SEARCH

SEARCH BY CITATION

Keywords:

  • Adult stem cells;
  • Cell culture;
  • Differentiation;
  • Microarray

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

Several adherent postnatal stem cells have been described with different phenotypic and functional properties. As many of these cells are being considered for clinical therapies, it is of great importance that the identity and potency of these products is validated. We compared the phenotype and functional characteristics of human mesenchymal stem cells (hMSCs), human mesoangioblasts (hMab), and human multipotent adult progenitor cells (hMAPCs) using uniform standardized methods. Human MAPCs could be expanded significantly longer in culture. Differences in cell surface marker expression were found among the three cell populations with CD140b being a distinctive marker among the three cell types. Differentiation capacity towards adipocytes, osteoblasts, chondrocytes, and smooth muscle cells in vitro, using established protocols, was similar among the three cell types. However, only hMab differentiated to skeletal myocytes, while only hMAPCs differentiated to endothelium in vitro and in vivo. A comparative transcriptome analysis confirmed that the three cell populations are distinct and revealed gene signatures that correlated with their specific functional properties. Furthermore, we assessed whether the phenotypic, functional, and transcriptome features were mediated by the culture conditions. Human MSCs and hMab cultured under MAPC conditions became capable of generating endothelial-like cells, whereas hMab lost some of their ability to generate myotubes. By contrast, hMAPCs cultured under MSC conditions lost their endothelial differentiation capacity, whereas this was retained when cultured under Mab conditions, however, myogenic capacity was not gained under Mab conditions. These studies demonstrate that hMSCs, hMab, and hMAPCs have different properties that are partially mediated by the culture conditions. STEM CELLS 2011;29:871–882


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

Postnatal stem cells can be obtained from different somatic tissues and have the capacity to self-renew and differentiate into multiple cell types. Human mesenchymal stem cells (hMSCs), also known as marrow stromal cells, were originally isolated from bone marrow (BM). They can be expanded in vitro and differentiate into adipocytes, osteoblasts, chondrocytes, tenocytes, myofibroblasts, and hematopoietic supportive stromal cells [1–4]. The clinical suitability of hMSCs for therapies of cartilage and bone defects is currently being tested [5]. Aside from the ability to generate these different cell types, hMSCs are also endowed with the ability to modify the immune response and produce a range of trophic factors, which in vivo leads to significant paracrine effects affecting tissue regeneration by endogenous cells [6, 7]. Many phase I–III studies are ongoing to test whether MSCs are suitable for therapy of immunological diseases (e.g., graft-versus-host disease [GVHD] and Crohn's disease) and ischemic disorders [8–10]. Recent studies have suggested that hMSCs may be a subpopulation of pericytes [11, 12].

A second population of cells, which are thought to be a subpopulation of pericytes, are mesoangioblasts (Mab). Mab were initially isolated from murine fetal aorta, but have since also been isolated from postnatal vessels of skeletal muscle and/or heart of larger mammalian species including human and dog [13]. Human Mab (hMab) are commonly isolated from muscle tissue and can, like MSCs, be expanded ex vivo adherent to plastic culture vessels. In contrast to hMSCs, hMab differentiate with high efficiency toward skeletal myofibers both in vitro and in vivo following transplantation in mice with muscular dystrophy [14]. A phase I–II clinical trial for Duchenne's muscular dystrophy will start shortly [15].

Several other plastic adherent stem/progenitor cell populations have been described that can be cultured from BM, blood, or other tissues [16–20]. Among these cell populations are human multipotent adult progenitor cells (hMAPCs) [21]. Even though hMAPCs are, like hMSCs and hMab, expanded adherent to culture vessels, differences in cell surface phenotype, expansion, and differentiation ability exist. For instance, hMAPCs can be expanded for more protracted periods of time than hMSCs and differentiate in vitro not only into mesenchymal cell types, but also into endothelial cells and cells with hepatocyte-like features [21–24]. Like hMSCs, hMAPCs have also significant trophic effects when grafted in vivo [25, 26]. Clinical studies with clinical grade large-scale expanded hMAPCs (Multistem) have been initiated for the prevention of GVHD as well as treatment of cardiac ischemia (ClinicalTrials.gov NCT00677859, NCT00677222).

Thus, as a number of adherent stem/progenitor cell populations, including hMSCs, hMAPCs, and hMab, are currently used clinically, it is of great importance to address the identity and potency of these different cell preparations. Therefore, in this study, we compared the phenotype and functional capacities of hMSCs, hMAPCs, and hMab using uniform standardized methods. We also determined if the differences in cell phenotype and function were mediated by differences in culture conditions.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

Isolation and Culture of hMAPCs, hMSCs, and hMab

Human MAPCs and hMSCs were isolated from bone fragments (femur) and hMab isolated from skeletal muscle fragments (quadriceps femoris) of children (5- to 15-year old) undergoing orthopedic surgery, after obtaining informed consent in accordance with the guidelines of the Medical Ethics Committee of the University Hospitals Leuven. Isolation and culture of the cells was performed as previously described [4, 14, 21]. