oscillating Pressure
To evaluate the influences of AMI on differentiation, we evaluated the effects of exposure to a combination of proteins extracted from normal myocardium or myocardium with AMI harvested at the acute phase of infarction and cyclic pressure changes, which simulate the in vivo conditions of myocardial infarction, on the differentiation of MSCs into cardiomyocytes. Proteins extracted from myocardium with AMI (MI proteins) were added to culture medium at a concentration of 25 μg/ml, or serum was added to culture medium at a final concentration of 10% (vol/vol). In addition, oscillating pressure was applied with specially designed oscillating pressure chamber  to simulate intracardiac pressure change. Cells were harvested after 3 days of exposure.
Induction or Inhibition of MSC Differentiation by Modulating Transforming Growth Factor-β1/Bone Morphogenetic Protein-2 Signaling
To investigate the signaling pathway involved in differentiation of MSCs into cardiomyocyte lineage, we targeted transforming growth factor-β1 (TGF-β1) and bone morphogenetic protein (BMP-2), which play important roles in cardiomyogenesis in embryo [12, 13]. Signal disruption of TGF-β1 and BMP-2 was performed by 10 μg/ml latency-associated peptide (LAP) and noggin (R&D Systems Inc., Minneapolis, http://www.rndsystems.com), respectively, before differentiation induction with MI proteins, serum, and oscillating pressure. Stimulation with human recombinant TGF-β1 (R&D Systems) and BMP-2 (R&D Systems) was performed with cross-combination of (a) 5 ng/ml TGF-β1 and (b) 5 ng/ml BMP-2 in 10% FBS. Reverse transcription (RT)-polymerase chain reaction (PCR) for cardiac-specific gene expression was done after 72 hours with same methods as described above.
RNA was prepared from cultured human MSCs with Trizol (Invitrogen). RNA from human myocardium was used as a positive control. After reverse transcription, complementary DNA for cardiac-specific genes (α-myosin heavy chain [α-MHC], β-MHC, connexin 43, cardiac troponin T [cTnT], GATA4, and atrial natriuretic peptide [ANP]) and glyceraldehyde-3-phosphate dehydrogenase were synthesized from extracted RNA. RT-PCR was performed with reverse transcription kit (Promega, Madison, WI, http://www.promega.com). To quantify the RT-PCR data, the hybridization signals were photometrically evaluated using TINA software (Raytest Isotopenmessgeräte, Straubenhardt, Germany, http://www.raytest.de). RT-PCR primers sequences and conditions were used in RT-PCR as described in supplemental online Table 1. These primers were human-specific, and there was no cross-reactivity with rat heart when tested (data not shown).
Dye Transfer Experiment for Analysis of Gap Junction Formation Between Cells
Dye transfer was used to assess the formation of gap junctional communication between differentiated MSCs . MSCs were preloaded with 2 μg/ml 1,1-dioctadecyl-3,3,3,3-tetramethylindocarbocyanine perchlocate (DiI) (Molecular Probes, Eugene, OR, http://probes.invitrogen.com) and 5 μg/ml calcein acetoxymethyl-ester (Molecular Probes) and cultured with unlabeled MSCs in the presence of an oscillating pressure chamber and MI proteins. At 24 hours after coculture, dye transfer was evaluated under an inverted microscope. The gap junction blocker heptanol (0.5 mM/ml) was added to culture medium to selectively block the gap junction communication between MSCs before coculture of labeled and unlabeled MSCs.
Cytosolic Ca2+ Measurement
Cardiomyocytes were prepared by primary cell culture from neonatal rat heart as previously described  and cocultured with MSCs that had been pretreated with MI proteins and pressure and then stained with DiI. Cytosolic [Ca2+] response was measured using fluorescence imaging. MSCs and cardiomyocytes were loaded by incubation with 2 μM Fura-2/AM (Sigma-Aldrich) plus 0.1% Pluronic F127 (BASF, Rhineland-Palatinate, Germany, http://www.corporate.basf.com) in normal Tyrode's solution for 20 minutes at room temperature. For fluorescence excitation, we used a polychromatic light source (xenon-lamp-based, Polychrome-IV; TILL-Photonics, Martinsried, Germany, http://www.till-photonics.com), which was coupled into the epi-illumination port of an upright microscope (BX51WI; Olympus, Tokyo, http://www.olympus-global.com) via a quartz light guide and a UV condenser. The imaging of MSC and cardiomyocytes was performed with a ×40 water immersion objective (numerical aperture 0.8; LUMPlanFl; Olympus) and an air-cooled slow-scan charge-coupled device (CCD) camera (SensiCam; PCO, Kelheim, Germany, http://www.pco.de). The monochromator and the CCD camera were controlled by a personal computer and ITC18, running custom-made software programmed with Microsoft Visual C++ (version 6.0) (Microsoft, Redmond, WA, http://www.microsoft.com). A standard two-wavelength protocol was used for fluorescence imaging of cells. Images were taken at 10 Hz, with double wavelength excitation at 340 and 380 nm, and the ratio of fluorescence intensity, F340/F380, obtained from each cell was displayed as an indication of changes in Ca2+ concentration.
