Dedifferentiated fat cells convert to cardiomyocyte phenotype and repair infarctedcardiac tissue in rats
Medet Jumabay a, Taro Matsumoto b,⁎, Shin-ichiro Yokoyama a, Koichiro Kano c, Yoshiaki Kusumi d,Takayuki Masuko b, Masako Mitsumata d, Satoshi Saito a, Atsushi Hirayama a,Hideo Mugishima e, Noboru Fukuda b,f,⁎a Division of Cardiology, Department of Medicine, Nihon University School of Medicine, Tokyo 173-8610, Japanb Division of Cell Regeneration and Transplantation, Advanced Medical Research Center, Nihon University School of Medicine, Tokyo 173-8610, Japanc Department of Animal Science, College of Bioresource Science, Nihon University, Fujisawa 252-8510, Japand Department of Pathology, Nihon University School of Medicine, Tokyo 173-8610, Japane Department of Pediatrics, Nihon University School of Medicine, Tokyo, 173-8610, Japanf Advanced Research Institute of the Science and Humanities, Nihon University Graduate School, Tokyo 102-0073, Japan
a b s t r a c ta r t i c l e i n f o
Article history:Received 25 May 2008Received in revised form 31 July 2009Accepted 3 August 2009Available online 15 August 2009
Adipose tissue-derived stem cells have been demonstrated to differentiate into cardiomyocytes and vascularendothelial cells. Here we investigate whether mature adipocyte-derived dedifferentiated fat (DFAT) cellscan differentiate to cardiomyocytes in vitro and in vivo by establishing DFAT cell lines via ceiling culture ofmature adipocytes. DFAT cells were obtained by dedifferentiation of mature adipocytes from GFP-transgenicrats. We evaluated the differentiating ability of DFAT cells into cardiomyocytes by detection of the cardiacphenotype markers in immunocytochemical and RT-PCR analyses in vitro. We also examined effects of thetransplantation of DFAT cells into the infarcted heart of rats on cardiomyocytes regeneration andangiogenesis. DFAT cells expressed cardiac phenotype markers when cocultured with cardiomyocytes andalso when grown in MethoCult medium in the absence of cardiomyocytes, indicating that DFAT cells havethe potential to differentiate to cardiomyocyte lineage. In a rat acute myocardial infarction model,transplanted DFAT cells were efficiently accumulated in infarcted myocardium and expressed cardiacsarcomeric actin at 8 weeks after the cell transplantation. The transplantation of DFAT cells significantly(pb0.05) increased capillary density in the infarcted area when compared with hearts from saline-injectedcontrol rats. We demonstrated that DFAT cells have the ability to differentiate to cardiomyocyte-like cells invitro and in vivo. In addition, transplantation of DFAT cells led to neovascuralization in rats with myocardialinfarction. We propose that DFAT cells represent a promising candidate cell source for cardiomyocyteregeneration in severe ischemic heart disease.
Research in the field of stem cell biology presents new opportu-nities for the use of cell-based therapies in disease areas with criticaland unmet medical needs. It has been shown that adult, umbilicalcord, embryonic, and fetal stem cells hold great potential as newtherapeutic strategies for the regeneration and repair of damaged
tissues or organs and may prove to be valuable in the treatment ofsevere cardiovascular diseases. In parallel, autologous cell therapy forthe induction of angiogenesis is thought to be a more progressiveregenerative treatment for vascular diseases [1,2]. The bone marrow(BM) possesses an abundance of stem cells [1] and endothelialprogenitor cells [2]. Results stemming from the Therapeutic Angio-genesis Using Cell Transplantation study demonstrated that BM cellimplantation (BMI) improves ischemic ulcers by restoring blood flowin cases of limb ischemia [3]. In addition, BMI has been considered as atreatment for ischemic heart diseases; preliminary clinical studiessuggest potential clinical benefits of cellular transplantation therapyin patients with acute myocardial infarction and chronic myocardialischemia [4–7]. Nevertheless, the development of this type oftherapeutic approach for the treatment of ischemic heart diseasesthat are refractory to conventional therapy presents with severalinherent shortcomings, such as long-term safety, optimal timing, and
Journal of Molecular and Cellular Cardiology 47 (2009) 565–575
⁎ Corresponding authors. N. Fukuda is to be contacted at Advanced Research Instituteof Science and Humanities, Nihon University, Ooyaguchi, kami-cho 30-1, Itabashi-ku,Tokyo 173-8610, Japan. Tel.: +81 3 3972 8111; fax: +81 3 3972 8666. T. Matsumoto,Division of Cell Regeneration and Transplantation, Advanced Medical Research Center,Nihon University School of Medicine, Tokyo 173-8610, Japan, Ohyaguchi, kami-cho 30-1, Itabashi-ku, Tokyo 173-8610, Japan. Tel.: +81 3 3972 8111; fax: +81 3 3972 8666.
a lack of adequate treatment strategies, especially those pertaining tothe risk of harvesting BMduring acute phases of cardiac disease and themethod of delivery of BM cells to the ischemic site in the heart. Finally,BMI requires large amounts of bone marrow, but marrow extractionunder general anesthesia is considered to be a high-risk procedurewhen performed on patients with severe heart disease.
Researchers in this field have recently shifted their focus towardcertain characteristics of stem cells derived from adipose tissue. Non-adipocyte cell fraction in adipose tissue known as stromal vascularfraction (SVF) contains approximately 3% of adherent colony-formingunit cell [8]. This type of cells, named adipose-derived stromal/stemcells (ASCs), can be enriched and expanded by culturing on plasticsurface. It has been shown that ASCs can differentiate along multiplelineages, including adipocytes, osteoblasts, chondrocytes, myocytes,neuronal cells, endothelial cells, and cardiomyocytes [9,10]. Sinceadipose tissue is often available in abundant and expendablequantities, ASCs exhibit potential advantages in cell therapy andtissue engineering applications in patients with cardiovasculardiseases. However, the efficiency of ASC isolation needs to beimproved because the procedures established so far are from a largevolume of adipose tissue obtained by liposuction. Moreover, thestromal cell fraction extracted from adipose tissue is highly complex,as it includes fibroblasts, preadipocytes, endothelial cells, and othertypes of cells [11]. Therefore, other adult stem cell sources that can beeasily isolated and expanded with a high purity are still needed.
