Bcl-2 Engineered MSCs Inhibited Apoptosis and Improved HeartFunction

WENZHONG LI,a NAN MA,a LEE-LEE ONG,a CATHARINA NESSELMANN,a CHRISTIAN KLOPSCH,a

YURY LADILOV,a DARIO FURLANI,a CHRISTOPH PIECHACZEK,b JEANNETTE M. MOEBIUS,b KAROLA LUTZOW,c

ANDREAS LENDLEIN,c CHRISTOF STAMM,d REN-KE LI,e GUSTAV STEINHOFFa

aDepartment of Cardiac Surgery, University of Rostock, Rostock, Germany; bMiltenyi Biotec, Bergisch Gladbach,Germany; cInstitute of Polymer Research, GKSS Forschungszentrum, Teltow, Germany; dCardiac Surgery, GermanHeart Institute Berlin, Berlin, Germany; eDivisions of Cardiovascular Surgery and Cardiology, Toronto GeneralHospital and the University of Toronto, Toronto, Ontario, Canada

Key Words. Gene therapy • Stem cells • Transplantation • Angiogenesis • Antiapoptosis

ABSTRACT

Engraftment of mesenchymal stem cells (MSCs) derivedfrom adult bone marrow has been proposed as a potentialtherapeutic approach for postinfarction left ventricular dys-function. However, limited cell viability after transplanta-tion into the myocardium has restricted its regenerativecapacity. In this study, we genetically modified MSCs withan antiapoptotic Bcl-2 gene and evaluated cell survival,engraftment, revascularization, and functional improve-ment in a rat left anterior descending ligation model viaintracardiac injection. Rat MSCs were manipulated to over-express the Bcl-2 gene. In vitro, the antiapoptotic and para-crine effects were assessed under hypoxic conditions. Invivo, the Bcl-2 gene-modified MSCs (Bcl-2-MSCs) were in-jected after myocardial infarction. The surviving cells weretracked after transplantation. Capillary density was quan-tified after 3 weeks. The left ventricular function was eval-

uated by pressure-volume loops. The Bcl-2 gene protectedMSCs against apoptosis. In vitro, Bcl-2 overexpression reducedMSC apoptosis by 32% and enhanced vascular endothelialgrowth factor secretion by more than 60% under hypoxicconditions. Transplantation with Bcl-2-MSCs increased 2.2-fold, 1.9-fold, and 1.2-fold of the cellular survival at 4 days, 3weeks, and 6 weeks, respectively, compared with the vector-MSC group. Capillary density in the infarct border zone was15% higher in Bcl-2-MSC transplanted animals than in vector-MSC treated animals. Furthermore, Bcl-2-MSC transplantedanimals had 17% smaller infarct size than vector-MSC treatedanimals and exhibited functional recovery remarkably. Ourcurrent findings support the premise that transplantation ofantiapoptotic gene-modified MSCs may have values for medi-ating substantial functional recovery after acute myocardialinfarction. STEM CELLS 2007;25:2118–2127

Disclosure of potential conflicts of interest is found at the end of this article.

INTRODUCTION

Cell transplantation utilizing fetal myocardial tissues [1], em-bryonic stem cells [2], skeletal myoblasts [3, 4], cardiomyocytes[5, 6], smooth muscle cells [7, 8], cardiac stem cells [9], bonemarrow cells [10], and hematopoietic stem cells [11] hasemerged as a promising therapeutic approach for the restorationof heart functions following myocardial infarction damage.Considerable experimental and clinical evidences have demon-strated that using different cell types to replace the necrotictissue in the myocardium is safe and can contribute to theimprovement of angiogenesis and heart functions [12–17].Among the various cell types investigated, bone marrow mes-enchymal stem cells (MSCs) are self-renewing clonal precursorsof nonhematopoietic stromal tissues [18]. They can be isolatedfrom the bone marrow or adipose tissue and expanded in culturebased on their ability to adhere to culture dishes and proliferatein vitro. MSCs are multilineage potential cells [19] and can giverise to osteoblasts, chondrocytes [20], neurons [21], skeletalmuscle [22], and cardiac muscle [23, 24] under appropriate

conditions. Furthermore, studies on human, baboon, and murineMSCs showed that MSCs are immunosuppressive [25, 26].Therefore, MSCs appear to be an appealing cell source forcardiac transplantation [13, 27] because of the ease of harvestand expansion ex vivo. However, the low cellular survival rateafter transplantation into an infarcted heart within the first fewdays engenders only marginal functional improvement [24, 28].Thus, it is necessary to reinforce MSCs against the arduousmicroenvironment incurred from ischemia, inflammatory re-sponse, and proapoptotic factors in order to improve the efficacyof cell therapy.

The 26-kDa Bcl-2 antiapoptotic protein belongs to the Bcl-2family of proteins, which was originally found to be overex-pressed in B-cell lymphoma [29]. It serves as a critical regulatorof pathways involved in apoptosis, acting to inhibit cell death[30]. Bcl-2 gene was found to be upregulated in failing hearts[31, 32] as well as aging hearts [33]. It acts to prevent pro-grammed cell death of ventricular myocytes [34].

Previous studies have demonstrated that Bcl-2 is the regu-lator of the metabolic functions of mitochondria during ischemicconditions, which can contribute to both cardiac and neuronal

Correspondence: Nan Ma, M.D., Ph.D., Department of Cardiac Surgery, University Rostock, Schillingallee 69, 18057 Rostock, Germany.Telephone: �49-381-494 6105; Fax: �49-381-494 6214; e-mail: nan.ma@med.uni-rostock.de Received November 24, 2006; accepted forpublication April 20, 2007; first published online in STEM CELLS EXPRESS May 3, 2007. ©AlphaMed Press 1066-5099/2007/$30.00/0 doi:10.1634/stemcells.2006-0771

TRANSLATIONAL AND CLINICAL RESEARCH: MESENCHYMAL STEM CELLS SERIES

STEM CELLS 2007;25:2118–2127 www.StemCells.com

protection under various stresses [34, 35]. The overexpressedBcl-2 can delay the onset of cell death and modestly augmentviable cell growth in the first 48 hours of apoptosis [36]. Bcl-2deficient mice showed abnormalities with loss of a death repres-sor in specific cells [37]. The protective effects of Bcl-2 forconditions of the diseased heart have been approved by previousstudies [34, 38]. Interestingly, overexpressing Bcl-2 in trans-genic mice improved the heart function and inhibited cardiomy-ocyte apoptosis [39].

In this study, adult rat bone marrow-derived MSCs weregenetically modified to overexpress Bcl-2, which aimed at viv-ifying the stem cells and enhancing their resistance to ischemicconditions from acute myocardial infarction after transplanta-tion into the heart. We proposed that genetic modification ofstem cells with Bcl-2 would armor stem cells settling into adeteriorative ischemic microenvironment and improve stem cellviability in the early post-transplanted period, thereby enhancingcardiac functional recovery after acute myocardial infarction.

MATERIALS AND METHODS

Isolation and Culturing of Rat MSCsMale Lewis rats were obtained from Charles River Laboratories(Wilmington, MA, http://www.criver.com). All animals receivedhumane care in compliance with the Guide for the Care and Use ofLaboratory Animals (NIH publication number 85–23, revised1996). Rat MSCs were isolated from the bone marrow of the femursand tibias of rats (Charles River Laboratories). A 21-gauge needlewas inserted into the shaft of the bone and flushed with 30 ml ofcomplete �-modified Eagle’s medium (�-minimal essential me-dium) containing 20% fetal bovine serum (lot selected for promot-ing rapid expansion of MSCs; HyClone, Logan, UT, http://www.hy-clone.com), 2 mM L-glutamine (PAA Laboratories, Linz, Austria,http://www.paa.at), 100 U/ml penicillin (PAA), 100 �g/ml strepto-mycin (PAA), and 25 ng/ml amphotericin B (PAA). The harvestedcells were filtered through a 70-�m nylon filter (Falcon; Becton,Dickinson and Company, Franklin Lakes, NJ, http://www.bd.com)and plated into one 75-cm2 flask per rat. The cells were grown at37°C and 5% CO2 for 3 days before the medium was replaced. Theadherent cells were grown to 90% confluence to obtain sampleshere defined as passage 0 cells.

