Toward an ideal animal model to trace donor cell fates after stem cell therapy: Production of stably labeled multipotent mesenchymal stem cells from bone marrow of transgenic pigs harboring enhanced green fluorescence protein gene¹

ABSTRACT

The discovery of postnatal mesenchymal stem cells (MSC) with their general multipotentiality has fueled much interest in the development of cell-based therapies. Proper identification of transplanted MSC is crucial for evaluating donor cell distribution, differentiation, and migration. Lack of an efficient marker of transplanted MSC has precluded our understanding of MSC-related regenerative studies, especially in large animal models such as pigs. In the present study, we produced transgenic pigs harboring an enhanced green fluorescent protein (EGFP) gene. The pigs provide a reliable and reproducible source for obtaining stable EGFP-labeled MSC, which is very useful for donor cell tracking after transplantation. The undifferentiated EGFP-tagged MSC expressed a greater quantity of EGFP while maintaining MSC multipotentiality. These cells exhibited homogeneous surface epitopes and possessed classic trilineage differentiation potential into osteogenic, adipogenic, and chondrogenic lineages, with robust EGFP expression maintained in all differentiated progeny. Injection of donor MSC can dramatically increase the thickness of infarcted myocardium and improve cardiac function in mice. Moreover, the MSC, with their strong EGFP expression, can be easily distinguished from the background autofluorescence in myocardial infarcts. We demonstrated an efficient, effective, and easy way to identify MSC after long-term culture and transplantation. With the transgenic model, we were able to obtain stem or progenitor cells in earlier passages compared with the transfection of traceable markers into established MSC. Because the integration site of the transgene was the same for all cells, we lessened the potential for positional effects and the heterogeneity of the stem cells. The EGFP-transgenic pigs may serve as useful biomedical and agricultural models of somatic stem cell biology.

bone marrow, enhanced green fluorescence protein, mesenchymal stem cell, myocardial infarction, pig

INTRODUCTION

Mesenchymal stem cells (MSC) derived from bone marrow possess a self-renewal capacity and the potential to differentiate into the many cell types of each embryonic germ layer (Charbord, 2010). These properties qualify MSC as an ideal system to study lineage commitment and differentiation in vitro. After transplantation, some MSC migrate to sites of tissue damage (Kawada et al., 2004; Chamberlain et al., 2007). The discovery that MSC can escape alloantigen recognition, display immunosuppressive properties (Tolar et al., 2010), or both, has stimulated interest in using MSC for therapies. However, the fate of the transplanted MSC in various disease models and the mechanisms leading to their therapeutic effect rely on a reliable tracking system.

Ectopic expression of the green fluorescent protein (GFP) gene under the control of housekeeping gene regulatory elements is an effective means of labeling MSC (Partridge and Oreffo, 2004; Chan et al., 2005). However, these exogenous reporters may fail to integrate into the host genome or can be silenced after prolonged culture (Bestor, 2000; McMahon et al., 2006). Isolation of stably integrated, GFP-tagged MSC from a transgenic donor animal can be an effective solution. Ogawa et al. (2004) provided evidence that MSC derived from GFP-reporter mice maintain stable GFP expression in all differentiated cell types. Previous work by our group demonstrated a reversal of osteoporotic symptoms in mice receiving GFP-integrated donor MSC (Hsiao et al., 2010).

We report herein the creation of transgenic pigs harboring enhanced green fluorescent protein (EGFP), a useful, reliable, and reproducible source of EGFP-labeled MSC. These cells can be easily tracked via GFP-fluorescence after transplantation. The EGFP-expressing donor MSC are readily identified in recipient animals despite the background autofluorescence of myocardial infarcts. The EGFP-transgenic pigs may provide a useful tool for transplantation biology and regenerative medicine studies.

MATERIALS AND METHODS

All experimental procedures were in accordance with the Guide for the Care and Use of Agricultural Animals in Research and Teaching (FASS, 2010) and were approved by the Institutional Animal Care and Use Committee of National Taiwan University.

EGFP-Transgenic Pig Production

The EGFP expression transgene (2.9 kb), CAG-EGFP, including the CAG promoter [a hybrid of cytomegalovirus (CMV) early enhancer element and chicken β-actin promoter] derived from the pCAGGS expression vector, the EGFP cDNA, and the rabbit β-globulin 3′ flanking sequence, was digested from the pCAGGS expression vector (pCX-EGFP; Niwa et al., 1991; Okabe et al., 1997; Lee et al., 2007) using ScaI and PstI, and purified by 2 CsCl₂ gradient ultracentrifugation steps. Each fertilized egg collected from superovulated gilts was microinjected with the purified 2.9-kb EGFP-containing transgene into a single pronucleus. These manipulated eggs were then transferred into synchronized foster mothers as described previously (Lee et al., 2003). Briefly, pubertal crossbred gilts aged 8 to 10 mo were fed Regu-Mate (20 mg/d; containing 0.4% altrenogest, Intervet International BV, Boxmeer, the Netherlands) for 15 d. All donor gilts were injected with 2,000 IU of PMSG (Folligon, Intervet International BV), and 80 h later, were injected with 1,500 IU of hCG (Chorulon, Intervet International BV). Recipient gilts were synchronized to the donors by injection with 1,000 IU of PMSG and 750 IU of hCG. Animals were artificially inseminated, and embryos were surgically collected 56 to 58 h after hCG injection by flushing from the oviduct with Dulbecco's PBS (Invitrogen, Grand Island, NY). The embryos were centrifuged at 12,000 × g for 10 min at room temperature, and a single pronucleus was microinjected with 1 to 2 pL of the DNA construct (1 ng/μL). Twenty to 30 normal-appearing injected embryos were then implanted into the oviducts of each recipient animal.