In Vivo Differentiation of Human MSCs in Rat Myocardium with AMI
AMI was produced by coronary artery ligation as previously mentioned . Rats were randomly assigned into two groups at 1 week after surgery and received 2 × 106 MSCs or PBS with a 24-gauge needle with a total volume of 100 μl at three different sites (basal anterior, mid-anterior, and basal lateral) in the peri-infarct area. To prevent rejection, rats were treated with busulfan (Jeil-Kirin Pharm., Seoul, Korea, http://www.jeilkirin.co.kr), administered intraperitoneally at a dose of 0.8 mg/kg 1 day before the surgical procedure.
MSCs were labeled with DiI (2.0ug/ml) before cell transplantation. Cell transplantation was performed 1 week after production of MI. Rats underwent echocardiography, performed by one blinded echocardiographer, twice, once before and once 4 weeks after cell transplantation. Echocardiograms were obtained using Acuson XP/10 (Simens, Mountain View, CA, http://www.medical.siemens.com) with an 8-MHz transducer, and the following measurements were obtained: left ventricular end-diastolic diameter (EDD), end-systolic dimension (ESD), and fractional shortening (FS), defined as FS = (EDD − ESD)/EDD. At 8 weeks after cell transplantation, rat hearts were harvested for the immunohistochemistry to confirm the fate of stem cells transplanted.
Immunocytochemistry, Western Blot Analysis, and Enzyme-Linked Immunosorbent Assay
Examinations were done under an inverted microscope (Olympus 1X70) and by confocal microscopy (PE2000; Nikon, Tokyo, http://www.nikon.com). Cells were dispersed using trypsin-EDTA for 15 minutes at 37°C, plated on laminin-coated glass, fixed using 4% formaldehyde with sucrose, and permeated using 0.5% Triton X-100 (Sigma-Aldrich). Cells were blocked with 1% bovine serum albumin and incubated with primary antibodies overnight at 4°C. Primary antibodies used were anti-cTnT (goat polyclonal antibody; Santa Cruz Biotechnology Inc., Santa Cruz, CA, http://www.scbt.com), anti-connexin 43 (rabbit polyclonal antibody; Santa Cruz Biotechnology), anti-α-MHC (mouse monoclonal antibody; Santa Cruz Biotechnology), and α-sarcomeric actin (mouse monoclonal antibody; Sigma-Aldrich). They were detected with fluorescein isothiocyanate (FITC)-conjugated rabbit anti-goat antibody (Zyme [Zymed Lab., South San Francisco, http://www.invitrogen.com]), FITC-conjugated swine anti-rabbit antibody (Dako, Glostrup, Denmark, http://www.dako.com), and FITC-conjugated sheep anti-mouse antibody (Chemicon, Temecula, CA, http://www.chemicon.com) as secondary antibodies.
Expressions of cardiac-specific markers of MSCs injected into the peri-infarct zone of myocardium were confirmed by staining with anti-cTnT and anti-connexin 43 at a dilution of 1:50 and secondary antibody as described above. The presence of MSCs was identified with immunofluorescence of DiI or immunostaining with anti-HLA-ABC (mouse monoclonal antibody; Chemicon) and Cy5-conjugated goat anti-mouse antibody as a secondary antibody. All examinations for immunocytochemistry and immunohistochemistry with immunofluorescence antibody were observed by confocal laser scanning microscopy (LSM 510; Carl Zeiss, Jena, Germany, http://www.zeiss.com).
