Xenotransplantation of Long-Term-Cultured Swine Bone Marrow-Derived Mesenchymal Stem Cells


  • Yasuhiro Nakamura,

    1. Cardiovascular Division, Departments of Medicine, University of Minnesota Medical School, Minneapolis, Minnesota, USA
    Search for more papers by this author
  • Xiaohong Wang,

    1. Cardiovascular Division, Departments of Medicine, University of Minnesota Medical School, Minneapolis, Minnesota, USA
    Search for more papers by this author
  • Chengsu Xu,

    1. Cardiovascular Division, Departments of Medicine, University of Minnesota Medical School, Minneapolis, Minnesota, USA
    Search for more papers by this author
  • Atsushi Asakura,

    1. Cardiovascular Division, Departments of Medicine, University of Minnesota Medical School, Minneapolis, Minnesota, USA
    Search for more papers by this author
  • Minoru Yoshiyama,

    1. Department of Internal Medicine and Cardiology, Osaka City University Medical School, Osaka, Japan
    Search for more papers by this author
  • Arthur H.L. From,

    1. Cardiovascular Division, Departments of Medicine, University of Minnesota Medical School, Minneapolis, Minnesota, USA
    Search for more papers by this author
  • Jianyi Zhang M.D., Ph.D.

    Corresponding author
    1. Cardiovascular Division, Departments of Medicine, University of Minnesota Medical School, Minneapolis, Minnesota, USA
    • University of Minnesota Academic Health Center, Mayo Mail Code 508, 420 Delaware Street SE, Minneapolis, Minnesota 55455, USA. Telephone: 612-624-8970; Fax: 612-626-4411
    Search for more papers by this author


Swine-derived MSCs were efficiently isolated and extensively expanded using a low fetal serum content growth medium to which selected growth factors were added. After ≥96 cell population doublings (PDs), MSCs were devoid of cytogenetic abnormalities. In vitro chondrogenic and osteogenic differentiation capacity was preserved after 80 PDs. To test therapeutic efficacy, 1 × 106 80-PD MSCs were injected directly into the peri-infarct zone of hearts of immunodeficient (non-obese diabetic/severe combined immunodeficient) mice at the time of acute myocardial infarction. Engrafted MSCs survived in the infarcted hearts for at least 4 weeks. Echocardiography at 2 and 4 weeks postinfarction revealed a significant preservation of the left ventricular ejection fractions of infarct hearts receiving MSCs compared with infarct hearts receiving saline. Peri-infarct zone capillarity was better preserved in MSC-treated hearts than other infarct groups of hearts, but infarct size was comparable in all groups. Only rare engrafted MSCs expressed cardiac-specific or endothelial cell-specific markers. Hence, 80-PD MSCs retained the capacity to promote functional improvement in the infarcted heart despite minimal differentiation of MSCs into cardiomyocytes or endothelial cells. These data suggest that the beneficial effects of MSC transplantation most likely result from the trophic effects of MSC-released substances on native cardiac and vascular cells. The capacity to massively expand MSC lines without loss of therapeutic efficacy may prove to be useful in the clinical setting where “off the shelf” MSCs may be required for interventions in patients with acute coronary syndromes.


Optimization of methods for isolation of MSCs from bone marrow is important because it has been reported that conventionally isolated MSCs are able to undergo only a moderate number of cell population doublings (PDs) in vitro [1]. It is possible, although untested to our knowledge, that multipassaging may decrease in vivo therapeutic effects of MSCs. It has been reported that successful isolation of marrow-derived MSCs appears to be critically dependent upon the specific serum products used in the process [1]. Because clinical trials of cell transplantation for myocardial repair are ongoing, highly reproducible methods of isolating and extensively expanding MSCs will be a necessity. It is also critical that populations of highly expanded MSC lines retain therapeutic efficacy. Recently, we reported that a culture medium containing a low serum concentration (2% fetal bovine serum [FBS]) in combination with additional multiple growth factors (platelet-derived growth factor [PDGF] and endothelial growth factor [EGF]) resulted in successful MSC isolation from all marrow aspirates studied [2].

In the current study, using the medium described above, MSCs were successfully isolated from all swine bone marrow aspirates. Five isolated MSC lines did not show evidence of cytogenetic abnormalities after 96 PDs. We hypothesized that extensively multipassaged MSCs would retain therapeutic potency in an acute myocardial infarction (AMI) model. Therefore, 80-PD MSCs were injected into the peri-infarct regions of immunodeficient (non-obese diabetic/severe combined immunodeficient [NOD/SCID]) mouse hearts at the time of AMI. A fraction of transplanted MSCs remained engrafted in the hearts for at least 4 weeks post-transplantation, and left ventricle (LV) contractile function was significantly better preserved in MSC-injected hearts at 2 and 4 weeks post-transplantation than in AMI hearts injected with saline. Few engrafted MSCs developed cardiomyocyte markers or endothelial cell markers, suggesting that the major benefits of transplantation did not derive from structural contributions of engrafted MSCs. The data support the concept that it is possible to extensively expand bone marrow-derived MSC lines and retain cytogenetic normality and therapeutic efficacy.