Analysis of Transgene Integration of EGFP-Transgenic Pigs

Genomic DNA of newly farrowed piglets was obtained from ear biopsies minced into fine particles with scissors, lysed in buffer (50 mM Tris-HCl, pH 8.0, 100 mM EDTA, 100 mM NaCl, 500 μg of proteinase K, and 1% SDS), and then extracted with phenol-chloroform. Screening of piglets was performed by PCR amplification of genomic DNA by using the EGFP primer set: 5′-GAATTAGCCACCATGGTGAGC-3′ and 5′-TGAATTCTTACTTGTACAGCTCGTCC-3′. The PCR reaction was carried out for 30 cycles, with denaturation at 94°C for 30 s, annealing at 55°C for 1 min, and extension at 72°C for 1 min, in a thermal cycler (GeneAmp2700, Applied Biosystems, Foster City, CA). The PCR products were analyzed on a 1.5%-agarose gel containing ethidium bromide and visualized by UV transillumination. Results were confirmed by Southern blot analysis. In brief, 15 μg of genomic DNA was digested overnight with EcoRI, electrophoresed, and transferred to a Hybond-N membrane (Amersham Bioscience, Buckinghamshire, UK). A fragment of EGFP-specific cDNA was used as a radioactive probe to hybridize to the membrane. Blots were scanned with a Typhoon 9200 Imager (Amersham Bioscience).

Isolation and Culture of Bone Marrow-Derived EGFP-MSC

Bone marrow aspirates were obtained from EGFP transgenic pigs. Anesthesia was induced with ketamine (10 mg/kg of BW, Sigma-Aldrich, St. Louis, MO) and maintained with inhalation of anesthetic halothane (Sigma-Aldrich, St. Louis, MO). The tibia area was prepared and sterilized, and approximately 5 mL of bone marrow was aspirated into a syringe containing 6,000 U of heparin. Bone marrow mononuclear cells were obtained by negative immunodepletion of CD3⁺, CD14⁺, CD19⁺, CD38⁺, CD66b⁺, and glycophorin A⁺ cells using a commercially available kit (RosetteSep, StemCell Technologies, Vancouver, British Columbia, Canada), according to the manufacturer's instructions (Lee et al., 2004). After a 20-min incubation at room temperature, the cell and antibody mixture was diluted with twice the volume of PBS supplemented with 2% fetal bovine serum (FBS; Hyclone, Logan, UT) and 1 mM EDTA (Invitrogen), layered over an equal volume of Ficoll-Paque medium (1.077 g/cm³; Amersham Bioscience), and centrifuged at 300 × g for 30 min at room temperature. Enriched cells were harvested from the buffy coat and washed twice with control medium, consisting of minimum essential medium (MEM) α (Sigma-Aldrich) supplemented with 20% FBS (Hyclone), 2 mMl-glutamine (Invitrogen), 100 U/mL of penicillin, and 100 μg/mL of streptomycin (Invitrogen). The cells were then seeded at a concentration of 1 × 10⁶ cells/cm². After incubation at 37°C in a humidified atmosphere containing 95% air and 5% CO₂ for 72 h, the nonadherent cells were removed by changing the medium. At this time, single cell-derived MSC colonies were observed, and subsequent medium changes were performed twice weekly. When these primary cultures were near confluence, the cells were lifted by incubation with 0.25% trypsin:1 mM EDTA (Invitrogen) for 5 min at 37°C, replated at 1:2 to 1:3 dilutions, and used for subsequent experiments, as described below.

Fluorescence-Activated Cell-Sorting Analysis

Passage 4 to 7 cells were collected by incubation with 0.25% trypsin-EDTA for 5 min at 37°C and resuspended at a density of 1 × 10⁶ cells/mL in ice-cold washing buffer consisting of PBS with 2% FBS. The cells were immunolabeled with phycoerythrin-conjugated mouse anti-pig-CD45, -CD29, -CD31, -CD44, and -CD90 (eBioscience, San Diego, CA) at 4°C for 30 min. After washing twice with washing buffer, cells were fixed with 1% formaldehyde (Sigma-Aldrich) in PBS containing 2% FBS. The cells were then analyzed using a fluorescence-activated cell-sorting (FACS) scan flow cytometer (Becton Dickinson, San Jose, CA).

Differentiation Assays

Adipogenic Differentiation.

To induce adipogenic differentiation, the cells were cultured to near confluence, and then treated with adipogenic induction medium consisting of MEM α (Sigma-Aldrich) supplemented with 10% FBS, 10 μg/mL of insulin (Sigma-Aldrich), 1 μM dexamethasone (Sigma-Aldrich), 0.5 mM isobutyl-methylxanthine (Sigma-Aldrich), and 100 μM indomethacin (Sigma-Aldrich) for 2 wk, with medium changes twice per week. At the end of the differentiation period, the cells were fixed with 10% formalin for 10 min, and lipid droplets were detected by Oil Red O (Sigma-Aldrich) staining (Peister et al., 2004).

Osteogenic Differentiation.