To evaluate the expression of growth factors in myocardium with AMI, immunohistochemistry and immunoblot analysis were performed. For immunohistochemistry, the myocardium of control and MI rat was totally excised, rinsed, embedded in Tissue-Tek OCT Compound (Sakura Finetek, Torrance, CA, http://www.sakura.com), and stored at −20°C. Then, 8-μm frozen sections were prepared. Tissue was incubated for 1 hour at room temperature with anti-TGF-β1 (rabbit polyclonal antibody; Santa Cruz Biotechnology) and anti-BMP-2 (goat polyclonal antibody; Santa Cruz Biotechnology) at a dilution of 1:50. Rabbit anti-goat biotinylated antibody (1:50) was used as secondary antibody. 3,3-diaminobenzidine was used as a chromagen. Immunoblot assays were performed as described previously . Primary antibodies were rabbit polyclonal TGF-β antibody and goat polyclonal BMP-2 at dilution of 1:200 (Santa Cruz Biotechnology). The secondary antibodies were anti-rabbit-horseradish peroxidase (HRP) (Promega) and anti-goat-HRP (Santa Cruz Biotechnology). The amount of TGF-β1 and BMP-2 in whole serum or myocardium-derived proteins was determined with TGF-β1 and BMP-2 enzyme-linked immunosorbent assay (ELISA) using anti-TGF-β1 and anti-BMP-2 monoclonal antibody as previously described .
Statistical significance was evaluated with an unpaired Student's t test for comparison between two groups. A probability value of <.05 was considered significant. Statistical analysis was performed with SPSS 11.0 for Windows (SPSS, Chicago, http://www.spss.com).
Effects of MI Proteins and Oscillating Pressure on the Differentiation of MSCs
We evaluated the effects of AMI on the differentiation of MSCs into cardiomyocyte lineage by using MI proteins combined with an oscillating pressure chamber to simulate the in vivo beating heart. The expression of cardiomyocyte-specific genes in MSCs was evaluated with RT-PCR and immunocytochemistry. As shown in Figure 2A and 2B, serum alone from rats of either sham operation or myocardial infarction did not induce MSCs to express cardiac markers at all (lanes C and F). MSCs cultured with myocardial proteins from sham-operated rats (lane D) weakly expressed the cardiac transcription factor GATA4 and expressed cTnT and ANP, whereas they did not express MHCs or connexin 43. In contrast, MSCs incubated with MI proteins from rats with AMI (lane G) expressed GATA4, MHC, cTnT, ANP, and connexin 43. In terms of the effect of pressure, MSCs that were incubated in the pressure chamber alone without myocardial proteins or serum (lane B) expressed ANP, connexin 43, and GATA4 but did not express MHC or cTnT. On the contrary, addition of oscillating pressure to myocardial proteins and serum from sham-operated rats augmented expression of all cardiac-specific genes (lane E), which was more prominent in combination with oscillating pressure and MI proteins and serum from rats with AMI (lane H). These differentiation responses were most powerful at a concentration of 25 μg/ml MI protein extracts (Fig. 2C). Western blot analysis of MSCs showed that protein expression of MHC and α-sarcomeric actinin was higher after induction of differentiation with MI proteins and oscillating pressure (Fig. 2D).
Immunocytochemistry confirmed the expression of α-MHC, cTnT, and connexin 43 in MSCs cultured with MI proteins and an oscillating pressure chamber (Fig. 3A–3C), which corroborated the results of RT-PCR. Connexin 43 was expressed preferentially at the interface of MSCs, suggesting the possibility of gap junction formation (Fig. 3D). In addition to α-MHC and cTnT, α-sarcomeric actin was also expressed (Fig. 3D).
To test whether the MSCs formed gap junctions, the unlabeled MSCs, after incubation with control or MI proteins, were cocultured with MSCs labeled with both DiI and calcein, both of which cannot be transferred by simple diffusion. MSCs stained with green fluorescence (calcein acetoxymethyl-ester) without red fluorescence (DiI) represent MSCs that were transferred calcein via functioning gap junctions from the dual-labeled MSCs, since calcein is able to pass only through gap junctions, whereas DiI is not. We observed that some of the initially unlabeled MSCs became green fluorescence-positive, suggesting gap junction formation, which was more prominent in MSCs exposed to both oscillating pressure and MI proteins (Fig. 4A, arrows). Dye transfer was completely blocked when MSCs were pretreated with the gap junction blocker heptanol (Fig. 4B). These results suggested that the MI proteins and oscillating pressure induced the formation of functioning gap junctions between differentiated MSCs. The extent of gap junction formation, measured by the number of MSCs with green fluorescence only, was greater in MSCs treated with MI proteins and oscillating pressure than in any other MSCs (Fig. 4C).