In contrast to ASCs, mature adipocytes are the most abundant celltype in adipose tissue. After collagenase digestion and filtration ofadipose tissue, mature adipocyte fraction can be separated from SVFas a white layer floating in the tube by centrifugation. We have shownthat mature adipocytes can revert to a more primitive phenotype andgain cell-proliferative ability when subjected to an in vitro dedifferen-tiation strategy, such as the ceiling culture method, and refer to thesecells as dedifferentiated fat (DFAT) cells [12]. DFAT cells lost theexpression of genetic markers specific to mature adipocytes butretained or gained expression of mesenchymal lineage-committedmarker genes, such as the peroxisome proliferator-activated receptor γ(PPAR-γ), RUNX2, and SOX9 [13]. The surface immunophenotype ofDFAT cells closely resembles that of BM-MSCs and ASCs. In vitrodifferentiation analysis revealed that DFAT cells had the capacity todifferentiate into adipocytes, chondrocytes, and osteoblasts underappropriate cell culture conditions even though the cells were clonallyexpanded [13]. These results indicate that DFAT cells represent a typeof multipotent progenitor cell. Because DFAT cells were obtained frompure mature adipocytes, they represent a more homogeneouspopulation than those obtained from ASCs. In addition, we haveshown that DFAT cells can be obtained and expanded from smallamounts (approximately 1 g) of subcutaneous adipose tissue indonors regardless of their age [13]. The accessibility and ease ofculture of DFAT cells support their potential application in cell-basedtherapies for patients with cardiovascular diseases.
In the present study, we investigated the capacity of DFAT cells todifferentiate into cardiomyocytes, in vitro and in vivo. We alsoexaminedwhetherDFAT cell transplantation contributes tomyocardialregeneration and neovascularization in a rat acute myocardialinfarction model.
2. Materials and methods
2.1. Animals
Sprague–Dawley (SD) rats and green fluorescent protein (GFP)-transgenic rats (SD TgN [act-EGFP] OsbCZ-004) were purchased fromJapan SLC Inc. (Hamamatsu, Japan). All rats were kept in microisolatorcagesona12:12-hday/night cyclewith freeaccess towaterandstandardchow diet. All animal experiments were performed in our laboratoryaccording to the Guide for the Care and Use of Laboratory Animals,
publishedby theUSNational Institutes ofHealth (NIHPublicationNo. 85-23, revised in 1996) andwere approvedby theAnimal Research andCareCommittee at the Nihon University School of Medicine.
2.2. Isolation of the DFAT cells from adipose tissue
Isolation of mature adipocytes from fat tissue was performed witha modification of the method described previously by Sugihara et al.[14]. Briefly, an isolation of approximately 2 g of fat tissue was mincedand digested using 0.2% (w/v) collagenase solution (C6885; Sigma-Aldrich, St. Louis, MO) at 37 °C for 45 min with gentle agitation. Afterfiltration through a 150-μm mesh and centrifugation at 135 g for3 min, the floating top layer containing unilocular adipocytes wascollected. To avoid contamination with stromal cells, the dissociatedfat cells were repeatedly pipetted, washed with phosphate-bufferedsaline (PBS), and centrifuged at least three times. The cells (5×104)were then placed in 25-cm2 culture flasks (NUNC, Roskilde, Denmark)filled completely with Dulbecco's modified Eagle's medium (DMEM;Invitrogen, Carlsbad, CA) supplemented with 20% fetal bovine serum(FBS; JRH Bioscience, Lenexa, KS, Lot 6G2146) and were incubated at37 °C in 5% CO2. Cells floated up and adhered to the top inner ceilingsurface of the flask. After 7 days, the medium was removed, and theflasks were inverted. The medium was changed every 3 days until thecells reached confluency. After splitting, the cells were used forexperiments before they reached passage 5. For microscopy, cellswere fixed in 4% paraformaldehyde (Sigma-Aldrich) for 15 min,washed once with PBS, and incubated for 10 min with AdipoRed(Cambrex, Walkersville, MD) to visualize the lipid droplet. Cells werethen washed in PBS and subjected to DNA staining with 5 μg/mlDAPI (Pierce, Rockford, IL). Staining was visualized and photo-graphed with a fluorescent microscope (Nikon Eclipse TE 2000-U;Nikon, Tokyo, Japan).
2.3. Differentiation of DFAT cells
Neonatal rat cardiomyocytes were isolated from SD rats aspreviously described [15]. Briefly, hearts from1- to 3-day-old neonatalrats were aseptically harvested. The atria and the aortawere discardedand the ventricles were minced and trypsinized at 37 °C with gentlestirring in G-solution (137 mM NaCl, 5.4 mM KCl, 1.47 mM KH2PO4,1.08 mM Na2HPO4, 6.1 mM glucose; pH 7.4) containing 0.1% trypsin(Invitrogen) for 10 min. The supernatant was collected and solutionwas added to the tissue, which was digested for an additional 10 min.This step was repeated until the tissue was completely digested. Thefirst two supernatant fractions were discarded, and all subsequentdigestion supernatants were rendered inactive by the addition ofneonatal calf serum (Invitrogen) to the cell–enzymemixture, followedby centrifugation and resuspension in DMEM containing 10% FBS,100 U/ml penicillin, and 100 μg/ml streptomycin. After a pre-incubation period at 37 °C for 2 h, the unattached cardiomyocyteswere seeded at a density of 1×105 cells/cm2 in 35 mm dishes.
We performed four different in vitro experiments to evaluate theability of DFAT cells to differentiate into cardiomyocytes: (1)coculture method—1×104 DFAT cells were directly cocultured with2×104 neonatal cardiomyocytes in 12-well dishes with DMEM/M199(4:1) medium in the presence of 10% FBS and 5% horse serum (JRHBiosciences); (2) conditioned medium method—1×105 DFAT cellswere cultured in the cardiomyocyte-conditioned medium. To preparethe medium, rat neonatal cardiomyocytes were cultured in T-25 cm2
flask with 5 ml of DMEM/M199 (4:1) medium containing 10% FBS and5% horse serum for 48 h. The resulting conditioned media werereplaced with equal volume of fresh media (DMEM/10% FBS).Subsequently, the conditioned media were filtered with a 0.22-μmfilter and used to the differentiation culture of DFAT cells. The mediumwas changed every 3 days during the differentiation culture; (3)transwell culture method—the DFAT cells were cultured on a transwell
566 M. Jumabay et al. / Journal of Molecular and Cellular Cardiology 47 (2009) 565–575
insert (Becton Dickinson Labware, Franklin Lakes, NJ) in DMEMsupplemented with 10% FBS using cardiomyocytes as feeder cells;and (4) stem cell mediummethod—1×105 DFAT cells were cultured instem cell methylcellulose medium (MethoCult GF M3534; StemCellTechnologies Inc., Vancouver, Canada) containing basic 1% methyl-cellulose in Iscove's modified Dulbecco's medium (Invitrogen) supple-mentedwith 1% bovine serum albumin (BSA; Sigma-Aldrich), 15% FBS,2-mercaptoethanol (0.1 mM), L-glutamine (2 mM), recombinanthuman insulin (10 μg/ml), human transferrin (200 μg/ml), recombi-nant murine interleukin 3 (IL-3; 10 ng/ml), recombinant human IL-6(10 ng/ml), and recombinant mouse stem cell factor (50 ng/ml).
2.4. Ischemic heart model
Male SD rats weighing 170–210 g were used in this study. Amyocardial infarction rat model was produced by ligation of the leftcoronary artery. Briefly, rats were anesthetized with sodium pento-barbital (30 mg/kg) and ventilated with a volume-regulatedrespirator. Left thoracectomy was performed through the fourthintercostal space, and the pericardium was opened to expose theheart. The left coronary artery was ligated between the pulmonaryartery conus and the left atrium (2–3 mm from its origin) using a 6-0Prolene suture. Three hours after ligation, 1×106 DFAT cells in 20 μlsaline (DFAT group, n=15) or 20 μl saline alone (Control group,n=25) were administered by intramuscular injection in five differentsites of the ischemic area. Five rats in each group were sacrificed at3 days after transplantation, and heart tissue was examinedimmunohistochemically to evaluate surviving transplanted cells.The other surviving rats were sacrificed at 8 weeks after transplan-tation, and heat tissue was examined immunohistochemically toevaluate surviving transplanted cells and neovascularization.
Total RNA was isolated from DFAT cells using the RNeasy Mini kit(Qiagen, Hilden, Germany). Aliquots of total RNA were reversetranscribed into single-stranded cDNA by incubation with avianmyeloblastoma virus reverse transcriptase (Clontech, Palo Alto, CA).The cDNA was PCR amplified for 30 cycles using the AdvantagePolymerase Mix (Clontech) in a GeneAmp PCR system, model 9700(PerkinElmer Life and Analytical Sciences). Copurified DNA wasamplified using the following primers: Nkx2.5 forward, 5′-ACCG-CCCCTACATTTTATCC-3′; reverse, 5′-GACAGGTACCGCTGTTGCTT-3′;GATA4 forward, 5′-CATGCTTGCAGTTGTGCTAG-3′; reverse, 5′-ATTCTCTGCTACGGCCAGTA-3′; Troponin-T forward, 5′-CAAGGAACA-GAGCTTTGTCGAA-3′; reverse, 5′-CACAACCTAGAGGCCGAGAAGT-3′;Cardiac α-actin forward, 5′-CTGAGGCGGCTACCTTACAC-3′; reverse,5′-AAGCCTGGCCTGGTTTATTT-3′. Each cycle consisted of denaturationat 94 °C for 30 s and annealing/extension at 70 °C for 1 min. 18Sribosomal RNA was used as an internal control.
2.6. Immunocytochemistry and immunohistochemistry
Cells grown in chamber slides were fixed in 4% paraformaldehyde/PBS, permeabilized with 0.1% Triton X-100, blocked with 1% normalgoat serum and 1% BSA in Tris-buffered saline (TBS), and incubatedovernight at 4 °C with mouse monoclonal anti-sarcomeric actinantibody (1:500; Sigma-Aldrich), rabbit anti-GATA4 antibody (1:200;Santa Cruz Biotechnology, Santa Cruz, CA), mouse monoclonal anti-Nkx2.5 antibody (1:200; R&D Systems, Minneapolis, MN), mousemonoclonal anti-troponin-T antibody (1:200; Thermo Fischer Scien-tific Anatomical Pathology, Fremont, CA), or mouse monoclonal anti-connexin 43 antibody (1:200; Sigma-Aldrich). The following day cellswere washed three times with PBS then incubated with theappropriate secondary antibodies (Alexa 594 goat anti-mouse IgG,or Alexa 594 goat anti-rabbit IgG antibodies, 1:500; Invitrogen) for 1 h
at room temperature. After washing three times with PBS, nuclei werecounterstained with 5 μg/ml DAPI. The samples were air dried,mounted in Fluoromount-G (Southern Biotechnology, Birmingham,AL), and photographed using a digital camera under a fluorescentmicroscope (Nikon Eclipse TE 2000-U).
For immunohistochemistry, rat hearts were fixed in 10% parafor-maldehyde/PBS for 8 h and subsequently processed through a gradedethanol series for dehydration, placed in xylene, then paraffinembedded. Paraffin-embedded tissues were sectioned at a thicknessof 5 μmandmounted on positively charged Labcraft glass slides. Slide-mounted tissue sections were deparaffinized by performing two 5-min xylene rinses followed by two 5-min 100% ethanol rinses. Antigenrecovery was carried out by microwave boiling the sections for 5 minin citrate buffer (0.01 M), followed by a cooling down period in citratesolution for 20 min before rinsing with PBS. Slides were incubatedovernight at 4 °C in a humidified slide chamberwith either rabbit anti-GFP antibody (1:200; Medical and Biological Laboratories, Nagoya,Japan) or mouse monoclonal anti-sarcomeric actin antibody (1:200),followed by incubation with Alexa 594 anti-mouse IgG and Alexa 488anti-rabbit IgG (1:500; Invitrogen) secondary antibodies. Stainingwas visualized and photographed with an immunofluorescencemicroscope (Nikon Eclipse TE 2000-U) or a multiphoton laserscanning microscope Fluoview FV1000 (Olympus, Tokyo, Japan). Incertain experiments, biotinylated anti-rabbit Ig antibody (1:200dilution, Vectastain ABC Elite Kit; Vector Laboratories, Burlingame,CA) was used for the secondary antibody. A positive reaction wasdetected by the ABC method and visualized by diaminobenzidine(DAB) reaction. Sections were counterstained with hematoxylin. As acontrol, for each antibody one slide was incubated with all reagentswithout primary antibody.
To evaluate capillary density in the ischemic myocardium, heartsamples were incubated overnight at 4 °C with biotinylated Griffoniasimplicifolia isolectin B4 (1:25 dilution; Vector Laboratories) andmouse monoclonal anti-α smooth muscle actin antibody (ASMA,1:200 dilution; DakoCytomation, Glostrup, Denmark). Samples werethen washed with PBS and incubated for 1 h at room temperaturewith FITC-conjugated streptavidin (1:100 dilution; BD Biosciences, SanJose, CA) andAlexa 594 goat anti-mouse IgG antibody (1:500) followedby nuclear staining with DAPI. Staining was visualized and photo-graphed with an immunofluorescence microscope (Nihon Eclipse TE2000-U). Vessels were counted and results were presented as thetotal number of isolectin B4- and ASMA-positive vessels normalizedto area.
2.7. Statistical analyses
Results were expressed as mean±SEM. One-way ANOVA wasused to compare each parameter. Post hoc Bonferroni t testcomparisons were subsequently performed to identify which differ-ences accounted for the significant overall two-way ANOVA. Mann–Whitney U-test was used for intergroup comparisons in evaluation ofvascular density.
3. Results
3.1. Differentiation of DFAT cells to cardiomyocytes
Morphological changes of adipocytes in the ceiling culture areshown in Fig. 1. When lipid-filled mature adipocyte fraction (Fig. 1A)isolated from adipose tissue are applied into culture flasks filledcompletely with media, the mature adipocytes attached to uppersurface of flasks and extend cytoplasm by day 3 of culture (Fig. 1B).Approximately 50% of the adherent cells enter cell cycle and divideasymmetrically (Supplementary video), generating fibroblast-likeDFAT cells containing small lipid droplets by days 3–5 (Figs. 1C andD). These findings suggest that the DFAT cells were directly produced
567M. Jumabay et al. / Journal of Molecular and Cellular Cardiology 47 (2009) 565–575
from adipocytes but not from cells attached to adipocytes. DFAT cellsthen divide symmetrically and form colonies by day 7 (Fig. 1E). Whenthe culture medium is changed and the flasks were inverted on day 7,cells enter a proliferative log phase, lost the lipid droplets completely,and exhibit spindle-shaped morphology (Fig. 1F). DFAT cells isolatedfrom subcutaneous adipose tissue of GFP-transgenic rat were used forcardiomyocyte differentiation experiments. These cells also exhibitedhomogenous spindle-shaped morphology (Fig. 1G), and approxi-mately 100% cells had GFP fluorescence (Fig. 1H).
To induce cardiogenic differentiation in DFAT cells, DFAT cellsderived from GFP-transgenic rats were directly cocultured withneonatal SD rat cardiomyocytes. From the third day in culture, therat cardiomyocytes formed myotubes and showed net-like structures(Fig. 2A). In this period, GFP-positive DFAT cells were efficientlyincorporated into cardiomyocyte myotubes (Figs. 2A and B). Immuno-cytochemical analysis revealed that cardiomyocytes expressed con-nexin 43, a major gap junction protein, in the myotubes and DFAT cellaggregates (Fig. 2C).
Fig. 1. Morphological changes of adipocytes during the ceiling culture method. (A) Morphology of isolated unilocular adipocytes from the floating cell layer after centrifugation ofcollagenase-digested adipose tissue. Cells were stained for neutral lipid by AdipoRed (Adipo) and for nuclei by DAPI. Merged view of bright-field (BF) and fluorescence images areshown. (B–F) Morphology of adipocytes on day 2 (B), day 3 (C and D), day 7 (E), and day 14 (F) of culture. Note that adherent adipocytes with a single nucleus (B) divideasymmetrically and generated fibroblast-like DFAT cells (C and D, arrows). The cells exhibited colony formation on day 7 (E) followed by morphologically homogenous spindle-shaped morphology on day 14 (F). (G and H) Morphology of DFAT cells derived from a GFP-transgenic rat. A phase-contrast image (G) and a fluorescent image (H) are shown. Scalebars represent 100 μm.
568 M. Jumabay et al. / Journal of Molecular and Cellular Cardiology 47 (2009) 565–575
To confirm the expression of cardiac-specific proteins in DFATcells, we performed immunostainings against cardiac nuclear proteinsGATA4 and Nkx2.5 and against cardiac cytoplasmic proteins such ascardiac sarcomeric actin and troponin-T. Seven days after the directcoculture, the GFP-positive DFAT cells were shown to express GATA4and Nkx2.5 (Figs. 2D and E, arrowheads) as well as cardiac sarcomericactin and troponin-T (Figs. 2F and G, arrowheads). Immunoreactivityof these marker proteins was not detected in DFAT cells before
coculture with cardiomyocytes. Quantitative analysis revealed that6.3±2.6%, 3.1±1.2%, and 5.6±0.8% (mean±SEM) of GFP-positivecells expressed GATA4, Nkx2.5, and troponin-T, respectively (Fig. 2H).On the other hand, differentiated GFP-positive cells did not exhibitimmunoreactivity forMyo D, a skeletalmuscle-specific transcriptionalfactor (data not shown). These results indicate that DFAT cellsdifferentiate into cardiomyocyte lineage when cocultured with nativecardiomyocytes.
Fig. 2. Morphology and cardiac markers expression of DFAT cells directly cocultured with neonatal rat cardiomyocytes. (A, B) Merged view of bright-field (BF) and fluorescenceimages (GFP) at day 3 of the coculture. Note that cardiomyocytes form net-like cellular structure that consists of myotubes (A). GFP-positive DFAT cells are efficiently incorporatedinto themyotubes (A and B). (C) Immunostaining of cells with connexin 43 (CX43) at day 3 of the coculture. Nuclei were stainedwith DAPI. Note that GFP-positive DFAT cells expressconnexin 43 and connect with neighborhood cardiomyocytes. Arrowheads and arrows indicate connexin 43-positive and -negative DFAT cells, respectively. (D–G) Expression ofcardiac phenotypic markers in DFAT cells directly cocultured 7 days with neonatal cardiomyocytes. Immunostainings of cells with anti-GATA4 (D), anti-Nkx2.5 (E), anti-sarcomericactin (Sr) (F), and anti-troponin-T (G). Arrowheads show GFP-positive DFAT cells expressing cardiac phenotypic markers. Arrows show DFAT cells do not express these markers.Scale bars represent 50 μm. (H) The percentage of each cardiac maker-positive cells in DFAT cells was quantified. Values are mean±SEM of triplicated dishes.
569M. Jumabay et al. / Journal of Molecular and Cellular Cardiology 47 (2009) 565–575
3.2. Expression of cardiac phenotype markers in DFAT cells
We next evaluated the expression of cardiac-specific genes in DFATcells after 1–3 weeks of DFAT cell culture using either the conditionedmedium from rat neonatal cardiomyocytes, indirect coculture with thecardiomyocytes by the transwell culture system, or methylcellulosesemisolidmediumMethoCult GFM3534. DFATcells displayed increasesin expression of Nkx2.5, troponin-T, GATA4, and cardiac actin over timein each culture condition (Fig. 3A). Quantitative analysis revealed thatlevels of these makers expression were significantly increased in thecells 3 week after the differentiation culture compared with that in thecells before induction (Fig. 3B). Interestingly, expression level of Nkx2.5was significantly (pb0.01) larger in DFAT cells cultured in the Metho-Cult medium compared to those in the cardiomyocyte-conditionedmedium or in the transwell coculture system.
In the culture of the MethoCult medium, we observed that DFATcells proliferated up to day 3, elongated at day 7, and aligned and
branched tomake clusters at day 15 (Figs. 4A–C). Immunocytochemicalanalysis revealed that DFAT cells cultured in the MethoCult mediumexpressed Nkx 2.5, troponin-T, and cardiac sarcomeric actin on day 15of the culture (Figs. 4D–F). These marker proteins were not detectedwhen the specific antibodies were replaced to Ig isotype control(data not shown). The percentage of troponin-T-positive cells inDFAT cells cultured in MethoCult medium was 13.6±0.8% (mean±SEM), which is significantly (pb0.05) larger than those in the directcoculture system (Fig. 4G). These results suggest that DFAT cellsconvert into cardiomyocytes phenotype without cardiomyocyte-coculture. However, we did not observe visible contraction activityin these cells (data not shown).
3.3. Differentiation of transplanted DFAT cells into myocardium
We next examined whether DFAT cell transplantation contributemyocardial regeneration in a rat acute myocardial infarction model.
Fig. 3. DFAT cells expressed cardiac-specific genes in vitro. DFAT cells were cultured in neonatal cardiomyocyte-conditioned medium (Cond-medium), in the transwell coculturesystem with cardiomyocytes (Transwell), or in metylcellulose semisolid mediumMethoCult GF M3534 (MethoCult) for 3 weeks. Total RNA was extracted at indicated time periods,and RT-PCR analysis was performed. (A) Expression of cardiac-specific genes evaluated by RT-PCR analysis. Note that Nkx2.5, troponin-T, GATA4, and cardiac actin were expressed inDFAT cells cultured in three different induction medium. Rat neonatal cardiomyocytes were used as a positive control of these genes. (B) The ratio of gene expression to 18Sribosomal RNA in DFAT cells before and 3 weeks after the three different culture was quantified by NIH image software. Data are expressed as mean±SEM of three independentsamples. ⁎pb0.01 from condition-medium or transwell coculture.
570 M. Jumabay et al. / Journal of Molecular and Cellular Cardiology 47 (2009) 565–575
Three hours after left coronary artery ligation, DFAT cells (1×106) orsaline (as control) were injected intramuscularly at five different sitesof ischemic area. Survival and differentiation of transplanted DFATcells into myocardium was assessed based on expression of GFP andcardiac sarcomeric actin. At 3 days after DFAT cell transplantation,GFP-positive cells were found clustered at the site of injectionwithoutany specific orientation (Fig. 5A). On the other hand, immuno-reactivity for GFP was not detected in myocardium of saline-injectedcontrols (data not shown). At 8 weeks after the DFAT cell trans-plantation, the GFP-positive cells were still alive and efficientlyaccumulated in infractedmyocardium (Fig. 5B, arrowheads). The GFP-positive cells were also observed in normal myocardium but lessamount (Fig. 5B, arrows). A large number of GFP-positive cellsexpressed cardiac sarcomeric actin. This finding was frequently seenin scared area (Figs. 5C and D). In scared area, cardiac sarcomericactin-positive DFAT cells often became fused and formed myotube-like structure (Fig. 5E). Occasionally, cardiac sarcomeric actin-positiveDFAT cells were observed in normal cardiomyotubes (Fig. 5F).Interestingly, GFP-positive cells differentiating the other cell lineages
such as adipocytes or bone cells were not observed in all animals weexamined. These results suggest that transplanted DFAT cells survivedin the ischemic area by at least 8 weeks after transplantation and mayhave contributed to the repairing of the damaged heart tissue toconvert into cardiomyocyte phenotype.
We next examined whether DFAT cell transplantation promoteangiogenesis in the infarcted heart tissue. Immunohistochemicalstaining for isolectin B4 and ASMA, markers for vascular endothelialcells and smoothmuscle cells, respectively, was performed. The resultsshowed that transplantation of DFAT cells significantly (pb0.05)increased the capillary density in the infarcted area when comparedwith hearts from saline-injected control animals (Figs. 6A–C). Thenumber of ASMA-positive capillary vessels was increased predomi-nantly in DFAT transplanted hearts (Fig. 6C), suggesting the DFAT celltransplantation promote maturation of capillary vessels as well asneovascularization.
Fig. 4.Morphological changes and cardiacmarker proteins expression of DFAT cells cultured inmethylcellulose medium. GFP-labeled DFAT cells were cultured in themethylcellulosemediumMethoCult GFM3534 for 15 days. Morphology of DFAT cells was assessed by a fluorescent microscope at day 3 (A), day 7 (B), and day 15 (C) of culture. On day 15 of culture,cells were immunostained for Nkx2.5 (D), troponin-T (E), and sarcomeric actin (F). Arrowheads showDFAT cells expressing cardiac phenotype marker proteins. Scale bars represent50 μm. (G) Percentage of troponin-T-positive cells was quantified from four different culture conditions. Values are mean±SEM of triplicated dishes. ⁎pb0.05 from the other threeculture conditions.