To induce osteogenic differentiation, the cells were grown to confluence and incubated in osteogenic induction medium consisting of MEM α supplemented with 10% FBS, 0.1 μM dexamethasone, 10 mM β-glycerophosphate (Sigma-Aldrich), and 50 μM ascorbic acid (Sigma-Aldrich) at 37°C in 5% CO₂ in air for 3 wk, with medium changes twice weekly. At d 14, the committed osteogenic cells were characterized by alkaline phosphatase assays (in 't Anker et al., 2003). Parallel plates of cells were fixed with 10% formalin at d 21, and bone matrix mineralization was determined by alizarin red S (Sigma-Aldrich) staining (Peister et al., 2004).

Chondrogenic Differentiation.

To induce chondrogenic differentiation, the cells were trypsinized and counted, and 2.5 × 10⁵-cell aliquots were centrifuged at 150 × g for 5 min at room temperature in 15-mL conical polypropylene tubes. The culture medium was then replaced with 1 mL of chondrogenic induction medium consisting of MEM α supplemented with 1% FBS, 6.25 μg/mL of insulin, 50 μM ascorbic acid, and 10 ng/mL of transforming growth factor-β1 (R&D Systems, Minneapolis, MN) and incubated at 37°C and 5% CO₂ in air for 3 wk with medium changes twice weekly. The production of proteoglycan was determined by toluidine blue staining (Johnstone et al., 1998).

Examination of EGFP Expression in the MSC Cells and Differentiated Cells

The EGFP-MSC and differentiated cells were examined regularly for EGFP expression by using an inverted epifluorescence microscope equipped with a UV light source and GFP filter set. To determine the percentage of EGFP-positive cells in each culture, the EGFP-MSC and differentiated cells, including osteogenic, adipogenic, and chondrogenic cells, were trypsinized with 0.25% trypsin-EDTA for 10 min at 37°C and pipetted to obtain single-cell suspensions. The chondrogenic cells from 3-dimensional aggregate cultures were dispersed with 1 mg/mL of collagenase type IA (Sigma-Aldrich) for 6 h at 37°C with shaking. Cells from each culture were washed twice with PBS supplemented with 2% FBS and analyzed by FACS.

Western Blot Analysis

Total protein was extracted with 0.5 mL of ice-cold lysis buffer containing 10 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1% Nonidet P-40 (Sigma-Aldrich), and 0.1% SDS for 15 min on ice. After being heated for 5 min at 95°C, 30 μg of protein lysate was analyzed on a 12% SDS-PAGE gel. Afterward, proteins were transferred to polyvinylidene fluoride membrane filters, probed with goat anti-GFP polyclonal antibody (Abcam, Cambridge, UK), and developed with enhanced chemiluminescent reagents (Amersham Bioscience).

Mitogen Proliferation Assays

The mononuclear spleen cells from C57BL/6 and BALB/c mice were prepared by using Ficoll-Paque density centrifugation (1.077 g/mL; Amersham Pharmacia Biotech, Piscataway, NJ), followed by 2 washes in RPMI 1640 (Invitrogen) supplemented with 50 µM 2-mercaptoethanol (Invitrogen), 10% FBS, 100 U/mL of penicillin, and 100 µg/mL of streptomycin. A graded number of mitomycin C (Sigma-Aldrich)-treated mouse MSC (mMSC) were seeded in triplicate in flat-bottomed 96-well plates and maintained at 37°C for 6 h. Splenocytes (2 × 10⁵ cells/well) from C57BL/6 mice containing 5 µg/mL of concanavalin-A (ConA; Sigma-Aldrich) were cultured with or without EGFP-MSC. Proliferation assays were performed after 3 d using a CellTiter 96 Aqueous Nonradioactive Cell Proliferation Assay kit (Promega, Madison, WI), according to the manufacturer's instructions (Zolnai et al., 1998). The absorbance was measured using a SpectraMax 190 ELISA plate reader (Molecular Devices, Sunnyvale, CA) at 490 nm.

Myocardial Infarction and Intramyocardial EGFP-MSC Cell Injection

Myocardial infarction (MI) was induced in mice (C57BL/6) 8 to 12 wk of age using permanent ligation of the left anterior descending (LAD) coronary artery, as described previously (Lian et al., 2011). Mice were randomized into 3 groups as follows: a) a control group, in which mice underwent a thoracotomy and cardiac exposure without ligation, but with PBS solution injected into the myocardium at 2 sites around the peri-infarct region, mimicking an EGFP-MSC injection; n = 6); b) mice with induced MI (MI group; n = 6); and c) MI + EGFP-MSC, in which mice in the MI group were injected with EGFP-MSC (MI-EGFP group; n = 6). Transplantation of MSC was performed within 15 min after LAD coronary artery ligation. A total of 5 × 10⁵ EGFP-MSC were injected at 2 sites into the peri-infarct regions with 10 μL of cell suspension per site (MI-EGFP group). The MI group received PBS injections (2 sites, 10 μL per site) instead of EGFP-MSC after LAD coronary artery ligation. Echocardiography was performed at 2 wk postoperatively to determine functional improvement in animals that received EGFP-MSC compared with those that received PBS. Cell therapy was monitored by optical fluorescence imaging (FI), using endogenous EGFP as the reporter. Results of FI were validated by ex vivo histological examination with a photon imager.

Immunohistochemistry

Immunohistochemistry was performed as described previously (Lian et al., 2011). Briefly, tissue slides were dewaxed, treated with epitope retrieval buffer (Thermal Scientific Inc.), and quenched with 30% H₂O₂. The slides were blocked with Power Block (Biogenex, San Ramon, CA) and incubated for 10 min at room temperature. Sections were then incubated with rabbit anti-Ki67 (Abcam, 1:50) or rabbit anti-GFP (Abcam, 1:250). Anti-rabbit antibody and horseradish peroxidase-labeled polymer (Nichirei, Tokyo, Japan) were applied to slides with diaminobenzidine detection buffer (Biogenex). The nuclei were stained with hematoxylin (J.T. Baker, Phillipsburg, NJ) for microscopic studies and quantification.

Echocardiography

Mice were anesthetized with pentobarbitol (50 mg/kg of BW, intraperintineally). The anterior chest was shaved, and the mouse was laid in a left decubitous position with application of gel on the chest wall for better scan-head–skin contact. Echocardiography was performed using a commercially available echocardiographic system (HDI 5000, Philips, Bloomfield, NJ) equipped with a 15-MHz transducer by an experienced operator blinded to the identity of the groups. Measurements included heart rate, left-ventricle (LV) dimension in both systolic and diastolic stages, LV fractional shortening, and ejection fraction. Data were analyzed using digitized pictures of the long-axis views.

Fluorescence Imaging for Transplanted Cell Survival

Cardiac fluorescence imaging (Φ Imager System, Fujifilm Life Science, Stamford, CT) for quantitative measurement of GFP signals from the LV side, including the ligation site and points of EGFP-MSC injection, was performed ex vivo after echocardiography analysis. The hearts were explanted and images were acquired using a 1- to 2-min interval until the peak signal was observed. The FI was quantified by creating polygonal regions of interest over the precordium.

Histological Examination

For histological examination, the chests of the anesthetized mice were opened and the hearts were perfused with 4% paraformaldehyde (Sigma-Aldrich) via LV stab. Fixed hearts from the study and control groups were harvested, kept in paraformaldehyde for 2 h, rinsed with PBS, and embedded in optimal cutting temperature compound (OCT; Miles Scientific, Naperville, IL). Frozen sections (10 μm thick) were processed for hematoxylin and eosin staining.

Statistical Analysis

The results of cardiac functions examined by echocardiogram are presented as means ± SE. Comparisons between experimental groups and the control were made using Student's 2-tailed t-test. P-values <0.05 were considered statistically significant.

RESULTS

Generation of Transgenic Pigs Harboring the EGFP Gene

To generate transgenic pigs harboring the EGFP gene, a DNA construct that carried the CAG hybrid promoter, EGFP cDNA, and rabbit β-globulin 3′ flanking sequence (Figure 1A) was first separated from plasmid pCX-EGFP, purified twice by CsCl₂ gradient ultracentrifugation, and microinjected into 256 pig embryos. The embryos were then transferred to 8 synchronized recipient gilts, 4 of which successfully became pregnant. Thirty-six potential transgenic founder pigs were born, and 3 were identified as being transgenic by PCR screening (data not shown). To estimate the transgene copy numbers and to confirm the integration of the transgene DNA in these transgenic pigs, the restriction enzyme EcoRI was used to digest the genomic DNA and was subjected to Southern blot analysis. On hybridization using the EGFP-specific cDNA as a probe, a 739-bp band representing the transgene was observed in all cases (Figure 1E). Different intensities of the hybridization signals indicated the presence of different transgene copy numbers in these pigs. In addition, we performed quantitative PCR analysis to further estimate the approximate copy number compared with the standard curve composed of control genomic DNA with 0, 0.5, 1, 2, 4, and 8 times the molar ratio of EGFP plasmids. The estimated copy numbers of all transgenic pigs were <2 (data not shown). The transgenic pig 64-9 had the smallest copy number, consistent with the results of the Southern blot analysis, as well as the brightness of GFP signals. The reason for pig 64-9 having less than 1 copy based on the quantitative analysis was most likely due to the mosaicism observed in the fibroblast of its ear biopsy.

Figure 1.

Characterization of enhanced green fluorescent protein (EGFP)-transgenic pigs generated by pronuclear microinjection. A) The 2.9-kb transgene consisting of CAG promotor [a hybrid of cytomegalovirus (CMV) early enhancer element and chicken β-actin promoter], EGFP cDNA, and rabbit β-globulin 3′ flanking sequence. B) Wild-type pig (left) and EGFP-transgenic pig (right) under natural light. C) Wild-type pigs (middle and right) and EGFP-transgenic pig (left) expressing robust amounts of EGFP under blue light and a green fluorescent protein (GFP) filter set. D) Three EGFP-transgenic pigs [number 64-8 (left), 64-9 (middle), and 84-1 (right)] under blue light and a GFP filter set. E) Southern blot assay of EcoRI-digested genomic DNA of EGFP-transgenic pigs (probe: 739 bp). M = marker; Pc = positive control; Nc = negative control. F to Q) Almost all tissues and organs of the EGFP-transgenic pig [right panel, Landrace-Yorkshire strain pig (LY pigs)] expressed robust amounts of EGFP under blue light and a GFP filter set, including that of F) ribs, G) kidney, H) heart, I) muscle, J) liver, K) brain, L) eyes, M) tongue, N) lung, O) intestines, P) testis, and Q) penis. Wild-type tissues and organs (Lee-Sung mini-pigs, left panel in F to O and upper sample in M) were used as the control.

Under daylight, the EGFP transgenic pigs have green-tinged skin, especially in their snout, eyes, and hooves, so they can be easily distinguished from wild-type pigs (Figure 1B). The transgenic pigs glow green throughout when exposed to blue light (390 nm; Figure 1C). All 3 transgenic pigs expressed robust quantities of EGFP (Figure 1D). Strong green fluorescence was also observed in most tissues and organs (Figures 1F to 1Q). Because of the mosaicism observed in 64-9, all the following EGFP-MSC related studies were based on the cells derived from EGFP-transgenic pigs 64-8 and 84-1.

Expression of Exogenous EGFP in the Bone Marrow-Derived EGFP-MSC

Marrow-derived nucleated cells were seeded at an initial density of 1 × 10⁶ cells/cm² in a flask containing MEM α supplemented with 20% FBS. By d 3, nonadherent cells were removed by decanting the medium, and plastic-adherent cells with elongated fibroblast-like morphology were apparent and formed symmetrical colonies. When these primary cells had grown to confluence, they were trypsinized and replated at a dilution of 1:2 to 1:3. These cells proliferated rapidly, with a population doubling time of 23 ± 0.9 h (n = 6), and were named EGFP-MSC. The EGFP-MSC appeared as spindle-shaped fibroblastoid cell populations (Figure 2A). They demonstrated no obvious reduction in mitogenic properties. They could be continuously propagated in culture, and their morphology was similar after 4 to more than 40 population doublings (data not shown).

Figure 2.

Morphology and surface marker expression of enhanced green fluorescent protein-labeled mesenchymal stem cells (EGFP-MSC). A) Phase contrast image of EGFP-MSC showing spindle-shaped morphology. B) The EGFP-MSC expressing green fluorescence. C) The EGFP protein expression of EGFP-MSC, detected by Western blotting. Cont. = control. D) The percentage of EGFP expressing MSC, as determined by fluorescence-activated cell sorting (FACS). E) Surface antigens of EGFP-MSC, analyzed by a FACS histogram. Numbers indicate the percentage of stained cells in the population (red) compared with the unstained control (white). The x-axes represent the relative fluorescence intensity (D and E). Scale bars represent 50 µm (A and B). M1 = marker 1.

Fluorescence microscopy was used to detect EGFP in the MSC. As shown in Figure 2B, these EGFP-MSC expressed uniform, increased abundance of EGFP. The EGFP expression in EGFP-MSC was also evaluated by Western blot analysis by using an antibody directed against the full-length AA sequence of EGFP (Figure 2C). Moreover, we showed that almost 100% of the EGFP-MSC were EGFP positive; these cells could be easily distinguished from wild-type MSC by FACS (Figure 2D). The EGFP expressed in EGFP-MSC was sustained at increased amounts without any decline, even when they stopped multiplying when aged in vitro (data not shown).

Phenotypic Characterization of EGF-MSC: FACS Analysis

To understand the phenotypic characteristics of EGFP-MSC, we analyzed their cell surface antigens. As shown in Figure 2E, 99.7% of the EGFP-MSC expressed CD44 (phagocytic glycoprotein I), an adhesion receptor that mediates cell attachment to hyaluronate and osteopontin; 15.2% expressed CD90 (thymus cell antigen-1), which is involved in the regulation of adhesion and signal transduction by T cells; and 100% expressed CD29 (integrin β1), which represents a receptor for the extracellular matrix. Almost none (0.5%) of the EGFP-MSC expressed CD45 (leukocyte common antigen), a pan-hematopoietic marker, and they were also negative (0.4%) for CD31 (platelet endothelial cell adhesion molecule), an endothelial cell marker. On the basis of their surface marker patterns, the EGFP-MSC were characterized as pig MSC (Lim et al., 2006).

Trilineage Differentiation and EGFP Examination of EGFP-MSC and Their Differentiated Progeny

To assess their osteogenic potential, EGFP-MSC were grown to confluence and cultured in osteogenic induction medium. In the initial culture, these cells continued to proliferate, formed multilayers, and showed calcium crystal deposition (Figure 3A). A greater (P < 0.05) alkaline phosphatase activity, a marker of osteoblastic differentiation, was observed after d 14 of osteogenic treatment (Figure 3C). These cultures were also studied for their ability to develop a mineralized matrix at d 21. After alizarin red S staining, red-mineralizing areas appeared within the cultures (Figure 3D). In addition, we showed that the EGFP-MSC differentiating toward the osteogenic pathway also retained their greater EGFP expression (Figure 3B); 98.5% of the osteogenic cells were EGFP positive when analyzed by FACS (Figure 3J).

Figure 3.