In addition, some MSCs that were cocultured with cardiomyocytes showed synchronous Ca2+ transients (Fig. 4D). Although the magnitude of Ca2+ transients in MSCs (Fig. 4Db, 4Dd) was much smaller than that in spontaneously beating myocytes (Fig. 4Db, 4Dd), this result may suggests the presence of electrical coupling between MSCs with cardiomyocytes. Although MSCs did not acquire the self excitability like cardiomyocytes and showed no spontaneous electrical activity, they showed contraction-like movement that is dependent on cocultured cardiomyocytes (supplemental online files 1, 2 [video clips]).
TGF-β1 and BMP-2 Pathways Are Necessary for the Expression of Cardiac-Specific Markers of MSCs by MI Proteins and Oscillating Pressure
To investigate whether the TGF-β1 and BMP-2 pathways are necessary for the differentiation of MSCs into cells with cardiomyocyte phenotype in the AMI milieu, MSCs were incubated with LAP and noggin, inhibitors of the TGF-β1 and BMP-2 pathways, respectively, in the presence of oscillating pressure and MI proteins. LAP and noggin independently suppressed the expression of α-MHC, β-MHC, cTnT, and connexin 43 (Fig. 5A). Combination of LAP and noggin even abolished GATA4 expression that was induced by MI proteins and oscillating pressure.
To investigate whether the TGF-β1 or BMP-2 pathway is sufficient for the differentiation of MSCs into cells with cardiomyocyte phenotype in the AMI milieu, we tested the effect of TGF-β1 or BMP-2 protein to induce the differentiation of MSCs rather than exposure to MI proteins. When MSCs were treated with 5 ng/ml TGF-β1 or BMP-2, they expressed cTnT and connexin 43, whereas they did not express α- and β-MHC in RT-PCR (Fig. 5B). Addition of oscillating pressure to TGF-β1 or BMP-2 induced expression of β-MHC in MSCs.
In Vivo Pathophysiologic Significance; Expression of TGF-β1 and BMP-2+ in the Myocardium of AMI
The pathophysiologic significance of the in vitro results described above was confirmed using a rat AMI model. To find out whether the myocardium with AMI expresses TGF-β1 and BMP-2, the concentrations of TGF-β1 and BMP-2 in MI proteins were measured by ELISA. The concentrations of TGF-β1 and BMP-2 in the protein extracts from the myocardium were significantly higher in the AMI rats than the sham-operated rats (TGF-β1, 143.6 ± 14.7 ng/mg in AMI rats vs. 48.3 ± 28.2 ng/mg in sham-operated rats [p = .013]; BMP-2, 168.7 ± 10.3 ng/mg in AMI rats vs. 77.5 ± 24.2 ng/mg in sham-operated rats [p = .008]) (Fig. 6A, 6C). The concentration of TGF-β1 and BMP-2 in serum was very low, although there was a statistically significant difference between AMI rats and sham-operated rats (TGF-β1, 0.051 ± 0.002 ng/ml in AMI rats vs. 0.044 ± 0.001 ng/ml in sham-operated rats [p = .03]; BMP-2, 4.82 ± 1.04 ng/ml in AMI rats vs. 3.03 ± 0.21 ng/ml in sham-operated rats [p = .002]) (Fig. 6B). Regarding the distribution of expression in the heart, we found in immunohistochemistry that both TGF-β1 and BMP-2 were strongly expressed in the peri-infarct zone of the myocardium with AMI compared with normal myocardium (Fig. 6D).