Fluorescence-Activated Cell Sorter AnalysisMSCs at passage 3 were analyzed for purity and epitope expressionusing fluorescence-activated cell sorter (FACS) analysis. The cellswere blocked with Fc Block blocking reagent (BD Biosciences, SanDiego, http://www.bdbiosciences.com) and incubated for 10 min-utes at 4°C with the following antibodies: anti-CD29.PE (cloneHa2/5; BD Biosciences), anti-CD44.FITC (Serotec Ltd., Oxford,U.K., http://www.serotec.com), anti-CD90 (Thy1).FITC (cloneMRC OX-7; Abcam, Cambridge, U.K., http://www.abcam.com),anti-CD45.PE (BD Biosciences), and anti-CD34.FITC (BD Bio-sciences). Isotype controls were purchased from BD Pharmingen(San Diego, http://www.bdbiosciences.com/index_us.shtml). Afterincubation, the cells were washed with 2 mM phosphate-bufferedsaline (PBS)/EDTA and analyzed using a FACSCalibur flow cy-tometer (Becton, Dickinson). Dead cells were excluded using pro-pidium iodide (PI) staining. Data analysis was performed withCellQuest software (BD Biosciences). Histograms of cell numberversus logarithmic fluorescence intensity were recorded for 10,000–20,000 cells per sample.

Genetic Modification of MSCsA total number of 4 � 105 cells per well were plated in a 6-wellplate 24 hours before transfection; jetPEI mediated transfection wasperformed according to the protocol given by the supplier(Polyplus-transfection, New York, http://www.polyplus-transfec-tion.com). In brief, dilution of 3.0 �g of DNA (pEGFP-N1,

pcDNA3.1/Bcl-2, or pcDNA3.1) and 6 �l of jetPEI per well wascarried out in 150 mM NaCl separately. The transfection mixturewas mixed immediately and added to the adherent MSCs with 1 mlof complete medium. Transfection using jetPEI was optimizedaccording to the supplier’s instructions by varying the amount ofDNA (pEGFP-N1) and the volumes of transfection reagent at thenumber of polynitrogen (N) per DNA phosphate (P) ratios between3:1 and 12:1. Determination of transfection efficiency was per-formed 24 hours after transfection by fluorescence microscopy(Leica, Heerbrugg, Switzerland, http://www.leica.com) and byFACS. For each experiment, at least three microscopic visual fields(200-fold magnification) were counted, and the ratios of enhancedgreen fluorescent protein-expressing cells to nonfluorescent cellswere calculated. MSCs without transfection and those transfectedwith pcDNA3.1/Bcl-2 or pcDNA3.1 are termed “MSCs,” “Bcl-2-MSCs,” and “vector-MSCs,” respectively. Bcl-2 protein expressionlevel was evaluated by Western blotting. Before cell injection,medium was removed from the cultures, cell layers were washedwith PBS, and cells were harvested by incubation with 0.25%trypsin/EDTA. All experiments and cell number determinationswere performed in triplicate.

In Vitro Functional Differentiation AssayTo induce adipogenic differentiation, MSCs after genetic modifica-tion were seeded at a density of 3 � 103 cells per cm2 and culturedfor up to 3 weeks in cell culture medium supplemented with 10�8

M dexamethasone, 2.5 �g/ml insulin, and 100 �M indomethacin.To induce chondrogenic differentiation, 3 � 105 MSCs were cul-tured in 1 ml of chondrogenic induction medium (cell culturemedium supplemented with 0.1 �M dexamethasone, 1 mM sodiumpyruvate, 0.17 mM L-ascorbic acid 2-phosphate, 0.35 mM L-proline,6.25 �g/ml insulin, 6.25 �g/ml transferrin, 6.25 ng/ml selenite, 5.33�g/ml linolic acid, 1.25 mg/ml bovine serum albumin, and 0.01�g/ml transforming growth factor-�3) in the tip of a 15-ml conicaltube to allow aggregation of the cells in suspension culture. Theinduction of chondrogenic differentiation was performed for 4weeks. To induce cardiomyocyte differentiation, MSCs were treatedwith the DNA demethylation agent 5-azacytidine following previ-ously reported protocols. The differentiation capacity toward dif-ferent cell lineages was verified by morphology changes and im-munostaining for specific markers, that is, aggrecan forchondrocytes, fatty acid binding protein (FABP-4) for adipocytes,and Homeobox protein NK-2 homolog E (Nkx2.5) for cardiomyo-cytes.

Western Blot AnalysisThe genetically modified MSCs were treated with cell lysis buffer(Promega, Madison, WI, http://www.promega.com). The proteinconcentration of the samples was determined by BCA protein assay(Pierce, Rockford, IL, http://www.piercenet.com). Twenty �g oftotal proteins were resolved by 12.5% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to a 0.2-mmnitrocellulose membrane (Schleicher & Schuell, Dassel, Germany,http://www.whatman.com). The membrane was blocked in PBSbuffer containing 0.2% Tween 20 and 5% skim milk overnight at4°C. Subsequently, the blot was incubated with human Bcl-2 pro-tein monoclonal antibody (1:200 dilution; Dako, Glostrup, Den-mark, http://www.dako.com) for 3 hours. Housekeeping protein�-actin was employed as loading control. Antibody binding wasdetected with horseradish peroxidase conjugated anti-mouse sec-ondary antibody (1:2,000; Pierce) and visualized by ECL kit (Am-ersham Biosciences, Piscataway, NJ, http://www.amersham.com).Proteins prepared from MSCs transfected with vector alone servedas a control.

Apoptosis AssaysMSCs grown on glass coverslips were treated with hypoxic precon-ditioning of 95% nitrogen and 5% carbon dioxide. At 24 hoursposthypoxic treatment, apoptotic cells were identified using theterminal deoxynucleotidyl transferase-mediated dUTP end labeling(TUNEL) method by staining the cells using the Chemicon Ap-opTag Assay Kit and counterstaining with PI. The coverslips were

2119Li, Ma, Ong et al.

www.StemCells.com

washed four times with PBS and mounted in FluorSave (Calbio-chem, San Diego, http://www.emdbiosciences.com). Apoptotic nu-clei were also assessed and quantified by staining with Hoechst33342 (Molecular Probes, Eugene, OR, http://probes.invitrogen-.com). Image acquisition was performed under phase contrast 400�on a Leica DMLB microscope. A total of 20 microscopic fieldswere quantitated for each coverslip. The results are representative ofthree independent experiments replicated with six different cover-slips.

Detection of Vascular Endothelial Growth FactorSecretionThe secretion of vascular endothelial growth factor (VEGF) by ratMSCs was evaluated by the Quantikine rat VEGF immunoassay(R&D Systems Inc., Minneapolis, http://www.rndsystems.com) ac-cording to the manufacturer’s protocol.

Cell LabelingTo identify the transplanted MSCs in the hearts, the cells werelabeled with bromodeoxyuridine (BrdU) (Sigma-Aldrich, St. Louis,http://www.sigmaaldrich.com) at a final concentration of 10 �mol/lin cell culture medium for 12 hours. The efficiency of labeling was42.1% � 3.5% (n � 3), which was evaluated by immunostaining ofthe cytospun cells with BrdU monoclonal antibody (Lab Vision,Fremont, CA, http://www.labvision.com). In addition, the cell en-graftment was also validated with 5-(and 6)-carboxyfluoresceindiacetate succinimidyl ester (Vybrant CFDA SE Cell Tracer; Mo-lecular Probes; 5 �M) labeling. The harvested cells were resus-pended in culture medium at a density of 106 per 100 �l and kepton ice (less than 1 hour) until transplantation.

Myocardial Infarction and MSC TransplantationMyocardial infarction was induced at 8–12 weeks of age (approx-imately 280 g of body weight). Rats were anesthetized by intraperi-toneal injection with Pentobarbital (50 mg/kg of body weight;Roche Diagnostics, Basel, Switzerland, http://www.roche-applied-science.com), endotracheally intubated, and mechanically venti-lated. The heart was exposed via a left thoracotomy. Anteriormyocardial infarction was created by ligation of the left anteriordescending artery with a 6–0 silk suture (myocardial infarction[MI]�) (supplemental online Fig. 1A). In the sham group, rats weresham operated (left thoracotomy without coronary artery ligation[MI�]). Successful infarction was determined by observing a palediscoloration of the left ventricular muscle and an ST elevation onelectrocardiograms. Immediately after ligation, 6 � 106 vector-MSCs or Bcl-2-MSCs were injected at six injection sites intoanterior and lateral aspects of the viable myocardium bordering theinfarction with a 31-gauge needle (BD Biosciences) (supplementalonline Fig. 1B, 1C). A further MI� rats with Dulbecco’s modifiedEagle’s medium injection served as medium controls. Animals weresacrificed at 4 days, 3 weeks, and 6 weeks after cell transplantation.Cell survival was examined at 4 days, 3 weeks, and 6 weeks.Angiogenesis, infarction area, and cell colocalization were evalu-ated at 3 weeks. The pressure-volume loop analysis was performedat 6 weeks.

Histological AnalysisAt the end of the observation period, animals were sacrificed. Thehearts were removed, washed with PBS, weighed, and snap frozenin liquid nitrogen. Frozen sections embedded in optimum cuttingtemperature medium (5 �m in thickness) were prepared and stainedwith Sirius red/Fast Green or hematoxylin-eosin. From every heart,three slices on mid level of the infarcted region were used formeasurement of scar size. Quantification of scar area and length wasperformed using imaging software ESI-Vision (Soft Imaging Sys-tem, Munster, Germany). The scar area was calculated as a percentof the whole area of the left ventricular section.

Quantitative Analysis of Cell SurvivalQuantitative analysis of BrdU-positive cells was performed on theimmunostained frozen sections. From every heart, five slices on mid

level of the infarcted region were used for measurement. Themyocardium extending 0.5–1.0 mm from the infarcted tissue orinfarct scar was considered to represent the border zone myocar-dium. To avoid contamination of the remote myocardium withborder zones, a myocardial area extending �1–2 mm from theborder zone area was not included in the statistical analysis [40]. Atleast 20 representative microscopic fields (with oil immersion phasecontrast 1,000� by a Leica DMLB) of each slide were randomlyselected from the region of interest and digitally photographed.BrdU positive cells were counted in the tissue bordering infarction(4 days after infarction), in the border zones of infarct scars (3 and6 weeks after infarction) with Leica IM 5 (V2.01) software. Cellsurvival was expressed as the proportion of the BrdU-positivenuclei to the total number of nuclei. The data were analyzed by twoinvestigators who were blinded with respect to the cell treatment.

Determination of Infarct SizeHeart tissue sections were stained with hematoxylin-eosin andSirius red F3BA (0.1% picric acid; Sigma-Aldrich). The infarctedarea with an increase in collagen content was shown as Siriusred-positive and was measured by computerized planimetry. Theratio of scar length and entire circumference defined the infarctextent for the endocardial and epicardial surfaces, respectively. Thefinal infarct size was determined as the average of endocardial andepicardial surfaces and was given in percent. Quantitative assess-ment of collagen deposition was performed with a multipurposecolor image processor (Envision; Dako). An investigator blinded tothe treatment performed the analysis.

ImmunostainingFor immunohistological detection of transplanted cells, frozen tis-sue sections were incubated with monoclonal mouse anti-BrdUantibody (Lab Vision). After blocking in Envision blocking buffer(Dako), sections were placed in primary antibody overnight at 4°Cto 8°C. On the following day, the sections were incubated withAlexa 568 (Molecular Probes) or horseradish peroxidase (SantaCruz Biotechnology Inc., Santa Cruz, CA, http://www.scbt.com)conjugated goat anti-mouse IgG. Endothelial-like cells were stainedfor von Willebrand factor (vWF) (primary antibody: polyclonalanti-vWF antibody, H-300, Santa Cruz; secondary antibody: Alexa568, Molecular Probes). Nuclei were counterstained with 4,6-dia-midino-2-phenylindole (Sigma-Aldrich). Cardiac-like cells werestained for Troponin T (primary antibody: monoclonal anti-Tropo-nin T antibody, cardiac isoform, clone CT-3, Lab Vision; secondaryantibody: Alexa 568, Molecular Probes). For colocalization study,BrdU-positive cells were identified (primary antibody: monoclonalmouse anti-BrdU antibody; secondary antibody: Alexa 488, Molec-ular Probes).

Determination of Capillary DensityThe capillary density in both infarcted and normal myocardial tissuewas determined as described by Weidner et al. [41]. Tissue sectionswere stained using anti-vWF antibody. For quantification of posi-tively stained vessels, five sections within the infarct zone of eachanimal were analyzed by an investigator who was blinded withrespect to the cell treatment. Capillaries were counted in 10 ran-domly chosen high-power fields (HPFs, 400�) in two sections peranimal. The results were expressed as capillaries per high powerfield.

Left Ventricular CatheterizationPressure-volume (PV) loops were recorded under isoflurane (2%)anesthesia with a conductance catheter (Millar 2 Fr catheter modelSPR-838 [Millar Instruments, Houston, TX, http://www.millarin-struments.com]) connected to Aria/Power Lab data-acquisitionhardware (Millar/Power Lab) by an investigator blinded to thetreatment 6 weeks after cell transplantation. The cuvette volumecalibration was performed according to the PVAN Pressure-VolumeAnalysis Software User’s Guide. All data were analyzed off-linewith IOX software (EMKA Technologies, Paris, http://www.emka-tech.com). Maximum dP/dt (max dP/dt), minimum dP/dt (min dP/dt), end-diastolic volume (EDV), end-systolic volume (ESV), and

2120 Bcl-2 Engineered MSCs

ejection fraction (EF) were measured from the steady-state PVloops. End-systolic pressure-volume relationship (ESPVR), vol-ume-axis intercept (V0), and maximal elastance (Emax) were ob-tained from PV loops measured by reducing preload through occlu-sion of the inferior vena cava.

Statistical AnalysisAll values are presented as mean � SD. One-way analysis ofvariance with Scheffe’s post hoc test for unequal sample sizes wasused to compare numeric data among the four experimental groups.Datasets consisting of two groups only were compared with the useof unpaired Student’s t tests. A level of p � .05 was consideredstatistically significant.

RESULTS

MSCs Express Markers Distinct from HematopoieticStem CellsCell isolation, expansion, characterization, and differentiation ofrat MSCs have been established according to previous reports[19]. Primary culture of the marrow cells was performed ac-cording to Dexter’s method [42]. Due to the presence of verylittle extracellular matrix in the bone marrow, gentle mechanicaldisruption can readily dissociate the stroma and hematopoieticcells into a single cell suspension. When plating the cells at lowdensity, the MSCs adhere and can be easily separated from thehematopoietic stem cells by repeated washing. The morphologyof rat MSCs from bone marrow displayed a homogenous spin-dle-shaped population and maintained a similar morphologyduring the subsequent passages. FACS analysis was employedto identify the surface marker expression. The MSC culture wasshown to be devoid of CD34 and CD45 (Fig. 1D, 1E), which arethe markers for hematopoietic cells. In contrast, a high expres-sion of CD29, CD44, and CD90 markers were observed (Fig.1A–1C). The results from FACS analysis were confirmed byimmunocytochemistry (data not shown).

Overexpression of Bcl-2 in Genetically ModifiedMSCsThe transfection efficiency of Polyethylenimine-TransfectionKit (Polyplus Transfection SA, Illkirch, France) for geneticallymodified MSCs was evaluated by pcDNA3.1/green fluorescentprotein (GFP) as an internal control after 24-hour gene delivery.

A representative GFP expression is shown in Figure 2A. Morethan 50% of MSCs were transfected based on FACS analysis(Fig. 2B). Bcl-2-MSCs and vector-MSCs have a phenotypesimilar to MSCs 3 days after transfection (data not shown). Toevaluate the expression of Bcl-2 protein in the genetically mod-ified MSCs in vitro, Western blot analysis was performed oncell samples at various time points (24 hours, 72 hours, 7 days,and 21 days) post-transfection. A significantly high expressionof Bcl-2 was observed in Bcl-2-MSCs as early as 24 hours,which remained detectable on day 7 (Fig. 2C). In contrast, aweak expression of Bcl-2 was detected in vector-MSCs 24 hoursafter transfection, indicating the low level of endogenous Bcl-2proteins in MSCs. The expression of internal housekeeping�-actin gene was at the same level. Our observation showed thatthe MSCs were successfully modified with Bcl-2. The upregu-lation of Bcl-2 was transient, and this may minimize the risk ofmalignant transformation or late cell failure. No apparentchange of MSC phenotype after transfection was observed (datanot shown).

Bcl-2 Modified MSCs Maintain TheirMultidifferentiation CapacityMSCs can differentiate into multiple lineages (such as bone,cartilage, adipose tissue, and cardiac muscle), and this ability istaken as a functional criterion defining MSC precursor cells. Toverify whether the Bcl-2 modification affected the differentia-tion capacity, the genetically modified MSCs underwent myo-genic, adipogenic, and chondrogenic differentiation using themethods previously described [23, 43] at 48 hours post-trans-fection. After 21 days of induction toward an adipogenic lin-eage, a characteristic morphological change with accumulationof lipid vacuoles was observed (Fig. 3A). Immunostaining re-vealed the presence of FABP-4, which is a marker protein foradipocytes. Chondrogenesis was assessed by immunostainingfor aggrecan after 4 weeks of culture under chondrogenic con-ditions. The chondrocyte-like cells showed positive staining foraggrecan protein (Fig. 3B–3D). For cardiomyocyte differentia-

Figure 1. Immunophenotypic profile of MSCs. Flow cytometry histo-grams after three passages show the expression (unshaded) of selectedsurface molecules; (A): CD29; (B): CD44; (C): CD90; (D): CD34; and(E): CD45. Control cells labeled without primary antibodies (shaded).The rat MSCs were positive for CD29, CD44, and CD90 but negativefor CD34 and CD45 surface markers, which are commonly found onhematopoietic cells. Figure 2. Transfection of MSCs with nonviral vector. (A): Represen-

tative photomicrograph of MSCs transfected with pcDNA3.1/GFP. (B):Fluorescence-activated cell sorter analysis showing the transfectionefficiency of GFP transfected MSCs (unshaded) compared with untrans-fected MSCs (shaded). (C): Representative Western blots showing thatBcl-2 was overexpressed in Bcl-2 modified MSCs and lasted for 7 daysafter transfection. Housekeeping protein �-actin served as loading con-trol. Abbreviations: d, days; GFP, green fluorescent protein; h, hours.

2121Li, Ma, Ong et al.

www.StemCells.com

tion, Bcl-2-MSCs were exposed to 5-azacytidine for 24 hours.After two weeks, cardiomyocyte-like cells and positive immu-nostaining for Nkx2.5 were noted (Fig. 3E–3G). Taken together,this evidence indicated that Bcl-2-MSCs retain their multidif-ferentiation potential into adipogenic, chondrogenic, and cardi-omyocyte lineages.

Bcl-2 Modified MSCs Protect Against ApoptosisIn VitroTo test the capability of Bcl-2-MSCs to protect against apopto-sis in vitro, the modified MSCs were treated under hypoxicconditions for 24 hours. The evaluation of apoptosis was carriedout using TUNEL assay (Fig. 4A) and Hoechst 33342 staining(Fig. 4B). Under hypoxia, the rates of cell apoptosis in MSCsand vector-MSCs exceeded those of Bcl-2-MSCs by 1.5-fold.The apoptotic cell number was significantly reduced by theBcl-2 genetic modification. Quantitative assay showed that thenumber of TUNEL-positive cells was decreased from 53% �3.5% of MSCs and 51% � 4.7% of vector-MSCs to 36% �1.8% (n � 6, p � .001) of Bcl-2-MSCs. However, there was nosignificant difference in apoptotic cell numbers among MSCs,vector-MSCs, and Bcl-2-MSCs when evaluated under nor-moxia. A further confirmatory assay was based on nuclear

chromatin morphology after staining with Hoechst 33342. Ap-optosis was evidenced by cell shrinkage, nuclear condensation,and DNA fragmentation that occurred in cells treated withhypoxia (Fig. 4B). The quantitative assay based on Hoechststaining showed that the number of apoptotic cells was de-creased from 46% � 3.1% of MSCs and 45% � 4.2% ofvector-MSCs to 30% � 2.6% of Bcl-2-MSCs (n � 6, p � .001).These findings demonstrated that the Bcl-2 modified MSCscould protect against apoptosis under hypoxic conditions. It isanticipated that Bcl-2 modified MSCs may retain their antiapop-totic properties in the ischemic myocardium.

Bcl-2 Modified MSCs Upregulated AngiogenicCytokine VEGF Under HypoxiaTo identify potential paracrine mechanisms responsible for thetherapeutic effect of Bcl-2-MSCs, we examined the secretion ofVEGF. MSCs, vector-MSCs, and Bcl-2-MSCs were culturedunder either normoxic or hypoxic conditions for up to 24 hours,and VEGF secretion was quantitated by Quantikine rat VEGFimmunoassay. A more than 60% increase in secretion of VEGFwas detected in Bcl-2-MSCs compared with MSCs alone andvector-MSCs when the cells were cultured under hypoxic con-ditions (Fig. 4C). However, there was no significant differenceamong the groups when the cells were cultured under normoxicconditions. The angiogenic cytokine VEGF was significantlyupregulated in Bcl-2-MSCs in response to hypoxic conditions.The high-level expression of VEGF from Bcl-2-MSCs mayprovide cardioprotective and proangiogenic effects.