Multilineage differentiation of enhanced green fluorescent protein-labeled mesenchymal stem cells (EGFP-MSC) in vitro. A to D) Osteogenesis differentiation of EGFP-MSC. A) The EGFP-MSC cultured in osteogenic medium for 3 wk. The arrow indicates calcium crystal deposition. B) Osteogenic differentiated EGFP-MSC express EGFP, as observed by fluorescence imaging. C) Osteogenic differentiated EGFP-MSC, stained with alkaline phosphatase at d 14. D) The intense red staining indicates deposition by the cells of a calcified extra cellular matrix, indicated by alizarin red S (ARS) staining. E and F) Chondrogenic differentiation of EGFP-MSC. E) Chondrogenic differentiated EGFP-MSC express EGFP, observed by fluorescence microscopy. F) Histological section of a micromass pellet, stained with toluidine blue and examined under bright field illumination. The uniform, intense purple staining reveals the rich glycosaminoglycan content of the pellet. G to I) Adipogenic differentiation of EGFP-MSC. G) The EGFP-MSC cultured in adipogenic medium contains oil drops. H) Adipogenic differentiated EGFP-MSC express EGFP, observed by fluorescence microscopy. I) Numerous Oil Red O-stained lipid vesicles are present within the cells. J to L) After J) osteogenic, K) chondrogenic, and L) adipogenic differentiation, EGFP-MSC still harbors greater abundance of EGFP expression. The x-axes represent the relative fluorescence intensity (J to L). Scale bars represent 50 µm (A to I). M1 = marker 1.

To investigate their chondrogenic potential, EGFP-MSC were first isolated by monolayer culture and then subjected to a pelleted micromass system in chondrogenic induction medium. After 3 wk of differentiation, the metachromatic nature of the matrix was visualized by toluidine blue staining (Figure 3F). We showed that EGFP-MSC expressed significant EGFP even in the pelleted micromass system (Figure 3E). The EGFP expression of these cells was shown by FACS, in which 99.1% of the cells were EGFP positive after enzymatic dispersion (Figure 3K).

To assess their adipogenic potential, EGFP-MSC were also grown to confluence and cultured in adipogenic induction medium. Small lipid droplets within the cytoplasm were visible after 14 d (Figure 3G), as illustrated by Oil Red O staining (Figure 3I). Adipogenic EGFP-MSC retained intense EGFP fluorescence (Figure 3H), with 99% of the cells positive for EGFP as determined by FACS (Figure 3L).

Immunosuppressive Characteristics of the EGFP-MSC

To determine whether EGFP-MSC had immunosuppressive properties in vitro, we cocultured mitomycin C-treated EGFP-MSC with splenocytes from C57BL/6 mice in the presence of ConA. The EGFP-MSC inhibited (P < 0.05) ConA-induced splenocyte proliferation in a dose-dependent manner, as compared with the corresponding negative control (100%) without EGFP-MSC (n = 3; Supplemental Figure 1; http://jas.fass.org/content/vol89/issue11).

Engraftment and Growth of the EGFP-MSC Cells

To identify the transplanted cells in myocardial sections, we injected 5 × 10⁵ undifferentiated EGFP-MSC into the myocardium surrounding the LAD coronary artery ligation site. Fluorescence microscopy showed that these EGFP-expressing MSC grew in situ and were present at the peri-infarct regions of the myocardium (Figure 4B) 2 wk after ligation of the LAD coronary artery compared with the autofluorescent signal present in the control group (Figure 4A). This observation was confirmed by ex vivo imaging of explanted whole hearts after EGFP-MSC transplantation (Figure 4L). The strong GFP fluorescent signals from the cryosectioned samples (Figure 4B), the GFP antibody staining signals from the paraffin sections in the MI + EGFP-MSC group (Figure 4E compared with Figure 4C and 4D), as well as the strong GFP signals from the ex vivo measurement (Figure 4L) indicated that the transplanted cells could survive within the ischemic myocardium during this period.

Figure 4.

Florescent and histological examination of infarcted hearts in mice after enhanced green fluorescent protein-labeled mesenchymal stem cell (EGFP-MSC) delivery repair for 2 wk. A) Fluorescence image of control myocardium (mice chests were opened without ligation of left anterior descending coronary artery, and PBS solution was injected into the myocardium), showing background autofluorescence. B) The presence of EGFP-MSC at the peri-infarct regions of the myocardium after myocardial infarction (MI). Arrows indicate the injected spindle-shaped EGFP-MSC. Nuclei are counterstained with 4′,6-diamidino-2-phenylindole dihydrochloride (DAPI; blue). C to H) Immunohistochemical staining showing green fluorescence protein (GFP)-positive cells (C to E) and Ki67-positive cells (F to G) retained within peri-infarct regions of the myocardium after MI. I to K) Hematoxylin-eosin staining of the left ventricle wall MI area under different conditions: I) control group, J) MI group, K) MI mice injected with EGFP-MSC. L) Ex vivo imaging confirming the presence of transplanted cells within the myocardium. Scale bars represent 50 µm (A and B), 20 µm (C to H), and 1 mm (I to K).

Histological evaluation revealed a thicker LV infarct wall area and attenuation of LV enlargement after intramyocardial delivery of EGFP-MSC (Figure 4I to 4K), associated with increased cell proliferation (Figure 4H compared with Figure 4F and 4G), compared with the MI group without EGFP-MSC treatment. The Ki67-positive signal observed in the MI group (arrow, Figure 4G) seemed to correlate with the formation of fibrosis after the MI condition.

Effects of EGFP-MSC on Cardiac Function After MI

We performed an echocardiographic study 2 wk after MI in all mice. We showed that transplantation of EGFP-MSC significantly increased systolic and diastolic thicknesses of the infarcted LV postwall (Figures 5H and 5I). The LV internal diameter at systolic and diastolic was shorter (P < 0.05) in the MI-EGFP group than in the MI group (Figures 5E and 5F). In addition, the intraventricular septum at systolic and diastolic was improved (P < 0.05) in the MI-EGFP group compared with the MI group (Figures 5B and 5C). Moreover, it was readily apparent that the fraction shortening (Figure 5G), ejection fraction (Figure 5D), and heart rate (Figure 5A) were better preserved (P < 0.05) in the infarcted heart with the MSC injection than in the infarcted heart with the PBS injection.

Figure 5.

Echocardiographic analysis of infarcted hearts in mice after enhanced green fluorescent protein-labeled mesenchymal stem cell (EGFP-MSC) delivery repair for 2 wk. A to I) The left ventricular functional variable showed a difference in each group in mice. FS = fraction shortening; EF = ejection fraction; IVSd = intraventricular septum at diastolic; LVIDd = left ventricular internal diameter at diastolic; LVPWd = left ventricular postwall at diastolic; IVSs = intraventricular septum at systolic; LVIDs = left ventricular internal diameter at systolic; LVPWs = left ventricular postwall at systolic. Control = chests of mice were opened without ligation of the left anterior descending coronary artery, and PBS was injected into the myocardium; MI = myocardial infarction mice; MI + EGFP-MSC = MI mice injected with EGFP-MSC. Results are the mean ± SE of 6 experiments in each group; *P < 0.05; **P < 0.01.

DISCUSSION

The generation of transgenic pigs harboring EGFP and the subsequent therapeutically functional EGFP-MSC lines have the following advantages. First, the domestic pig, Sus scrofa, is a valuable model organism for biomedical research because of its suitable body size and similar physiology compared with humans. Second, stem cells from GFP-transgenic pigs are very useful in cell transplantation studies because they provide a continuous source of green cells, which can be tracked easily. Under the control of the CAG promoter [the combination of the CMV early enhancer element and chicken β-actin promoter], we demonstrated that EGFP-transgenic pigs generated by pronuclear microinjection expressed EGFP in all tissues and organs tested. Third, although other groups have generated GFP-tagged MSC through in vitro transfection/transduction (Conget and Minguell, 2000; Partridge and Oreffo, 2004; Hoelters et al., 2005; McMahon et al., 2006), our EGFP-MSC derived from EGFP-transgenic pigs 64-8 and 84-1 had the major advantages of sustained, ubiquitous EGFP expression in stem cells and all their differentiated progeny. Because the EGFP-MSC derived from the same transgenic donor usually share the common DNA integration site(s), we therefore lessened the concern regarding positional effects and cell heterogeneity, compared with transfecting GFP transgene into established MSC lines in vitro, which by itself is a challenging procedure in this cell type (Helledie et al., 2008). Fourth, the other advantage compared with transfecting transgenes into established MSC cultures in vitro is that our transgenic model has the advantages of obtaining stem and progenitor cells in earlier passages when the stem cells have better potency. Finally, we demonstrated that the EGFP-MSC could be detected after transplanting it into infarcted myocardium and could restore the heart function. The heart is one of the organs that exhibits the strongest autofluorescence (Nussbaum et al., 2007). The greater expression of EGFP from our EGFP-MSC still enabled us to distinguish the donor cells from the background autofluorescence even in this organ, suggesting the feasibility of tracking our EGFP-MSC-derived cells for various MSC-related therapeutic studies in pig models of human diseases.

Transgenic pigs harboring the EGFP gene are a useful animal model in regenerative medicine. One of the best examples was demonstrated in a recent publication in which transgenic porcine embryonic fibroblasts harboring the hOCT4-EGFP (also termed OG2) transgene were used as donor cells to make OG2-transgenic pigs by somatic cell nuclear transfer (Nowak-Imialek et al., 2011). These OG2-transgenic pigs pave ways to insightful studies on pig germ cell development, as well as reprogramming studies, because any non-EGFP expressing differentiated cells types can be reprogrammed back to the hOCT4-EGFP transgene expressing pluripotent status by fusion with murine embryonic stem cells or by overexpressing the human OCT4 (octamer-binding transcription factor 4),SOX2 [SRY (sex determining region Y)-box 2], KLF4 (Kruppel-like factor 4), and c-MYC (c-Myc proto-oncogene; Nowak-Imialek et al., 2011). Many mechanistic studies, including the detailed epigenetic regulatory cascade, can be monitored in the selected EGFP⁺ cells during the course of the reprogramming process.

Although the tissue- or cell type-specific transgenic model has great potential in various lines of research in regenerative medicine, one would need to generate transgenic pigs having (almost) all their cell types expressing the reporter EGFP if the task is to track the cell fate of transplanted stem or progenitor EGFP⁺ cells in various pig models of human diseases. Several groups have indeed produced GFP-transgenic pigs under the control of housekeeping regulatory sequences either by transfection of the GFP transgene into various cell types followed by somatic cell nuclear transfer (Park et al., 2001; Brunetti et al., 2008; Klassen et al., 2008; Kawarasaki et al., 2009; Nowak-Imialek et al., 2011), or by using sperm vector-mediated transgenesis (Naruse et al., 2005; Webster et al., 2005; Yong et al., 2006). However, the utility of these pigs was compromised because the GFP expression was limited to certain tissues. None of these studies reported strong GFP signals in EGFP-transgenic pig-derived MSC cultures before and after their differentiation.