571M. Jumabay et al. / Journal of Molecular and Cellular Cardiology 47 (2009) 565–575
572 M. Jumabay et al. / Journal of Molecular and Cellular Cardiology 47 (2009) 565–575
4. Discussion
The pathologic ventricular remodeling brought about by cardiacischemia leads to damage of the myocardium, which is replaced byfibrous scar tissue, resulting in the loss of cardiac function. In an effortto replace the cardiomyocytes lost after infarction, cellular transplan-tation has been investigated as a potential therapy. This has promptedan interest in the identification of cell types capable of replenishingthe injured myocardium with healthy cells and restoring heartfunction. Adult stem cells ontogenetically derived from the primitivemesoderm such as BM-MSCs and ASCs are attractive cell source in celltherapy for myocardial infarction, because of the ease of isolation and
their rapid growth in vitro while maintaining their differentiationability. In adult stem cells, BM-MSCs [16], resident cardiac stem cells[17], and ASCs [18] have been shown to directly differentiate intocardiomyocytes and improve cardiac function after transplantationinto the affected cardiac areas.
In the present study, we demonstrated that mature adipocyte-derived DFAT cells also could convert into cardiomyocyte phenotype.We have reported that the surface immunophenotype of DFAT cellsclosely resembles that of BM-MSCs andASCs (CD13+, CD29+. CD44+,CD90+, CD105+, CD34−, CD56−, HLA-DR−) [13]. Because BM-MSCsand ASCs are obtained by expansion of a very small number of stemcells from highly heterogeneous cell populations, by exploiting their
Fig. 6. DFAT cell transplantation promoted neovascularization in the infarcted myocardium. Eight weeks after cell transplantation, cardiac tissue was stained with isolectin B4 andanti-α-smooth muscle actin (ASMA) antibody. Nuclei were stained with DAPI. (A) Photomicrographs of immunostaining of capillary vessels in infracted myocardium. Note thatvascular formation was increased in the DFAT transplanted heart (DFAT) compared to that in the saline-injected heart (Control). Scale bars represent 50 μm. (B) The capillary vesseldensity was determined by counting the total number of isolectin B4-positive and ASMA-positive vessels in each heart by analyzing 20× optical fields. Results were given as totalnumber of vessels/area from five different sections. Bars represent means±SD. ⁎pb0.05 (Mann–Whitney U-test).
Fig. 5. Immunohistochemical analysis of rat hearts received DFAT cell transplantation. Three hours after ligation of the left coronary artery of SD rats, DFAT cells (1×106) or salinewere injected in the ischemic myocardium. (A) A photomicrograph of infracted cardiac tissue 3 days after DFAT cell transplantation. The sample was immunostained for GFP and theimmunoreactivity was visualized by DAB. Scale bar indicates 100 μm. (B–F) Photomicrographs of horizontal section of cardiac tissue 8 weeks after DFAT cell transplantation. Thesamples were double stained with anti-GFP and anti-sarcomeric actin antibodies. Nuclei were stained with DAPI. (B) GFP-positive DFAT cells were efficiently accumulated ininfracted areas (arrowheads), although the cells were also distributed in normal areas (arrows). Scale bar represents 1 mm. (C) Higher magnification views of the open square inpanel B. DFAT cells were engrafted in scarred cardiac tissue and expressed sarcomeric actin (Sr). Scale bars represent 200 μm. (D–F) Confocal microscopy confirmed sarcomeric actinexpression in DFAT cells. Scale bars represent 200 μm.
573M. Jumabay et al. / Journal of Molecular and Cellular Cardiology 47 (2009) 565–575
ability to adhere to plastic, these stem cell cultures are frequentlycontaminated by certain types of differentiated cells [19,20]. Our FACSanalysis revealed that ASCs at passage 1 contained high number ofsmooth muscle cells (18.6%), endothelial cells (2.7%), and blood cells(12.8–13.3%) [13]. Conversely, we observed thatmore than 97% of cellsisolated from adipose tissues used for ceiling cultures were lipid-filledadipocytes with a single nucleus, and approximately 40% of adherentadipocytes produced DFAT cells followed by rapid proliferation,resulting DFAT cell cultures are rarely (b0.1%) contaminated by othercell types, even at first passage. In addition, DFAT cells are reproduciblyisolated with significant expansion within a few passages fromapproximately 1 g of subcutaneous adipose tissue regardless of donor'sage [13]. These properties of DFAT cells may render them a safer andmore efficacious clinical cell therapy tool for application in regenerativemedicine.
In the present study, we showed that DFAT cells expressed cardiac-specific phenotypic markers when directly cocultured with neonatalcardiomyocytes. These findings suggest that DFAT cells can convertinto cardiomyocyte phenotype under appropriate microenvironmentwithout any modification of cells such as 5-azacytidine treatment[21]. We also showed that several cardiac-specific genes wereexpressed when DFAT cells were grown on neonatal cardiomyocyte-conditioned medium or when cultured by the transwell method.These findings suggest that soluble factors released from cardiomyo-cytes alone were sufficient to induce differentiation of DFAT cells intocardiac lineage and that physical contact between the cardiomyocytesand the DFAT cells is not requisite. Moreover, we found that DFAT cellsexpressed cardiac phenotypic markers when grown on semisolidmethylcellulose medium (MethoCult GF M3534) in the absence ofcardiomyocytes, indicating that DFAT cells have the potential tospontaneously convert to cardiomyocyte phenotype. Interestingly,our results demonstrated that percentage of troponin-T-positive cellsin DFAT cells cultured in MethoCult medium was significantly largerthan those in the other culture system. The MethoCult medium cansupport hematopoietic clonogenic expansion that allowed three-dimensional structure and microconcentration of cells [22]. Indeed,the semisolid culture makes DFAT cells aligned and branched to formclusters (see Figs. 4A–C). The three-dimensional culture environmentwith appropriate cytokines may be required to differentiation of DFATcells into cardiac phenotype. Planat-Benard et al. [18] reported thatASCs could differentiate into cardiomyocyte lineage in the MethoCultmedium. They showed that some ASC-derived cardiomyocyte-likecells had spontaneous beating activity although the frequency wasvery low (0.02–0.05%). However, we could not detect any sponta-neous beating activity in DFAT cells in the culture. Because ASCs areshown to be composed of stem cells/progenitor cells of differentdegree of maturity [23], ASCs but not DFAT cells may containimmature cell population that can differentiate into functionalcardiomyocytes. Recently, Rose et al. [24] identified a BM-MSCpopulation displaying plasticity toward the cardiomyocyte lineagewhile retaining MSC properties, including a nonexcitable electro-physiological phenotype. DFAT cells expressing cardiac markers maybe similar type of cells that retain both cardiac and mesenchymalprogenitor cell phenotype. To induce spontaneous beating activity inDFAT cells, certain genetic reprogramming procedures may berequired such as 5-azacytidine treatment or reprogramming factorstransduction.
To investigate the in vivo differentiation properties of DFAT cellsand to examine the effect of their transplantation into injured cardiactissue, DFAT cells were injected into the infarcted myocardium in rats.By 8 weeks after injection, GFP-positive DFAT cells were foundincorporated predominantly into the center of the infarct heart.Survival of engrafted cells may be especially compromised in thepresence of unresolved myocardial ischemia or inflammation. Themost of engrafted DFAT cells expressed cardiac sarcomeric actin,suggesting that DFAT cells differentiated to cardiomyocyte lineage in
vivo, although cell fusion events between the transplanted cells andnative cardiomyocytes cannot be ruled out [25,26]. However, cellfusion is not likely in this experiment because GFP- and sarcomericactin-double-positive cells were observed just center of infarct scarwhere it did not express sarcomeric actin. Kajstura et al. [27] arguedagainst cell fusion as a major mechanism of cardiac repair, since intheir study the transplantation of donor BM stem cells taken frommale GFP-positive mice into female mice following myocardialinfarction led to all GFP-positive cells having only one set of X and Ychromosomes. Our finding that engrafted DFAT cells in the infarctheart efficiently converted cardiomyocyte phenotype support thecommon assumption that the damaged tissue will direct and restrictthe cellular fate of transplanted adult stem cells [28]. We did not findany mature adipocytes and bone formation presumably derived fromDFAT cells in the infarct heart tissue at 8 weeks after transplantation,although the DFAT cells efficiently differentiate into mature adipo-cytes when the cells are injected subcutaneously [29]. Recently, it hasbeen reported that the injection of unfractioned BM cells or BM-MSCsinto infarcted myocardium carries a considerable risk for boneformation [30]. DFAT cells may lower the risk of uncontrolleddifferentiation including bone formation due to their homogenousproperty. Further studies are needed to evaluate the safety and long-term stability of DFAT cell transplants.
In the present study, the DFAT cell transplantation significantlyincreased capillary density in the infarcted area. It has been proposedthat the transplanted adult stem cells released soluble factors that,acting in a paracrine fashion, protect the heart, attenuate pathologicalventricular remodeling, induce neovascularization, and promoteregeneration [31]. Because we could not detect any GFP-positiveblood vessels in the in vivo study, it is possible that the transplantedDFAT cells induced neovascularization by releasing angiogeniccytokines and chemokines in the ischemic tissues. It has beenshown that BM-MSCs have ability to promote arteriogenesis viaproducing proarteriogenic factors as well as proangiogenic factors[32]. Our data showed that DFAT cell transplantation predominantlyincreased number of ASMA-positive small vessels, indicating thatDFAT cells also promote arteriogenesis. Our preliminary data ofcytokine array analysis revealed that DFAT cells express severalproangiogenic and proarteriogenic factors such as vascular endo-thelial growth factor-A (VEGF-A), placenta growth factor (PlGF),stromal cell-derived factor-1 (SDF-1), transforming growth factor-β1(TGF-β1), and platelet-derived growth factor-BB (PDGF-BB), some ofwhich are significantly upregulated by hypoxia (data not shown). InBM-MSCs and ASCs, the efficacy of cell transplantation therapies hasbeen reported to be mainly mediated by cytokines secreted by thetransplanted cells [33,34]. Paracrine mechanisms contributing to thereparative effects of DFAT cell transplantation may be predominantlymediated by the release of soluble cytokines as well as direct contactbetween DFAT cells and resident cardiac cells.
In conclusion, this study showed that DFAT cells convert tocardiomyocyte phenotype in vitro and in vivo. Moreover, thetransplantation of DFAT cells into infarcted cardiac tissue led toneovascularization in rats with myocardial infarction. Becauseadipose tissue is abundant and easily accessible at most ages inhumans, we propose that DFAT cells are an ideal cell source forcardiomyocyte regeneration in severe ischemic heart disease.
Acknowledgments
This work was supported in part by financial grants from theMinistry of Education, Science, Sports and Culture of Japan (20590707to T.M., 20591156 to H.M. and S0801033), by financial grants from theJapan Science and Technology Agency (08030216), and by a NihonUniversity Multidisciplinary Research Grant (08-020 and 09-012 toT.M.). We thank Mr. Yoshiki Taniguchi and Ms. Chii Yamamoto fortheir technical supports for our experiments.
574 M. Jumabay et al. / Journal of Molecular and Cellular Cardiology 47 (2009) 565–575
Appendix A. Supplementary data
Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.yjmcc.2009.08.004.
References
[1] Herzog EL, Chai L, Krause DS. Plasticity of marrow-derived stem cells. Blood2003;102:3483–93.
[2] Asahara T, Masuda H, Takahashi T, Kalka C, Pastore C, Silver M, et al. Bone marroworigin of endothelial progenitor cells responsible for postnatal vasculogenesis inphysiological and pathological neovascularization. Circ Res 1999;85:221–8.
[3] Tateishi-Yuyama E, Matsubara H, Murohara T, Ikeda U, Shintani S, Masaki H, et al.Therapeutic angiogenesis for patients with limb ischaemia by autologoustransplantation of bone-marrow cells: a pilot study and a randomised controlledtrial. Lancet 2002;360:427–35.
[4] Assmus B, Schachinger V, Teupe C, Britten M, Lehmann R, Dobert N, et al.Transplantation of Progenitor Cells and Regeneration Enhancement in AcuteMyocardial Infarction (TOPCARE-AMI). Circulation 2002;106:3009–17.
[5] Schachinger V, Assmus B, Britten MB, Honold J, Lehmann R, Teupe C, et al.Transplantation of progenitor cells and regeneration enhancement in acutemyocardial infarction: final one-year results of the TOPCARE-AMI Trial. J Am CollCardiol 2004;44:1690–9.
[6] Wollert KC, Meyer GP, Lotz J, Ringes-Lichtenberg S, Lippolt P, Breidenbach C, et al.Intracoronary autologous bone-marrow cell transfer after myocardial infarction:the BOOST randomised controlled clinical trial. Lancet 2004;364:141–8.
[7] Schachinger V, Erbs S, Elsasser A, Haberbosch W, Hambrecht R, Holschermann H,et al. Improved clinical outcome after intracoronary administration of bone-marrow-derived progenitor cells in acute myocardial infarction: final 1-yearresults of the REPAIR-AMI trial. Eur Heart J 2006;27:2775–83.