Cell Engraftment with Bcl-2 Modified MSCs in theInfarcted HeartTo assess the efficacy of Bcl-2-MSC transplantation, 6 � 106

Bcl-2-MSCs or vector-MSCs from passage 3 were transplantedinto the viable left ventricular (LV) myocardium borderinginfarction. Medium was injected into control animals. The as-sessment of cell engraftment was carried out 4 days, 3 weeks,and 6 weeks after cell transplantation. BrdU-positive nucleiwere identified by immunostaining (Fig. 5A–5E). Figure 5A and5B depicts representative images 4 days following cell trans-plantation. Figure 5C–5E shows representative images of Bcl-2-MSC cells in the heart of a MI� rat sacrificed 3 weeks aftercell injection. The number of surviving cells in the Bcl-2-MSCgroup was greater than in the vector-MSCs following cell trans-plantation, and the differences proved to be statistically signif-icant (Fig. 5F), with 2.2-fold enhancement of cell survival onday 4 (n � 10, p � .01), 1.9-fold enhancement on day 21 (n �9, p � .05), and 1.2-fold enhancement after 6 weeks (n � 6, p �.05). Manipulation of MSCs with Bcl-2 gene enhanced cellularsurvival after myocardial infarction.

Furthermore, double immunofluorescence staining withBrdU and anti-vWF antibody revealed that at least some of theMSCs (3.1% � 1.3%) appeared to display endothelial cell-likephenotype. Occasionally, blood vessels consisted of BrdU-pos-itive cells (Fig. 6A, 6B). Three weeks after cell transplantation,we observed a very low number of MSCs colocalized withcardiac Troponin T (cTnT) (Fig. 6C–6E). The colocalizationwas further confirmed by the three-dimensional reconstructionof the tissue acquired by a Leica TCP2 confocal microscope(Fig. 6F–6H). The frequency of cTnT-BrdU double-positivecells from the engrafted stem cells was extremely low (Bcl-2-MSC group 0.05% � 0.02% cTnT). There was no significantdifference between Bcl-2-MSCs and the MSC group. It was notclear whether the transplanted cells had fused or differentiatedinto cardiomyocytes. There is no evidence for tumor formationand MSC differentiation into bone, adipose, and cartilage after3 and 6 weeks of cell transplantation (data not shown).

Figure 3. Differentiation capacity of Bcl-2-MSCs. MSCs transfectedwith Bcl-2 were cultured in adipogenic, chondrogenic, and myogenicmedium for up to 2 months. (A): Adipogenic differentiation. Immuno-staining with fatty acid binding protein-4 (brown). (B–D): Chondro-genic differentiation. Immunostaining for aggrecan ([B, D], red). Nucleiwere counterstained with 4,6-diamidino-2-phenylindole ([C, D], blue).(E–G): Cardiomyocyte-like differentiation. Immunostaining with anti-Nkx2.5 antibody ([E, G], green). Nuclei were counterstained withpropidium iodide ([F, G], red).

2122 Bcl-2 Engineered MSCs

Capillary DensityCapillary density at the border zone of the acute myocardialinfarction (AMI) was determined based on vWF immunostain-ing 3 weeks after cell transplantation. Representative images areshown in Figure 7A. There was a significant increase in capil-lary density in vector-MSC (20.5 � 1.9 vessels per HPF) andBcl-2-MSC (23.6 � 1.3 vessels per HPF) groups when com-pared with medium-treated hearts with AMI (14.4 � 2.1 vesselsper HPF) (n � 8, p � .001) (Fig. 7B), and no obvious angio-genesis was detected in the medium group. In addition, thecapillary density was 15% higher in Bcl-2-MSC treated heartsthan in vector-MSC treated hearts with significant difference(n � 8, p � .002).

Infarct Size and Cardiac FunctionLeft anterior descending (LAD) ligation consistently resulted intransmural myocardial infarction, exhibiting typical histologicchanges including thinning of the left ventricular free wall andextensive collagen deposition 3 weeks after myocardial infarc-tion. Representative left ventricular sections 3 weeks followingLAD ligation with medium or MSC or Bcl-2-MSC injection areshown in Figure 7C. Both vector-MSC treated and Bcl-2-MSCtreated MI� animals showed smaller infarction size (34.2% �6.0% in vector-MSC group; 28.3% � 5.8% in Bcl-2-MSCgroup) compared with the medium treated animals (47.8% �6.3%), and the difference was statistically significant (n � 8,p � .001 vector-MSC group vs. medium group; n � 8, p � .001Bcl-2-MSC vs. medium group). Furthermore, the collagen con-tent was 17% lower in Bcl-2-MSC treated hearts than in vector-MSC treated hearts (34.2% � 6.0% of vector-MSC group vs.28.3% � 5.8% of Bcl-2-MSC group) (n � 8, p � .029) (Fig.7D).

Hemodynamic parameters of left ventricular function (sup-plemental online Table 2) demonstrated that max dP/dt and mindP/dt were improved significantly more in the Bcl-2-MSCgroup than in the vector-MSC group (Fig. 7E) (n � 6, p � .001).Compared with the medium group, both Bcl-2-MSCs and thevector-MSC group had significant improvements on hemody-namic parameters (ESV, EF, max dP/dt, and min dP/dt). Fur-thermore, there were significant improvements in the Bcl-2-MSC group on EDV, ESPVR, and Emax compared with themedium group (Bcl-2-MSC group: 239.2 � 35.9 �l of EDV,0.58 � 0.14 mmHg/�l slope ESPVR, 1.50 � 0.34 mmHg/�lEmax vs. medium group: 292.1 � 39.0 �l of EDV, 0.36 � 0.16mmHg/�l slope ESPVR, 0.93 � 0.39 mmHg/�l Emax) (n � 6,p � .05). However, there was only a positive trend towardimprovement on EDV, ESPVR, and Emax between vector-MSCgroup and medium group, and the difference did not reachstatistical significance. No significant difference was observedon the values of V0 between the Bcl-2-MSC group (83.4 � 14.1�l) and the vector-MSC group (79.4 � 13.0 �l) or between theBcl-2-MSC group and the medium group (87.2 � 15.8 �l).

DISCUSSION

MSCs derived from adult bone marrow have been proposed asa promising cell source for the regeneration and restoration ofinfarcted heart function. However, poor survival of implantedcells hampered the therapeutic efficacy. In this study, the keyroles Bcl-2 played in protecting grafted MSC survival underhypoxia-ischemia were identified in vitro and in vivo. Weconfirmed that Bcl-2 genetically modified MSCs retained theirdifferentiation capacity and could give rise to different lineagesin vitro. Overexpression of Bcl-2 reduces MSC death and apo-

Figure 4. Antiapoptotic effect and upregulation of VEGF secretion of Bcl-2-MSCs under hypoxic conditions. (A): Representative photomicrographsof TUNEL-positive (left panels) and total (right panels) cells after 24 hours of hypoxic treatment. (B): Representative photomicrographs of apoptoticcells (arrows) with chromatin condensation by Hoechst 33342 staining after 24 hours of hypoxic treatment. (C): VEGF secretion by Bcl-2-MSCsunder hypoxic conditions. After 24 hours of incubation, conditioned medium from normoxia (solid bar) and hypoxia (empty bar) treated cells (n �6) was subjected to VEGF enzyme-linked immunosorbent assay (ELISA) assay. VEGF concentration values are mean � SD (� p � .001). ELISAdata are representative of three independent experiments. Abbreviations: PI, propidium iodide; TUNEL, terminal deoxynucleotidyl transferase-mediated dUTP end labeling; VEGF, vascular endothelial growth factor.

2123Li, Ma, Ong et al.

www.StemCells.com

ptosis under hypoxic conditions. In vivo transplantation of Bcl-2-MSCs enhances the survival rate of MSCs in the LAD ligationmodel. It could attenuate postinfarction LV remodelling and restoreLV function, which may have attributed to the enhanced antiapop-totic properties of the modified MSCs. These observations suggestthat genetic modification of MSCs with Bcl-2 could be of signif-icant value in improving the efficacy of stem cell therapy followinga broad range of cardiac diseases. Our results agree with the earlierfindings that transgenic mice overexpressing Bcl-2 in the heartreduced infarct size and improved recovery of cardiac functionafter ischemia/reperfusion injury [38, 39, 44].