We showed here that our CAG hybrid promoter-driven EGFP-MSC derived from transgenic pigs 64-8 and 84-1 had the major advantage of greater sustained EGFP expression in stem cells and in all their differentiated progeny. More than 99% of the cells were defined as EGFP-positive cells by FACS. We did not find any differences between EGFP-MSC established from these 2 transgenic donors. It has been documented that the CMV promoter, or related viral sequences, may attract de novo methylation and therefore impair transgene expression (Kong et al., 2009). In our case, using CMV enhancers to facilitate the activity of the β-actin promoter that drives the EGFP gene did not seem to cause the epigenetic silencing effect. This was judged from the greater EGFP expression in all tested cell types of the EGFP-transgenic pigs throughout their lives. In addition, the isolated EGFP-MSC was still GFP positive even after approximately 35 passages when the cells became senescent. We also observed strong EGFP signals from various EGFP-MSC-derived in vitro differentiated cell lineages.

Using flow cytometry, we found that EGFP-MSC exhibited immunophenotypes consistent with those reported in the literature for pig bone marrow MSC (Lim et al., 2006). We also detected transcripts of NG2 (chondroitin sulfate proteoglycan) and PDGFRβ (platelet-derived growth factor receptor β) (data not shown) that are expressed in human MSC and pericytes from multiple organs (Crisan et al., 2008). In addition, the EGFP-MSC are capable of differentiating toward adipogenic, chondrogenic, and osteogenic pathways, which has been reported as a standard method for defining MSC (Pittenger et al., 1999; Horwitz et al., 2005). In addition, we demonstrated that the EGFP-transgene did not affect the multilineage potential of EGFP-MSC compared with the MSC isolated from the wild-type pigs (data not shown). However, our results showed that these EGFP-MSC presented a greater capacity for osteogenic and chondrogenic differentiation, but were restricted for adipogenic differentiation. This may be a strain-specific feature of the LY (hybrid of Landrace and Yorkshire) pigs that were used for generating our EGFP-tagged transgenic pigs. When we compared the trilineage differentiation potential between MSC from wild-type LY pigs and MSC from Lee-Sung pigs (a Taiwanese strain of mini-pig), we found that MSC from the Lee-Sung pigs demonstrated better potential in differentiating to adipocytes (data not shown). This is consistent with literature reports in rodents indicating that bone marrow-derived MSC isolated from different inbred strains may have different surface epitopes, rates of proliferation, and differentiation potential (Peister et al., 2004).

Intramyocardial injection of the EGFP-MSC significantly increased the heart function, as indicated by the increased indexes of the ejection fraction and fraction shortening, and it inhibited the development of the LV enlargement in mice with chronic heart failure secondary to MI. Our results suggest that transplantation of EGFP-MSC improves cardiac function and can be easily identified through ex vivo fluorescence imaging. Because none of the EGFP-positive transplanted cells acquired the cardiomyocyte fate for the lack of evident α-smooth muscle actin expression, our tracking system enabled us to demonstrate that cellular differentiation was not the primary mechanism underlying the therapeutic effect. This is consistent with reports that MSC-derived cardiomyocyte rarely are detected after delivery in MI hearts (Miyahara et al., 2006; Nakamura et al., 2007). The improved cardiac function after MSC transplantation may therefore be attributed to the cytokine or growth factors, such as vascular endothelial growth factor produced by the transplanted MSC, that promote myogenic repair and prevent fibrosis (Pak et al., 2003; Miyahara et al., 2006). In addition, the therapeutic effect may be attributed to the ability of MSC to inhibit regional apoptotic effects or paracrine pathways to trigger angiogenesis (Fuchs et al., 2001; Kocher et al., 2001; Kinnaird et al., 2004; Miyahara et al., 2006; Tolar et al., 2010).

In summary, we have taken advantage of the EGFP-tagged transgenic pigs generated by pronuclear microinjection to generate traceable MSC. Our studies demonstrate that the EGFP protein is expressed stably after in vitro differentiation of the EGFP-tagged MSC into various cell lineages, including adipocytes, osteoblasts, and chondrocytes. These undifferentiated EGFP-MSC cells and in vitro-differentiated EGFP-tagged cells are therefore the best tools for further study of MSC homing and differentiation activity after transplantation into red fluorescent protein-tagged recipient animals. This will distinguish real differentiation of implanted EGFP-tagged MSC from cell fusion events, as demonstrated in other studies (Li et al., 2007). Moreover, the long-term stable expression of the EGFP marker also allows us to isolate transplanted cells to identify critical molecular regulators, as well as their target genes from the posttransplanted EGFP cells.

Footnotes

1

We thank David Welchman (Global Health Institute, EPFL, Lausanne, Switzerland) and Bernhard Payer (Massachusetts General Hospital/Harvard Medical School, Boston, MA) for discussions and comments on the manuscript; and Masaru Okabe and Jun-Ichi Miyazaki (Osaka University, Osaka, Japan) for providing the the pCAGGS expression vector-enhanced green fluorescent protein plasmid. This work was supported by grants from the National Taiwan University (Taipei; Grant 97R0066-40 to W. T.-K. Cheng; Grant 97R0066-41 to S.-C. Wu; Grant 97R0066-46 to S.-P. Lin) and the National Science Council (Taipei; Grant NSC95-2317-B-002-011- to S.-C. Wu).

Author notes

2

These authors contributed equally to this work.

© American Society of Animal Science 2011