In Vivo Differentiation of MSCs Transplanted in Peri-Infarct Area and the Effects on Cardiac Function
To confirm the in vivo differentiation of MSCs in the AMI situation that was suggested by in vitro experiments using MI proteins with oscillating pressure, a simulation of the in vivo AMI milieu, MSCs were labeled with DiI and injected into the peri-infarct area of AMI rats. In myocardium 8 weeks after transplantation, DiI-positive cells were observed to survive and express the gap junction protein connexin 43 between transplanted MSCs themselves, as well as between transplanted MSCs and host cardiomyocytes (Fig. 7A). To confirm the presence of transplanted human MSCs in rat myocardium, myocardial tissues were stained for human-specific antigen using anti-HLA antibody. MSCs with green fluorescence (HLA), which indicated the transplanted human MSCs, expressed red fluorescence of cardiac troponin T (Fig. 7A). Echocardiographic evaluation in rats that underwent cell transplantation showed the reduction of left ventricular EDD (Fig. 7B) and ESD (Fig. 7C), as well as improvement of FS (Fig. 7D), compared with the control group. These results suggest that transplanted human MSCs survived, differentiated, and integrated into host myocardium, leading to improved cardiac function after infarction.
In the present study, we demonstrated that in vitro treatment with myocardial infarct proteins and oscillating pressure simulating in vivo heart of AMI induced the commitment of umbilical cord blood-derived human MSCs to cardiomyocyte-like cells with intermediate phenotype and that this induction was mediated by TGF-β1 and BMP-2 pathways. In vitro cellular electrophysiologic experiments showed that these differentiated MSCs expressing cardiomyocyte-specific markers were able to make electrical coupling with cardiomyocytes but not to do self-beating. Furthermore, the pathophysiologic significance of in vitro results was confirmed using a rat AMI model. We confirmed that TGF-β1 and BMP-2 were significantly activated in the myocardium of AMI and that, when MSCs were transplanted to the peri-infarct zone of heart, they expressed cardiomyocyte-specific markers, leading to improved cardiac function. These in vitro and in vivo results suggest that infarct-related biological and physical factors in AMI induce commitment of MSCs to cardiomyocyte-like cells through TGF-β/BMP-2 pathways.
Biological and Physical Factors of Microenvironment in the Infarcted Heart Are Important for Commitment of MSCs to Cardiomyocyte Lineage
The effect of transplanted MSCs on infarcted myocardium has been vigorously investigated for the past decade [17, –19]. Conversely, however, the effect of microenvironments in host heart on the transplanted MSCs themselves has rarely been studied. Behfar et al. reported that the cytokines from myocardium affecting transplanted stem cells in a paracrine pathway are required for stem cell differentiation in chronic myocardial infarction model . Different conditions of host factors, including phase of myocardial infarction and vascularity of myocardium, were also reported to influence the fate of transplanted stem cells [21, –23].
In our study, we used an in vitro culture system including proteins extracted from myocardium with AMI and an oscillating pressure chamber. This system biologically simulates the microenvironment of AMI, where various cytokines or growth factors are activated [24, 25], and it physically simulates the changing pressure of the beating heart. We found that it was possible to induce MSCs to be cardiomyocyte-like cells through biological and physical stimulation in vitro.
Myocardial proteins from normal heart induced expression of certain cardiac-specific genes, such as GATA 4, ANP, and cTnT, but not connexin 43 or MHC. In contrast, MI proteins could induce the expression of connexin 43 and MHC in addition to GATA4, ANP, and cTnT. We also showed that these MSCs committed to cardiomyocyte lineage expressed gap junctions among themselves, as well as between themselves and cardiomyocytes in vivo. It means that myocardial proteins from normal heart have only partial effects on the differentiation of stem cells. However, myocardial infarction allows the release of more cytokines from myocardial proteins, which are favorable for the cardiac differentiation of stem cells.
Oscillating pressure alone did not induce the differentiation, but it significantly augmented the expression of cardiac-specific genes in committed MSCs. This suggested that cytokines or paracrine networks in the infarcted myocardium are involved in differentiation induction in the conditions of AMI, which can be augmented by oscillating pressure as in the beating heart.
Role of TGF-β1 and BMP-2 Pathways for MSC Differentiation in the Acutely Infarcted Myocardium
We demonstrated that the TGF-β1 or BMP-2 pathway is necessary in MSC differentiation in the AMI setting in our study. TGF-β1 and BMP-2 have been reported as important cytokines for differentiation of both embryonic and adult stem cells [20, 26, 27]. Furthermore, exogenous TGF-β augmented the engraftment and differentiation of stem cells injected to the myocardium after ischemic injury [28, 29]. Despite this indirect evidence of its involvement in the differentiation of stem cells, direct evidence of the influence of microenvironment in AMI on stem cell differentiation and its underlying mechanism has not yet been reported. In the present study, we confirmed that TGF-β1 and BMP-2 are significantly elevated in AMI. We also showed that the protein extracts from myocardium with AMI, which we applied to simulate the AMI milieu in vitro, contained significantly higher levels of TGF-β1 and BMP-2 compared with those from normal myocardium. In addition, pretreatment with a specific blocker for these molecules inhibited the differentiation of MSCs into cardiomyocyte lineage by MI proteins and oscillating pressure, whereas TGF-β1/BMP-2 alone did not completely reproduce the results from the addition of MI proteins. These findings suggest that TGF-β1/BMP-2 pathways are necessary but not sufficient to induce the commitment of MSCs to cardiomyocyte lineage.
Mechanism of Cyclic Pressure Changes to Induce Cell Differentiation
We also demonstrated the role of cyclic pressure change to induce the expression of connexin 43, GATA-4, and ANP and augmentation of cardiac marker expression in MSCs. Although the mechanism of differentiation by cyclic pressure is unclear, cyclic stretching was reported to increase expression of connexin in neonatal cardiomyocytes [30, 31] and vascular smooth muscle cells  via c-fos expression. A recent study reported that expression of TGF-β receptors and production of their corresponding ligands are increased in bone marrow MSCs by cyclic pressure . Considering our results that oscillating pressure-induced connexin 43 and GATA4 expression was suppressed by inhibitors of the TGF-β1 and BMP-2 pathways, there is a possibility that cyclic pressure also involved TGF-β/BMP-2 pathways.
Umbilical Cord Blood-Derived MSC as a Source of MSCs
In this study, the source of MSCs was human umbilical cord blood. At present, the bone marrow is the most common source of MSCs, which have been widely investigated. However, bone marrow-derived MSCs have several limitations. Aspiration of bone marrow needs an invasive procedure, and because of the limited number of MSCs in aspiration products, ex vivo culture for expansion is required. Moreover, MSCs obtained from aged subjects could be dysfunctional and limited in number compared with MSCs obtained from younger subjects . With these points considered, although allogenic source is a limitation for cord blood MSCs, they may be a promising alternative to the MSCs from bone marrow. They have same immunophenotype and multilineage differentiation characteristics  as of MSCs from the bone marrow and, furthermore, have advantages because they are derived from young and healthy stem cells and can escape the possibility of senescence  and are readily available for cell therapy in the clinic.
There are some limitations in the present study. First, we could not observe full and terminal differentiation of MSCs to authentic beating cardiomyocytes in in vitro culture, although we confirmed that MSCs injected into myocardium survived, expressed cardiac markers, and improved cardiac function after 8 weeks of transplantation. Thus, we guess that biological and physical factors of infarcted heart may help MSCs commit to cardiomyocyte lineage but not terminally differentiate to authentic cardiomyocytes. Second, although transplantation of MSCs enhanced the ventricular function after AMI, we cannot entirely prove the mechanism of this phenomenon. Myocardial regeneration by stem cell differentiation could be a plausible mechanism [35, –37]. However, the findings from a previous study and our study, that MSCs improved myocardial function in the absence of terminal differentiation to cardiomyocytes, suggest the possibility of a transient paracrine effect of transplanted MSCs . This is an important issue that needs further investigation. Third, differentiation induction of human MSCs with proteins from rat myocardium is a limitation that comes from technical difficulty in obtaining a large amount of human myocardium.
Treatment of proteins extracted from myocardium with AMI and application of oscillating pressure induce the commitment of human cord blood MSCs into the cardiomyocyte-like phenotype via TGF-β1- and BMP-2-dependent pathways. Our results suggest that infarct-related biological and physical factors in the microenvironment of AMI help MSCs commit to the intermediate phenotype from MSCs to cardiomyocyte and make electrical coupling with cardiomyocytes, which may contribute to the improvement of myocardial function after infarction.
Disclosure of Potential Conflicts of Interest
The authors indicate no potential conflicts of interest.
This study was supported by a grant for Stem Cell Research Center (SC3150), the Korean Society of Circulation, and the Innovative Research Institute for Cell Therapy. S.-A.C. and E.J.L. contributed equally to this work.