There are several mechanisms contributing to the high levelof stem cell death within 4 days after implantation into ischemichearts [24, 28]. These include host inflammatory response, lossof survival signal from matrix attachments or cell-cell contact,delivery of oxygen and substrates via diffusion [28], variousproapoptotic or cytotoxic factors in the ischemic myocardium[45], and ischemic/reperfusion damage to the implanted stemcells incurred from repeated bouts of ischemia [46]. Further-more, MSCs are extremely sensitive to the hypoxic and inflam-matory environment in ischemic hearts [24]. Our present studiesreveal that genetic modification of MSCs with Bcl-2 effectively

protects transplanted MSCs against ischemia and increases cellsurvival after implantation. Our investigation, consistent withprevious findings from Mangi et al. [18] and Tang et al. [46],showed that the therapeutic efficacy of MSCs was closelyrelated to the in situ survival of cells implanted in the hostileenvironment of hypoxia, inflammation, and scarring from myo-cardial infarction. Hence, the essential survival factors forMSCs, such as cytokines, chemokines, integrins, and otheradhesion molecules, need to be further addressed quantitatively.

It was recently reported that persistent overexpression ofVEGF in nonischemic myocardium may cause the formationof vascular tumors [47]. The present study for the first timedemonstrated the high-level expression of VEGF from Bcl-2-MSCs in response to the hypoxic conditions. Therefore,transplantation of Bcl-2-MSCs could provide adequate mag-nitude and duration of VEGF expression in the ischemicmyocardium without formation of vascular tumors. The high-level expression of VEGF from Bcl-2-MSCs transplanted to

Figure 5. Engraftment of Bcl-2-MSCs in ischemic myocardium. (A,B): Representative images from immunostaining of grafted MSCs in theinfarcted rat myocardium 4 days following MSC injection. BrdU-la-beled MSCs (brown) were clearly identified. (C–E): Immunofluores-cence staining of BrdU-labeled cells in the Bcl-2-MSC group 3 weeksafter cell treatment with high magnification. (C): BrdU (red). (D): DAPI(blue). (E): Merged image (�1,000 in all panels). Arrowheads indicatethe cell engraftments. (F): Quantitative assessment of engrafted MSCsat 4 days, 3 weeks, and 6 weeks (� p � .05; �� p � .01). Abbreviations:BrdU, bromodeoxyuridine; DAPI, 4,6-diamidino-2-phenylindole.

Figure 6. Bcl-2-MSCs in ischemic myocardium. (A, B): Doubleimmunofluorescence staining revealed BrdU-labeled Bcl-2-MSCs(green) incorporated into the endothelial lining of blood vessels (red)near the infarct border zone 3 weeks after transplantation. (C–E): Threeweeks after transplantation, sections near the infarct zone were double-stained for BrdU (green) and cardiac marker Troponin T (red). BrdU-positive cell colocalized with cardiac Troponin T. (F–H): Three-dimen-sional reconstruction of the tissue section using confocal microscopyillustrates that BrdU-positive nucleus (green) is within a cell body thatexpresses cardiac Troponin T. Cardiac-like cells were stained for Tro-ponin T (primary antibody: monoclonal anti-Troponin T antibody, car-diac isoform, clone CT-3, Lab Vision; secondary antibody: Alexa 568);BrdU-positive cells were identified (primary antibody: monoclonalmouse anti-BrdU antibody; secondary antibody: Alexa 488). Abbrevi-ations: BrdU, bromodeoxyuridine; cTnT, cardiac Troponin T.

2124 Bcl-2 Engineered MSCs

the ischemia-damaged myocardium would provide cardiopro-tective effects and induce functional collateral vessels, whichcontribute to the salvaging of ischemic myocardium anddecrease the infarct area.

However, it is still not clear whether intramyocardiallytransplanted MSCs function as newly differentiated cardiomy-ocytes and endothelial cells or serve as the paracrine cells bysecretion of cardioprotective proteins. Our study and those ofother groups [46, 48, 49] confirmed that grafted implanted

MSCs can secrete important survival factors, including VEGF,insulin-like growth factor, hepatocyte growth factor, stromalcell-derived factor-1, and basic fibroblast growth factor. Con-sidering the low frequency of new cardiomyocytes and endo-thelial cells [46, 50], this paracrine effect could be even moreimportant for the functional recovery of the infarcted heart thanthe transdifferentiation potential of stem cells. Therefore, aquantitative assay of the paracrine cytoprotective factors fromstem cells cocultured with cardiomyocytes subjected to hypoxia

Figure 7. Functional effects of Bcl-2 modified MSCs. (A, B): Bcl-2-MSCs enhanced angiogenesis in infarcted heart. (A): Representativephotomicrographs of the infarct zone obtained after immunostaining for von Willebrand factor (vWF) antibody. (B): Quantitative capillary densitydata based on vWF immunostaining. Values are mean � SD (� p � .002, �� p � .001). (C): Representative picrosirius red staining of transversesections through rat hearts 3 weeks following left anterior descending ligation with medium, vector-MSC, or Bcl-2-MSC injection. Thinning of theleft ventricular free wall and extensive collagen deposition (red) in scar tissue was noted in medium and vector-MSCs. (D): The scar size wassignificantly reduced in both the vector-MSC (� p � .001 vs. medium group) and Bcl-2-MSC groups (� p � .001 vs. medium group). The scar sizewas 17% smaller in the Bcl-2-MSC group than in the vector-MSC group (�� p � .029). (E): Effects of MSC transplantation on left ventricularmaximum dP/dt and minimum dP/dt. Values are means � SD (� p � .001). Abbreviation: HPF, high-power field.

2125Li, Ma, Ong et al.

www.StemCells.com

deserves significant attention. The cocktail of paracrine cyto-protective factors may bring fresh therapeutic options to anumber of diseases, including heart failure.

Our current work, together with many other studies, dem-onstrated that bone marrow MSC implantation could inducetherapeutic angiogenesis in infarcted heart [11, 51–53] or isch-emic tissue [54]. Our data suggested that the viability of trans-planted cells is one of the important factors for preservation ofmyocardial function, and therapeutic angiogenesis is not linearlycorrelated with the survival rate of the transplanted cells. Theunderlying mechanism of angiogenesis is very complicated. Forinstance, MSCs are able to differentiate into vascular endothelialcells in the ischemic myocardium and generate capillary-likestructures [11, 51, 53]. Meanwhile, MSCs enhance angiogenesispartly by increasing endogenous levels of vascular endothelialgrowth factor and vascular endothelial growth factor type 2receptor [54]. Our evidence supports the theory that the para-crine effect of MSCs might be one of the major reasons fortherapeutic angiogenesis. Using the genetic modification ap-proach may further enhance the cytoprotective effect and restoreMSC angiogenesis effect.

In the present study, a significant increase in nucleosomalDNA fragmentation and DNA condensation by TUNEL andHoechst 33342 nuclear staining was observed in MSCs subjectedto hypoxic conditions compared with the normoxic control cells, afinding concordant with previous data verifying that hypoxia pro-vokes apoptosis of MSCs [18]. It is known that hypoxia can induceBcl-2 downregulation through nuclear factor-�B (NF-�B) in endo-thelial cells [55] and myocytes [56]. We provide the first directevidence that Bcl-2 upregulation from exogenous delivery activatesa survival pathway that is sufficient to suppress hypoxia-inducedapoptosis of MSCs. The underlying mechanism by which Bcl-2protects MSCs under hypoxia is still unknown. We speculate thatNF-�B might play a pivotal role in the cytoprotective effect ofBcl-2 on MSCs under hypoxic conditions. Further studies need tobe conducted in order to support the hypothesis.

Our hemodynamic data are consistent with previous reportsthat demonstrated cardiac functional improvement after MSCtransplantation [57, 58]. We found that Bcl-2-MSC transplan-tation further reduced EDV and significantly enhanced the sys-tolic functional recovery determined by slope ESPVR andEmax. These beneficial effects of Bcl-2-MSCs may, in part,attribute to VEGF secretion in response to the hypoxic environ-ment, angiogenesis, and the improved cellular survival afterpretreatment with the antiapoptotic Bcl-2 gene.

In this investigation, MSCs were genetically modified withBcl-2 by a cationic polymer based vector with several uniqueadvantages, such as easy surface modification with well defined

structure and chemical properties [59], relatively low toxicity,high capability for carrying large therapeutic genes, and lack ofimmune responses. More importantly, transient overexpressionof Bcl-2 modified MSCs that have benefited from the nonviralgene delivery vector may minimize the potential risk of tumor-igenesis while providing safe and sufficient protection for trans-planted MSCs from short-term ischemic damage, which plays acritical role in transplanted stem cell death [24]. Further work onthe long-term persistence, function, and phenotype of Bcl-2 mod-ified MSCs treated with nonviral vectors needs to be done to safelyrealize the full potential of MSCs for myocardial regeneration.

CONCLUSION

In summary, we have confirmed that genetic modification withthe antiapoptotic Bcl-2 gene resulted in high-level VEGF ex-pression in response to the hypoxic conditions. It enhanced thesurvival of engrafted MSCs in the heart after acute myocardialinfarction. The transplantation of Bcl-2 modified MSCs amelio-rated LV remodeling and improved LV function. Geneticallyengineering cells by Bcl-2 using a nonviral vector could be aneffective strategy for increasing cell survival after cell trans-plantation while minimizing the potential risk of tumorigenesis.Transplantation of gene-engineered MSCs may provide a noveland effective approach in the treatment of acute myocardialinfarction.

ACKNOWLEDGMENTS

This work was supported by Helmholtz Gemeinschaft, Meck-lenburg-Vorpommern (Nachwuchsgruppe Regenerative Medi-zin Regulation der Stammzellmigration Forderkennzeicher0402710), BMBF, Miltenyi Biotec, START-MSC project(project 6: Kardiovaskulare Differenzierung und Applikationdefinierter mesenchymaler Stammzellpopulationen), and Stein-beis Transfer Zentrum fuer Herz-Kreislaufforschung, Rostock,Germany. Authors thank Miriam Nickel, Margit Fritsche, andDaniela Kurzhals for technical assistance.

DISCLOSURE OF POTENTIAL CONFLICTS

OF INTEREST

The authors indicate no potential conflicts of interest.

REFERENCES

1 Leor J, Patterson M, Quinones MJ et al. Transplantation of fetalmyocardial tissue into the infarcted myocardium of rat. A potentialmethod for repair of infarcted myocardium? Circulation 1996;94(suppl 9):332–336.

2 Min JY, Yang Y, Sullivan MF et al. Long-term improvement of cardiacfunction in rats after infarction by transplantation of embryonic stemcells. J Thorac Cardiovasc Surg 2003;125:361–369.

3 Murry CE, Wiseman RW, Schwartz SM et al. Skeletal myoblasttransplantation for repair of myocardial necrosis. J Clin Invest 1996;98:2512–2523.

4 Taylor DA, Atkins BZ, Hungspreugs P et al. Regenerating functionalmyocardium: Improved performance after skeletal myoblast transplan-tation. Nat Med 1998;4:929–933.

5 Jia ZQ, Mickle DA, Weisel RD et al. Transplanted cardiomyocytessurvive in scar tissue and improve heart function. Transplant Proc 1997;29:2093–2094.

6 Li RK, Jia ZQ, Weisel RD et al. Cardiomyocyte transplantation improvesheart function. Ann Thorac Surg 1996;62:654–660.

7 Li RK, Jia ZQ, Weisel RD et al. Smooth muscle cell transplantation intomyocardial scar tissue improves heart function. J Mol Cell Cardiol1999;31:513–522.

8 Yoo KJ, Li RK, Weisel RD et al. Autologous smooth muscle celltransplantation improved heart function in dilated cardiomyopathy. AnnThorac Surg 2000;70:859–865.

9 Leri A, Kajstura J, Anversa P. Cardiac stem cells and mechanisms ofmyocardial regeneration. Physiol Rev 2005;85:1373–1416.

10 Tomita S, Li RK, Weisel RD et al. Autologous transplantation of bonemarrow cells improves damaged heart function. Circulation 1999;100(suppl 19):247–256.

11 Kocher AA, Schuster MD, Szabolcs MJ et al. Neovascularization ofischemic myocardium by human bone-marrow-derived angioblasts pre-vents cardiomyocyte apoptosis, reduces remodeling and improves car-diac function. Nat Med 2001;7:430–436.

12 Li RK, Mickle DA, Weisel RD et al. Optimal time for cardiomyocytetransplantation to maximize myocardial function after left ventricularinjury. Ann Thorac Surg 2001;72:1957–1963.

2126 Bcl-2 Engineered MSCs

13 Stamm C, Westphal B, Kleine HD et al. Autologous bone-marrowstem-cell transplantation for myocardial regeneration. Lancet 2003;361:45–46.

14 Menasche P, Hagege AA, Vilquin JT et al. Autologous skeletal myoblasttransplantation for severe postinfarction left ventricular dysfunction.J Am Coll Cardiol 2003;41:1078–1083.

15 Assmus B, Schachinger V, Teupe C et al. Transplantation of progenitorcells and regeneration enhancement in acute myocardial infarction (TOP-CARE-AMI). Circulation 2002;106:3009–3017.

16 Ma N, Ladilov Y, Moebius JM et al. Intramyocardial delivery of humanCD133� cells in a SCID mouse cryoinjury model: Bone marrow vs. cordblood-derived cells. Cardiovasc Res 2006;71:158–169.

17 Pittenger MF, Martin BJ. Mesenchymal stem cells and their potential ascardiac therapeutics. Circ Res 2004;95:9–20.

18 Mangi AA, Noiseux N, Kong D et al. Mesenchymal stem cells modifiedwith Akt prevent remodeling and restore performance of infarcted hearts.Nat Med 2003;9:1195–1201.

19 Pittenger MF, Mackay AM, Beck SC et al. Multilineage potential ofadult human mesenchymal stem cells. Science 1999;284:143–147.

20 Pereira RF, Halford KW, O’Hara MD et al. Cultured adherent cells frommarrow can serve as long-lasting precursor cells for bone, cartilage, andlung in irradiated mice. Proc Natl Acad Sci U S A 1995;92:4857–4861.

21 Chen J, Li Y, Wang L et al. Therapeutic benefit of intracerebral trans-plantation of bone marrow stromal cells after cerebral ischemia in rats.J Neurol Sci 2001;189:49–57.

22 Ferrari G, Cusella-De Angelis G, Coletta M et al. Muscle regenerationby bone marrow-derived myogenic progenitors. Science 1998;279:1528 –1530.

23 Makino S, Fukuda K, Miyoshi S et al. Cardiomyocytes can be generatedfrom marrow stromal cells in vitro. J Clin Invest 1999;103:697–705.