[8] Mitchell JB, McIntosh K, Zvonic S, Garrett S, Floyd ZE, Kloster A, et al.Immunophenotype of human adipose-derived cells: temporal changes instromal-associated and stem cell-associated markers. Stem Cells 2006;24:376–85.
[9] Zuk PA, Zhu M, Ashjian P, De Ugarte DA, Huang JI, Mizuno H, et al. Human adiposetissue is a source of multipotent stem cells. Mol Biol Cell 2002;13:4279–95.
[10] Halvorsen YD, Franklin D, Bond AL, Hitt DC, Auchter C, Boskey AL, et al.Extracellular matrix mineralization and osteoblast gene expression by humanadipose tissue-derived stromal cells. Tissue Eng 2001;7:729–41.
[11] Tholpady SS, Llull R, Ogle RC, Rubin JP, Futrell JW, Katz AJ. Adipose tissue: stemcells and beyond. Clin Plast Surg 2006;33:55–62 vi.
[12] Yagi K, Kondo D, Okazaki Y, Kano K. A novel preadipocyte cell line established frommouse adult mature adipocytes. Biochem Biophys Res Commun 2004;321:967–74.
[13] Matsumoto T, Kano K, Kondo D, Fukuda N, Iribe Y, Tanaka N, et al. Matureadipocyte-derived dedifferentiated fat cells exhibit multilineage potential. J CellPhysiol 2008;215:210–22.
[14] Sugihara H, Yonemitsu N, Miyabara S, Yun K. Primary cultures of unilocular fatcells: characteristics of growth in vitro and changes in differentiation properties.Differentiation 1986;31:42–9.
[15] Chen J, Larsson L, Haugen E, Fedorkova O, Angwald E, Waagstein F, et al. Effects ofautoantibodies removed by immunoadsorption from patients with dilatedcardiomyopathy on neonatal rat cardiomyocytes. Eur J Heart Fail 2006;8:460–7.
[16] Pittenger MF, Martin BJ. Mesenchymal stem cells and their potential as cardiactherapeutics. Circ Res 2004;95:9–20.
[17] Beltrami AP, Barlucchi L, Torella D, Baker M, Limana F, Chimenti S, et al. Adultcardiac stem cells are multipotent and support myocardial regeneration. Cell2003;114:763–76.
[18] Planat-Benard V, Menard C, Andre M, Puceat M, Perez A, Garcia-Verdugo JM, et al.Spontaneous cardiomyocyte differentiation from adipose tissue stroma cells. CircRes 2004;94:223–9.
[19] Clark BR, Keating A. Biology of bone marrow stroma. Ann N Y Acad Sci 1995;770:70–8.
[20] Zuk PA, ZhuM,Mizuno H, Huang J, Futrell JW, Katz AJ, et al. Multilineage cells fromhuman adipose tissue: implications for cell-based therapies. Tissue Eng 2001;7:211–28.
[21] Makino S, Fukuda K, Miyoshi S, Konishi F, Kodama H, Pan J, et al. Cardiomyocytescan be generated from marrow stromal cells in vitro. J Clin Invest 1999;103:697–705.
[22] Miller CL, Lai B. Human and mouse hematopoietic colony-forming cell assays.Methods Mol Biol 2005;290:71–89.
[23] Guilak F, Lott KE, Awad HA, Cao Q, Hicok KC, Fermor B, et al. Clonal analysis of thedifferentiation potential of human adipose-derived adult stem cells. J Cell Physiol2006;206:229–37.
[24] Rose RA, Jiang H, Wang X, Helke S, Tsoporis JN, Gong N, et al. Bone marrow-derived mesenchymal stromal cells express cardiac-specific markers, retain thestromal phenotype, and do not become functional cardiomyocytes in vitro. StemCells 2008;26:2884–92.
[25] Alvarez-Dolado M, Pardal R, Garcia-Verdugo JM, Fike JR, Lee HO, Preffer K, et al.Fusion of bone-marrow-derived cells with Purkinje neurons, cardiomyocytes andhepatocytes. Nature 2003;423:968–73.
[26] Nagaya N, Fujii T, Iwase T, Ohgushi H, Itoh T, Uematsu M, et al. Intravenousadministration of mesenchymal stem cells improves cardiac function in rats withacute myocardial infarction through angiogenesis and myogenesis. Am J PhysiolHeart Circ Physiol 2004;287:H2670–6.
[27] Kajstura J, Rota M, Whang B, Cascapera S, Hosoda T, Bearzi C, et al. Bone marrowcells differentiate in cardiac cell lineages after infarction independently of cellfusion. Circ Res 2005;96:127–37.
[28] LaflammeMA, Murry CE. Regenerating the heart. Nat Biotechnol 2005;23:845–56.[29] Nobusue H, Endo T, Kano K. Establishment of a preadipocyte cell line derived from
mature adipocytes of GFP transgenic mice and formation of adipose tissue. CellTissue Res 2008;332:435–46.
[30] BreitbachM, Bostani T, Roell W, Xia Y, Dewald O, Nygren JM, et al. Potential risks ofbone marrow cell transplantation into infarcted hearts. Blood 2007;110:1362–9.
[31] Gnecchi M, Zhang Z, Ni A, Dzau VJ. Paracrine mechanisms in adult stem cellsignaling and therapy. Circ Res 2008;103:1204–19.
[32] Kinnaird T, Stabile E, Burnett MS, Lee CW, Barr S, Fuchs S, et al. Marrow-derivedstromal cells express genes encoding a broad spectrum of arteriogenic cytokinesand promote in vitro and in vivo arteriogenesis through paracrine mechanisms.Circ Res 2004;94:678–85.
[33] Kinnaird T, Stabile E, Burnett MS, Epstein SE. Bone-marrow-derived cells forenhancing collateral development: mechanisms, animal data, and initial clinicalexperiences. Circ Res 2004;95:354–63.
[34] Nakagami H, Maeda K, Morishita R, Iguchi S, Nishikawa T, Takami Y, et al. Novelautologous cell therapy in ischemic limb disease through growth factor secretionby cultured adipose tissue-derived stromal cells. Arterioscler Thromb Vasc Biol2005;25:2542–7.
575M. Jumabay et al. / Journal of Molecular and Cellular Cardiology 47 (2009) 565–575