Materials and Methods

All experimental procedures were approved by the University of Minnesota Animal Resources Committee. The investigation conformed to the NIH Guide for the Care and Use of Laboratory Animals [3].

Isolation of Swine MSCs

MSCs from bone marrow were isolated by gradient density centrifugation as described in detail previously [1, 2, 4]. In brief, bone marrow was aspirated from the sternum of healthy young Yorkshire pigs at 45 days of age into a syringe containing 6,000 U of heparin and diluted with Dulbecco's phosphate-buffered saline (PBS) in a ratio of 1 to 1. The marrow sample was carefully layered onto Ficoll-Paque-1077 (Sigma-Aldrich, St. Louis, http://www.sigmaaldrich.com) in a 50-ml conical tube and centrifuged at 400g for 30 minutes at room temperature. The mononuclear cells were collected from the interface, washed with 2–3 volumes of Dulbecco's PBS, and collected by centrifugation at 1,000 rpm. The cells were resuspended and seeded at a density of 200,000 cells per cm2 in a T-75 flask coated with 10 ng/ml fibronectin and cultured in medium consisting of 60% low-glucose Dulbecco's modified Eagle's medium (DMEM) (Gibco-BRL, Gaithersburg, MD, http://www.gibcobrl.com), 40% MCDB-201 (Sigma-Aldrich), 1× insulin-transferrin-selenium, 1× linoleic acid-bovine serum albumin, 0.05 μM dexamethasone (Sigma-Aldrich), 0.1 mM ascorbic acid 2-phosphate, 2% fetal calf serum, 10 ng/ml PDGF, 10 ng/ml EGF, 100 U/ml penicillin, and 100 U/ml streptomycin. This medium was used for all the cell cultures except for in vitro differentiation experiments (described below). Within 3 days after plating, nonadherent cells were removed by replacing the medium. The attached cells grew and developed colonies in approximately 5–7 days. At approximately 10 days, the primary cultures reached nearly 90% of confluence, and then the cells were subcultured by trypsinizing. The first-passaged cells were plated into a 96-well plate at density of 1,000 cells per well and cultured in the same medium. The cultures were then scaled up to the six-well plate and thereafter to T-25, T-75, and T-150 flasks at a density of 2,000 cells per cm2. The cells from five marrow aspirations were maintained through 96 population doublings. The cells were stored in liquid nitrogen for cell stocks. Karyotype was checked every 20 PDs. An identical isolation method was used with similar success in 48 pigs in separate research projects for autologous transplantation [2, 4].

Adenoviral Transduction of MSCs

Nuclear LacZ adenoviruses were prepared and titrated by the Gene Transfer Vector Core Laboratory at University of Iowa, in which Escherichia coli β-galactosidase was subcloned into a recombinant Ad5 vector (Rous sarcoma virus [RSV] promoter, AdRSV-lacZ). Swine MSCs at 80 PDs were plated at 5,000 cells per cm2 in modified DMEM with 2% FBS. One day later, the cells were washed with serum-free modified DMEM and infected overnight with AdRSV-LacZ at a multiplicity of infection (MOI) of 1,500. The supernatant was removed next day, and the cells were washed with PBS and then refed with fresh normal MSC culture medium as previously described [2, 4]. The medium was repeatedly changed over 2 days (≥6 changes) to ensure complete removal of viral particles and to allow for the internalization of any particles remaining on the surface. On the day of surgery, part of the adhesive cells were used to perform 5-bromo-4-chloro-3-indolyl-β-d-galactoside (X-gal) staining according to the manufacturer's directions (Invitrogen, Carlsbad, CA, http://www.invitrogen.com) to determine the transduction efficiency, and the rest of the cells were harvested with 0.25% trypsin-EDTA, washed with PBS, and resuspended at a concentration of 2 × 107 cells per milliliter in saline for cell transplantation. We could obtain >95% transduction efficiency of swine MSC with AdRSV-LacZ at an MOI of 1,500.

Karyotype Analysis

Karyotype analysis was performed according to the method described by Liu et al. [2] and Shim et al. [5]. MSCs were seeded at a density of 4,000 cells per cm2 in a T-150 flask and cultured in medium as described above. When 80% confluence was achieved, medium was exchanged with 15 ml of fresh medium containing Demecolcine solution (Sigma-Aldrich) at final concentration of 0.33 μg/ml and incubated at 37°C in 5% CO2, 95% air for 6–8 hours. After Demecolcine incubation, nonadherent mitotic cells were collected and placed in 50-ml conical tubes. The cells were then rinsed with Ca2+-, Mg2+-free Dulbecco's PBS, and the rinsing fluid was collected and added into the same conical tube. The attached cells were trypsinized and then neutralized by adding back-pooled fluid to conical tube, and the cells were then harvested via centrifugation at 1,000 rpm for 10 minutes. The resulting pellet was suspended in 10 ml of prewarmed (37°C) hypotonic solution (0.56% KCl in H2O, wt/vol) and incubated at room temperature for 20 minutes. The cells were then centrifuged at 1,000 rpm for 5 minutes and fixed in 14 ml of Carnoy's fixative (3:1, vol/vol, absolute methanol to glacial acetic acid, freshly made) for 15 minutes. After another wash and centrifugation, the cells were resuspended in 0.5 ml of fixative. The cell suspension was dropped onto microscope slides, and after air drying, the slides were stained with diluted Giemsa staining solution (catalog no. GS-500; Sigma-Aldrich) for 15–30 minutes. The stained slides were rinsed with tap water, air-dried, and observed at a magnification of ×400 with immersion oil (catalog no. 16484; Cargille Laboratories Inc., Cedar Grove, NJ, http://www.cargille.com) Approximately 10 to 20 metaphase spreads from each sample were examined, and karyotypes from each sample were arranged.

In Vitro Chondroblastic and Osteoblastic Differentiation of MSCs

To demonstrate the in vitro pluripotency, MSCs that had undergone 80 PDs were cultured under conditions that have previously been reported [1, 2] to cause transdifferentiation of stem cells into chondroblasts or osteoblasts. To study chondroblast differentiation, MSCs were trypsinized, and aliquots of 25 × 104 cells in 0.5 ml of chemically defined medium (with porcine 10 ng/ml transforming growth factor-β [TGF-β] and 10−7 M dexamethasone) were centrifuged in 15-ml conical tubes at 500g. The pelleted cells were left at the bottom of the tubes and incubated in a CO2 incubator. Within 24 hours after incubation, the sedimented cells formed a spherical mass (an aggregate) at the bottom of the tube that did not adhere to the tube wall. Differentiation to chondroblasts was shown by toluidine blue staining and type II collagen staining [1].

To study osteoblast differentiation, trypsinized cells were plated onto six-well plates at 3 × 103 cells per cm2 in 10% serum DMEM. On the following day, the medium was replaced with fresh medium containing 100 μM dexamethasone (Sigma-Aldrich), 10 mM β-glycerophosphate (Sigma-Aldrich), and 50 μM ascorbic acid-2-phosphate (Sigma-Aldrich), with changes of medium every 3–4 days. On days 4, 8, 12, 16, and 20 of culture, samples were assayed for alkaline phosphatase activity and mineral deposition using histochemical staining with Sigma-Aldrich Kit 85 and the Von Kossa method, respectively [1].

AMI Induction and MSC Transplantation

The mouse model of myocardial infarction was produced as previously described [6]. Briefly, immunodeficient (NOD/SCID) mice were anesthetized with pentobarbital sodium (25 mg/kg, i.p.) and lidocaine hydrochloride (10 mg/kg, i.p.). After tracheal intubation, the mice were ventilated with a small-animal respirator. AMI was produced by permanent ligation of the left anterior descending coronary artery (LAD) at mid-level with a 9-0 nylon surgical suture (during visualization with a dissecting microscope). Mice that survived the LAD ligation were then randomized into three groups; one received MSCs suspended in saline (n = 12), another received saline injection only (n = 11), and a sham-operated group was used as normal controls (n = 6). MSC transplantation was performed within 15 minutes post-LAD ligation, following which the chest was closed. Each mouse received either saline or four injections of MSC suspension (total of 1 × 106 cells per heart in 50 μl of saline) delivered into the peri-infarct regions with a 32-gauge needle. Sham-operated hearts were exposed to an identical surgical protocol, but LAD ligation and intramyocardial injections were not done. Postprocedure, the chest was closed in layers, and the mice were allowed to recover. Groups of animals returned to the laboratory at week 2 or 4 for echocardiographic measurements of LV function. Following the echocardiographic examination, mice were sacrificed with a pentobarbital overdose (i.p.), and their hearts were excised. Whole heart samples were stained with X-gal (Invitrogen) overnight to visualize transplanted cells as described previously [4, 7].


Mice were lightly anesthetized with ketamine HCl (50 mg/kg, i.p.) and xylazine (16.5 mg/kg, i.p.). Echocardiography was performed using a commercially available echocardiographic system equipped with a 15.6-MHz phased-array transducer (SONOS 5500; Phillips Medical Systems, Best, The Netherlands, http://www.medical.philips.com). A two-dimensional short-axis view of the LV was obtained at the level of the papillary muscles. The LV internal dimensions (end-diastolic dimension [LVDd] and end-systolic dimension [LVDs]) were measured by the leading-edge method from at least eight consecutive cardiac cycles. The LV ejection fraction (EF) was calculated as follows: EF = (LVDd2 − LVDs2/LVDd2 × 100 (%). After echocardiography was preformed, the animals were sacrificed with an overdose of barbiturates.

Determination of Infarct Size

Following heart explantation, the left ventricle was excised, cut open from base to apex, and laid on a flat white paper. Infarct size was calculated and expressed as a percentage of LV surface area, as previously described [8]. Color digital images were obtained, and the ratio of LV scar area (whitish-yellow) to total LV surface area was evaluated with an image software analysis system (NIH ImageJ program, http://rsb.info.nih.gov/ij). Because of significant scar thinning, the ratio of scar volume to total LV volume would substantially underestimate the severity of initial LV damage; therefore, a volumetric analysis would be misleading, and we believe that the ratio of scar surface area to total LV surface area better describes the severity of the AMI. However, this method of normalization of infarct size to total epicardial surface area also has limitations and will underestimate relative infarct severity in hearts in which uninfarcted myocardial regions have dilated as a consequence of LV remodeling.

Analysis of Myocardial Capillarity

Hearts from sham-operated, infarcted saline-injected, and infarcted cell-transplanted hearts obtained 2 weeks after AMI were transversely sectioned into 8-μm slices using a cryostat and stained for CD31 (BD Biosciences, San Diego, http://www.bdbiosciences.com) antibodies. The sections were visualized using fluorescence-labeled secondary antibodies (Molecular Probes Inc., Eugene, OR, http://probes.invitrogen.com). Images were taken at a magnification of ×20 using an Olympus microscope (BX51/BX52; Tokyo, http://www.olympus-global.com). The number of capillaries was counted in blind fashion in three fields per each section of the peri-infarct zone, and total of five sections per heart were analyzed (n = 4 for each group). The quality of the computer analysis (ImageJ) was checked against manual counting.

Total Number of Engrafted Cells and Engraftment Rate of Transplanted MSCs

After X-gal staining, hearts were embedded in Tissue-Tek OCT compound (Fisher Scientific International, Hampton, NH, http://www.fisherscientific.com) and frozen in liquid nitrogen-cooled isopentane. Eight-μm-thick frozen tissue sections were transversely sectioned on a cryostat from base to apex of the entire heart. Total cell nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI) (Sigma-Aldrich). Engrafted MSCs were identified by counting the X-gal-positive nuclei in every 10th serial section of the whole heart and then multiplying by 10 to obtain the total number of engrafted MSCs per heart. The engraftment rate of transplanted MSCs was calculated by dividing the total number of engrafted MSCs by 1 × 106 (the number of MSCs injected) and multiplying by 100%.

Antibodies, Immunocytochemistry, and Cytohistochemistry

Immunofluorescence and immunohistochemistry staining of the heart tissues for assessment of possible cardiomyocyte and endothelial cell differentiation has been described previously [9]. Double immunostaining of 8-μm-thick heart sections was performed with rabbit polyclonal anti-myocyte enhancer factor-2 (anti-MEF2) (Santa Cruz Biotechnology Inc., Santa Cruz, CA, http://www.scbt.com) followed by Alexa 488-conjugated anti-rabbit IgG secondary antibodies (Molecular Probes). This was followed by monoclonal mouse anti-cardiac troponin T (anti-cTnT; NeoMarkers, Fremont, CA, http://www.labvision.com/), followed by rhodamine-conjugated anti-mouse IgG secondary antibodies (Chemicon, Temecula, CA, http://www.chemicon.com). In addition, sections were treated with rabbit polyclonal anti-laminin (Sigma-Aldrich) followed by Alexa 594-conjugated anti-rabbit secondary antibodies (Molecular Probes). Subsequently, mouse monoclonal anti-N-cadherin (Zymed Laboratories, South San Francisco, www.invitrogen.com/zymed) and mouse anti-connexin-43 (Chemicon), followed by Alexa 488-conjugated anti-mouse IgG secondary antibodies, were used. Moreover, anti-von Willebrand factor and anti-CD31were also used for immunohistochemistry using the Vectastain ABC kit (Vector Laboratories, Burlingame, CA, http://www.vectorlabs.com).

To determine whether osteoblastic or osteochondral differentiation of MSCs occurred in vivo, X-gal-stained sections of the transplanted hearts underwent histochemical staining for alkaline phosphatase activity and mineral deposition (osteoblastic markers) or toluidine blue and type II collagen (osteochondral markers) as described for the in vitro studies.


All data are expressed as a mean ± SEM. Differences among groups were compared by one-way analysis of variance, with post hoc comparisons performed using the Scheffe test of multiple comparisons. Values of p < .05 were considered significant.


Isolation and Maintenance of MSCs Using Modified Culture Media

Five independently isolated 96-PD MSC lines retained normal phenotype and karyotype. Figure 1demonstrates an example of representative cytogenetic data obtained from one line of MSCs. These results indicate that the modified culture conditions permit MSCs to undergo extensive population doublings of MSCs without the occurrence of cytogenetic abnormalities.

Figure Figure 1..

Representative karyotypic data from a first-passage MSC and a 96-population doubling MSC. There was no change in the karyotypic features of the multipassaged MSCs.

Osteogenic and Chondroblastic Differentiation

The osteogenic and chondroblastic differentiation results for 80-PD MSCs are depicted in Figure 2. Differentiation to chondroblasts was shown by toluidine blue staining and type II collagen staining (Fig. 2A). When cell aggregates were cultured in chemically defined medium containing 0.05 μM dexamethasone without TGF-β, toluidine blue staining revealed no evidence of cartilaginous matrix production.

Figure Figure 2..

Chondroblast and osteogenic differentiation. (A): Typical differentiation to chondroblasts was shown by toluidine blue staining and type II collagen staining when 80-population doubling (PD) MSCs were cultured in chemically defined medium containing 0.05 μM dexamethasone plus transforming growth factor-β (TGF-β). Staining revealed evidence of matrix production. This was not present in MSCs cultured in medium without dexamethasone and TGF-β. (B, C): Osteogenic differentiation of 80-PD MSCs (over a period of 21 days) was demonstrated by an increase in alkaline phosphatase activity (B) (purple) and the accumulation of calcium (C) (black).These changes were accompanied by morphological changes from spindle-shaped to cubodial or polygonal cells.

For osteogenic differentiation, 80-PD MSCs after 8–14 days in osteogenic culture medium, more than 80% of cells were alkaline phosphatase-positive. Calcium accumulation was shown after 1 week and increased over time (Fig. 2B, 2C). These data indicate that these MSCs maintained differentiation potential after extensive passaging.

An important question would be whether some of these MSCs, after being transplanted into heart with AMI, would differentiate into osteoblasts/chondroblasts in the heart. Therefore, the in vitro histochemical methods were applied to myocardial sections harboring engrafted MSCs; no evidence of chondroblastic or osteogenic differentiation was found (data not shown). These data support the concept that specificity of stem cell differentiation is determined by the microenvironment where the engrafted MSCs reside.

MSC Engraftment Rate

Figure 3 illustrates the time course for the engraftment rate of MSCs after transplantation into the infarcted NOD/SCID mouse heart. At week 1, the engraftment rate was 4.4% ± 1.2% of injected MSCs (n = 4). At 2 and 4 weeks, the engraftment rates were decreased to 3.4% ± 1.4% (n = 4) and 2.1% ± 0.9% (n = 4; p < .05 vs. week 1), respectively. A typical H&E-stained pathological section showed engrafted MSCs present 4 weeks after transplantation (Fig. 3B). A decline in engrafted cell numbers over time has been previously reported by our group and others [4, 9, 10]. The precise mechanism of engrafted MSC loss over time remains to be determined.

Figure Figure 3..

Engraftment rate. (A): Engraftment rate of the multipassaged MSCs in the peri-infarct zones of the infarcted non-obese diabetic/severe combined immunodeficient (NOD/SCID) mouse hearts. The y-axis indicates engrafted cell numbers relative to total cell numbers (described in Materials and Methods) at 1, 2, and 4 weeks after 1 × 106 80-population doubling (PD) MSC transplantation. Values are mean ± SEM. ∗, p < .05 vs. week 1. (B): H&E staining of the heart tissues from NOD/SCID mouse 2 weeks after xenotransplantation of adenovirus-labeled swine MSCs. Panel (a) shows that the engrafted cells (5-bromo-4-chloro-3-indolyl-β-d-galactoside-stained nuclei) are distributed in peri-infarct region (magnification, ×100); panel (b) shows the same section at high magnification (magnification, ×400).

LV Contractile Function and Infarct Size

To determine whether transplanted MSCs were capable of enhancing preservation of LV contractile function in post-AMI hearts, we used echocardiography to measure the LV EFs at 2 and 4 weeks after AMI. Representative M mode echocardiograms are shown in Figure 4A. It is readily apparent that contractile performance and LV chamber geometry are better preserved in the infarcted heart with MSC injection than in the infarcted heart with saline injection. Table 1 and Figure 4 B show that at week 2, the EF was significantly higher (p < .01) in the MSC group (42% ± 1%; n = 4) than in saline-injected group (26% ± 1%; n = 5). At week 4, the corresponding values were 42% ± 3% (n = 4) and 26% ± 2% (n = 6), respectively (p < .01 MSC vs. saline group). Although infarct size tended to be slightly smaller in the MSC-treated group, the differences were not statistically significant (Table 1).

Figure Figure 4..

A significant functional improvement after swine MSC transplantation into MI hearts of non-obese diabetic/severe combined immunodeficient (NOD/SCID) mice. (A): Typical M-mode echocardiograms obtained from hearts of NOD/SCID mice 4 weeks after sham operation or MI: (a), a control heart; (b), a heart with postinfarction left ventricle (LV) remodeling without MSC transplantation; and (c), a heart with postinfarction LV remodeling plus 80-population doubling (PD) MSC transplantation. Marked LV dilation and decreased systolic wall motion was present in the untreated acute myocardial infarction (AMI) heart compared with the control heart, and these abnormalities were attenuated in the MSC-transplanted AMI heart. (B): LV ejection fraction of control hearts (n = 6), hearts with AMI without MSCs (n = 5), and AMI hearts with MSC transplantation (n = 5) at 2 and 4 weeks after AMI. Values are mean ± SEM; *, p < .05. Abbreviations: EDD, end-diastolic diameter; MI, myocardial infarction.

Table Table 1.. Echocardiography data and infarct size in NOD/SCID mice
original image

Immunocytochemistry and Histochemistry

To determine whether multipassaged MSCs might differentiate to cardiomyocytes in recipient hearts, we performed immunostaining against several cardiomyocyte markers at 2 weeks after cell transplantation. Sections were examined to identify engrafted nuclear LacZ-stained cells coexpressing cardiomyocyte markers. Multiple consecutive sections were used, and care was taken to avoid the confounding possibility that an undifferentiated small MSC overlaying a larger cardiomyocyte would be identified as a trans-differentiated MSC. Approximately 0.2% (n = 4) of engrafted MSCs coexpressed cTnT in their cytoplasm, and some of them coexpressed the cardiac transcription factor MEF2 simultaneously in their nuclei (Fig. 5A). Moreover, approximately 0.2% (n = 4) of troponin T-stained-positive MSCs also expressed N-cadherin and connexin-43 at their junctions with neighboring cells (Fig. 5B, 5C). Engrafted MSCs coexpressing cardiac-specific proteins were mostly single-nucleated (Fig. 5A–5C). A few of these cells were double-nucleated, with both nuclei cells expressing β-galactosidase. Staining for laminin, a marker for basal lamina of cardiomyocytes, indicated that lacZ-positive cells were connected with lacZ-negative cardiomyocytes at their N-cadherin locales (Fig. 5D). In this study, the number of MSCs that showed apparent trans-differentiation is in line with our previous work and that of others [4, 9, 10]. Last, it should be noted that cell fusion (i.e., between an engrafted MSC and a native cardiomyocyte) may be the origin of some or all of the apparently differentiated cells expressing both the donor-derived cell marker LacZ and proteins characteristic of cardiomyocytes.

Figure Figure 5..

Possible donor-derived cell differentiation in vivo. (A): Two weeks following MSC transplantation, immunofluorescence staining of peri-infarct zone microscopic sections showed MSC-derived cells expressing nuclear LacZ (a) coexpressed MEF2 (c, e, f; green) in their nuclei and also expressed cTnT (d, e, f; red) in their cytoplasm (arrows). DAPI staining indicates nuclei (b; blue). Scale bars = 20 μm. (B): Immunofluorescence staining shows cells with LacZ-stained nuclei (a; arrows) coexpressing troponin T in the cytoplasm (c, e, f; green) and connexin-43 (d, e, f; red) at the junctions with neighboring cardiomyocytes (arrowheads). DAPI staining indicates nuclei (b; blue). Scale bars = 20 μm. (C): Immunofluorescence staining showing lacZ-positive cells (a; arrows) that coexpressed troponin T in the cytoplasm (c, e, f; green) and N-cadherin at the junctions (d, e, f; red) with neighboring cardiomyocytes (arrowheads). DAPI staining indicates nuclei (b; blue). Scale bars = 20 μm. (D): Immunofluorescence staining showing lacZ-staining-positive cells (a) (arrows) that expressed N-cadherin (c, e, f; green) at the junctions with neighboring cardiomyocytes (arrowheads), which was evident by laminin-positive basal lamina (d, e, f; red). DAPI staining indicates nuclei (b; blue). Scale bars = 20 μm. Abbreviations: cTnT, cardiac troponin T; DAPI, 4,6-diamidino-2-phenylindole; MEF2, myocyte enhancer factor-2; nLacZ, nuclear LacZ.

Because the percentage of apparently transdifferentiated MSCs in relation to the total number of cardiomyocytes [11] in the normal mouse left ventricle (that is, ∼3.5 × 106) is only ∼0.003% or less, the therapeutic effects of MSCs most likely had a basis other than their direct structural contributions to LV.

In addition, approximately 0.3% (n = 4) of engrafted MSCs also appeared to differentiate into endothelial cells as characterized by expression of von Willebrand factor and CD31 proteins (endothelium-specific markers) (Fig. 6A); their small numbers also preclude the possibility that they made a significant structural contribution to angiogenesis.

Figure Figure 6..

Increased capillarity and rare endothelial cell differentiation of engrafted MSCs. (A): At 2 weeks, rare MSCs that expressed lacZ in the nucleus also coexpressed vWF (arrow) and CD31 (arrow). Scale bars = 20 μm. (B): Representative immunofluorescence images showed CD31-staining-positive capillaries in a normal heart, a saline-injected acute myocardial infarction (AMI) heart, and an MSC-transplanted AMI heart. (C): Mean number of CD31-stained capillaries in the peri-infarct regions of control hearts, saline-injected AMI hearts, and MSC-transplanted hearts 2 weeks after myocardial infarction. Abbreviations: MI, myocardial infarction; VWF, von Willebrand factor.

To determine whether transplanted swine MSCs can affected angiogenesis via an indirect (i.e., paracrine) mechanism, capillary density was evaluated in normal, saline-injected infarcted hearts and in MSC-transplanted infarcted hearts 2 weeks post-AMI. The capillary density was significantly reduced in the peri-infarct zone of saline-treated AMI hearts. However, capillarity remained normal in the MSC-transplanted group (Fig. 6A–6C). Because, as noted above, very few cells showed concordant X-gal endothelial cell marker staining, the preservation of capillary numbers was not dependent on differentiation of MSCs into endothelial cells.

To summarize, the capillarity data, taken together with the modest number of engrafted MSCs and their low rates of apparent differentiation, suggest that the main beneficial effect of the transplanted cells is not structural but rather derives from trophic effects on stressed native cardiomyocytes and the native vasculature [12, [13]–14].


The major findings of this study are: (a) that modification of the cell culture medium (i.e., reduction of fetal serum concentration to 2%) and the addition of selected growth factors (PDGF and EGF) permits isolation of swine bone marrow-derived MSCs with a very high success rate; (b) that these MSCs can be expanded through at least 96 PDs without suffering cytogenetic damage; (c) that these MSCs can be expanded through at least 80 PDs without loss of the capacity to undergo in vitro osteoblastic and chondroblastic differentiation; and (d) that after 80 PDs, these MSCs had significant therapeutic effects in AMI hearts of immunodeficient mice, evidenced by persistent MSC engraftment and persistent, significant improvement of LV function. There was a very low frequency of differentiation of donor-derived cells into cells bearing cardiomyocyte or endothelial cell markers. The current data refute the view that multipassaged MSCs lose functional competence and therapeutic efficacy in the treatment of AMI.

Isolation of MSCs

The characteristics of bone marrow-derived MSCs were established by Pittenger et al. [1]. Those investigators used 10% FBS in the culture medium and emphasized that the selection of a specific FBS lot number was important for the success of MSC isolation [1]. It is likely that the different serum lots used in their culture media contained various concentrations of growth factors essential for successful MSC isolation. In the present study, the use of a culture medium with a low serum concentration (2% FBS) and added PDGF and EGF (to eliminate potential variability of essential growth factor contents) resulted in routinely successful MSC isolation. Hence, a growth factor-enriched, low-serum culture medium may simplify the successful isolation of MSCs and, as indicated by the aforementioned data, permit extensive expansion of these MSC isolates.

The concept of a growth factor-enriched, low-serum culture medium for the successful isolation of multipotent adult progenitor cells (MAPCs) was established by Jiang et al. [15]. MAPCs are human leukocyte antigen (HLA) class I- and CD44 surface marker-negative, but they do express embryonic stem cell maker Oct-4 [15]. MSCs are HLA class I- and CD44 surface marker-positive and Oct-4-negative [1, 2]. Therefore, MAPCs are distinctly different from the multipassaged MSCs used in the present study. MAPCs are the only known adult stem cell line that has multilineage differentiation potential [15].

We have previously reported in a swine postinfarction model, in which we used first-passaged autologous MSCs for total of 46 transplantations, that MSC isolation was always successful [2]. With appropriate inducers, these MSCs differentiated into osteoblasts, chondrocytes, adipocytes, and other cell types in vitro, and following transplantation, they were effective in attenuating left ventricular remodeling due to acute infarction or pressure overload in our studies; early passaged cells have been reported to be similarly successful in the studies of others [2, 4, 16, [17], [18]–19]. In the present study, using modified culture medium, we have considerably extended previous findings (our own and those of others [2, 4, 19]) by demonstrating that MSCs isolated using the modified culture medium have no apparent karyotypic abnormalities after ≥96 PDs and retain in vitro differentiation capacity and in vivo therapeutic effects after at least 80 PDs.

MSC Engraftment and In Vivo Differentiation Potential of Engrafted MSCs

A number of studies have suggested that cardiomyocytes may be generated from circulating bone marrow cells and transplanted bone marrow-derived cells [16, [17]–18, 20, 21]. However, subsequent studies from Murry et al. [22] and Wagers et al. [23] have suggested that even though functional improvement may be obtained using these approaches, the degree of differentiation to cells with cardiac phenotype may be much lower than described in the initial papers. In addition, a number of studies have shown that apparent differentiation of stem cells may really reflect fusion between cells of bone marrow origin and resident cardiomyocytes [24, 25]. There is also evidence that new cardiomyocytes may be differentiated from resident cardiac stem cells [26]. In the present study, the transplanted MSCs appeared to differentiate to cardiac cells with a very low frequency. However, as pointed out above, the possibility that cell fusion accounts for donor-derived cells with cardiomyocyte markers has not been excluded. More importantly, in the current model, the total number of donor-derived cells showing LacZ staining and costaining for cardiac or endothelial markers was very small. Therefore, it is most likely that the main contribution of engrafted MSCs was via trophic effects on native cardiac cells rather than direct contributions to structural improvement.

Possible Mechanisms of Preservation of LV Contractile Function in MSC-Transplanted Hearts

Echocardiographic evaluation of cardiac function of hearts 2 and 4 weeks after acute myocardial infarction revealed significant preservation of ejection fraction values in hearts that received MSCs compared with the nontransplanted groups. The exact mechanisms underlying the preservation of cardiac function in response to MSC transplantation remain unclear. The engraftment rate is low, and the rates of transdifferentiation to cardiomyocytes and endothelial cells are also quite low. Therefore, the data from the present study refute the view that the transplanted MSCs make significant direct structural contributions to cardiomyocyte regeneration, neovascularization, or cardiac contractile performance, and the fact that infarct size in MSC-treated hearts was not significantly reduced also indicates that cardiomyocyte regeneration from native cardiac stem cells was not significant.

Therefore, it is likely that the preservation of contractile function induced by MSC engraftment is the result of “trophic” (i.e., paracrine) effects mediated by cytokines and/or other agents secreted by engrafted MSCs; we hypothesize that these substances act upon spared host cardiac cells to (a) attenuate apoptosis [27, 28] and/or (b) otherwise “protect” surviving cardiomyocytes, and (c) induce neovascularization. The postulated paracrine effects likely result from the effects of multiple cytokines and growth factors that are released from the engrafted cells. The exact paracrine effects remain to be defined.

The point that the effects of the transplanted cells may be paracrine does not mitigate the need for an off-the-shelf product that, because of its constant availability, could be used in urgent situations such as percutaneous interventions for acute coronary syndromes, which are quite frequent in the U.S. and other countries. Indeed, Gnecchi et al. have shown the protective effects of very early cell and conditioned medium therapy in an ischemia-reperfusion rat model of infarction [12]. This type of dramatic response will not occur clinically if treatment is delayed because of the need for expansion of the patient's own MSC population.

The hypothesized paracrine effects of MSC transplantation on cardiomyocytes may be mediated by limitation of the activation of well-known maladaptive hypertrophy-associated signaling pathways turned on by increased myocyte stretch and neurohumoral activation induced receptor stimulation associated with AMI [4, 27]. As pointed out earlier, Gnecchi et al. have reported that MSCs and concentrated conditioned medium, used to culture MSCs with constitutive AKT activation, was markedly protective when injected into peri-infarct regions of rats with acute coronary ligation and reperfusion [12]. In this model, infarct size was markedly reduced because of decreased loss of stressed cardiomyocytes, presumably as a consequence of decreased apoptotic and necrotic cell death [12].

We recently reported that conditioned medium from modified porcine MSCs that overexpress vascular endothelial growth factor (VEGF) inhibit apoptosis in isolated HL-1 cardiac cells subjected to stress [4]. In the corresponding in vivo study, the rates of VEGF-MSC engraftment and apparent differentiation were also low, arguing against a direct structural contribution of the engrafted MSCs to beneficial effects despite marked protective effects on LV contractile function and bioenergetic characteristics. Taken together, all these experimental data support the view that MSCs are capable of secreting substances that are beneficial to stressed native cardiomyocytes. The exact paracrine stimuli and altered patterns of signaling pathway activation in transplanted hearts remain to be defined in future studies.


A modified culture medium with low serum content that contained selected growth factors facilitated a high efficiency of isolation of bone marrow-derived MSCs. Importantly, when these MSCs were maintained and expanded through 96 PDs, normal cytogenetic characteristics were preserved. In vitro differentiation capacity was maintained for at least 80 PDs, and transplantation of 80-PD MSCs into immunodeficient mouse hearts at the time of AMI resulted in significant, persistent engraftment. The apparent differentiation of the engrafted cells into donor-derived myocyte or endothelial cells was quite modest, suggesting that a trophic mechanism is the basis of the beneficial effects. The potential clinical significance of the present study is that we report a simple method that can be used for routinely successful isolation and extensive expansion of (autologous or allogeneic) MSC lines. The methodology presented is useful in facilitating clinical transplantation therapy (of both autologous and allogenic MSCs) in patients with acute coronary syndromes and/or in hearts with chronic cardiomyopathy.


The authors indicate no potential conflicts of interest.


This work was supported by U.S. Public Health Service Grants HL67828, HL50470, HL61353, and HL70970. Y.N. and X.W. contributed equally to this article.