24 Toma C, Pittenger MF, Cahill KS et al. Human mesenchymal stem cellsdifferentiate to a cardiomyocyte phenotype in the adult murine heart.Circulation 2002;105:93–98.

25 Le Blanc K, Tammik C, Rosendahl K et al. HLA expression andimmunologic properties of differentiated and undifferentiated mesenchy-mal stem cells. Exp Hematol 2003;31:890–896.

26 Djouad F, Plence P, Bony C et al. Immunosuppressive effect of mesen-chymal stem cells favors tumor growth in allogeneic animals. Blood2003;102:3837–3844.

27 Wollert KC, Meyer GP, Lotz J et al. Intracoronary autologous bone-marrow cell transfer after myocardial infarction: The BOOST random-ised controlled clinical trial. Lancet 2004;364:141–148.

28 Zhang M, Methot D, Poppa V et al. Cardiomyocyte grafting for cardiacrepair: Graft cell death and anti-death strategies. J Mol Cell Cardiol2001;33:907–921.

29 Gross A, McDonnell JM, Korsmeyer SJ. BCL-2 family members and themitochondria in apoptosis. Genes Dev 1999;13:1899–1911.

30 Reed JC. Bcl-2 family proteins. Oncogene 1998;17:3225–3236.31 Misao J, Hayakawa Y, Ohno M et al. Expression of bcl-2 protein, an

inhibitor of apoptosis, and Bax, an accelerator of apoptosis, in ventricularmyocytes of human hearts with myocardial infarction. Circulation 1996;94:1506–1512.

32 Feuerstein G, Ruffolo RR Jr., Yue TL. Apoptosis and congestive heartfailure. Trends Cardiovasc Med 1997;7:249–255.

33 Liu L, Azhar G, Gao W et al. Bcl-2 and Bax expression in adult rat heartsafter coronary occlusion: Age-associated differences. Am J Physiol1998;275:R315–R322.

34 Kirshenbaum LA, de Moissac D. The bcl-2 gene product preventsprogrammed cell death of ventricular myocytes. Circulation 1997;96:1580–1585.

35 Murphy AN, Bredesen DE, Cortopassi G et al. Bcl-2 potentiates themaximal calcium uptake capacity of neural cell mitochondria. Proc NatlAcad Sci U S A 1996;93:9893–9898.

36 Nunez G, London L, Hockenbery D et al. Deregulated Bcl-2 geneexpression selectively prolongs survival of growth factor-deprived he-mopoietic cell lines. J Immunol 1990;144:3602–3610.

37 Veis DJ, Sorenson CM, Shutter JR et al. Bcl-2-deficient mice demon-strate fulminant lymphoid apoptosis, polycystic kidneys, and hypopig-mented hair. Cell 1993;75:229–240.

38 Imahashi K, Schneider MD, Steenbergen C et al. Transgenic expressionof Bcl-2 modulates energy metabolism, prevents cytosolic acidificationduring ischemia, and reduces ischemia/reperfusion injury. Circ Res 2004;95:734–741.

39 Chen Z, Chua CC, Ho YS et al. Overexpression of Bcl-2 attenuatesapoptosis and protects against myocardial I/R injury in transgenic mice.Am J Physiol Heart Circ Physiol 2001;280:H2313–H2320.

40 Palojoki E, Saraste A, Eriksson A et al. Cardiomyocyte apoptosis andventricular remodeling after myocardial infarction in rats. Am J PhysiolHeart Circ Physiol 2001;280:H2726–H2731.

41 Weidner N, Semple JP, Welch WR et al. Tumor angiogenesis andmetastasis—correlation in invasive breast carcinoma. N Engl J Med1991;324:1–8.

42 Dexter TM, Allen TD, Lajtha LG. Conditions controlling the prolifera-tion of haemopoietic stem cells in vitro. J Cell Physiol 1977;91:335–344.

43 Barry F, Boynton RE, Liu B et al. Chondrogenic differentiation ofmesenchymal stem cells from bone marrow: Differentiation-dependentgene expression of matrix components. Exp Cell Res 2001;268:189–200.

44 Brocheriou V, Hagege AA, Oubenaissa A et al. Cardiac functionalimprovement by a human Bcl-2 transgene in a mouse model of ischemia/reperfusion injury. J Gene Med 2000;2:326–333.

45 Geng YJ. Molecular mechanisms for cardiovascular stem cell apoptosisand growth in the hearts with atherosclerotic coronary disease andischemic heart failure. Ann N Y Acad Sci 2003;1010:687–697.

46 Tang YL, Tang Y, Zhang YC et al. Improved graft mesenchymal stemcell survival in ischemic heart with a hypoxia-regulated heme oxygen-ase-1 vector. J Am Coll Cardiol 2005;46:1339–1350.

47 Lee RJ, Springer ML, Blanco-Bose WE et al. VEGF gene delivery tomyocardium: Deleterious effects of unregulated expression. Circulation2000;102:898–901.

48 Yoon YS, Wecker A, Heyd L et al. Clonally expanded novel multipotentstem cells from human bone marrow regenerate myocardium after myo-cardial infarction. J Clin Invest 2005;115:326–338.

49 Tang YL, Zhao Q, Qin X et al. Paracrine action enhances the effects ofautologous mesenchymal stem cell transplantation on vascular regener-ation in rat model of myocardial infarction. Ann Thorac Surg 2005;80:229–236.

50 Gnecchi M, He H, Liang OD et al. Paracrine action accounts for markedprotection of ischemic heart by Akt-modified mesenchymal stem cells.Nat Med 2005;11:367–368.

51 Uemura R, Xu M, Ahmad N et al. Bone marrow stem cells prevent leftventricular remodeling of ischemic heart through paracrine signaling.Circ Res 2006;98:1414–1421.

52 Kawamoto A, Gwon HC, Iwaguro H et al. Therapeutic potential of exvivo expanded endothelial progenitor cells for myocardial ischemia.Circulation 2001;103:634–637.

53 Nagaya N, Fujii T, Iwase T et al. Intravenous administration of mesen-chymal stem cells improves cardiac function in rats with acute myocar-dial infarction through angiogenesis and myogenesis. Am J Physiol HeartCirc Physiol 2004;287:H2670–H2676.

54 Chen J, Zhang ZG, Li Y et al. Intravenous administration of human bonemarrow stromal cells induces angiogenesis in the ischemic boundaryzone after stroke in rats. Circ Res 2003;92:692–699.

55 Matsushita H, Morishita R, Nata T et al. Hypoxia-induced endothelialapoptosis through nuclear factor-kappaB (NF-kappaB)-mediated bcl-2suppression: In vivo evidence of the importance of NF-kappaB in endo-thelial cell regulation. Circ Res 2000;86:974–981.

56 Regula KM, Baetz D, Kirshenbaum LA. Nuclear factor-kappaB re-presses hypoxia-induced mitochondrial defects and cell death of ventric-ular myocytes. Circulation 2004;110:3795–3802.

57 Berry MF, Engler AJ, Woo YJ et al. Mesenchymal stem cell injectionafter myocardial infarction improves myocardial compliance. Am JPhysiol Heart Circ Physiol 2006;290:H2196–H2203.

58 Amado LC, Saliaris AP, Schuleri KH et al. Cardiac repair with intramyo-cardial injection of allogeneic mesenchymal stem cells after myocardialinfarction. Proc Natl Acad Sci U S A 2005;102:11474–11479.

59 Ma N, Wu SS, Ma YX et al. Nerve growth factor receptor-mediated genetransfer. Mol Ther 2004;9:270–281.

See www.StemCells.com for supplemental material available online.

2127Li, Ma, Ong et al.

www.StemCells.com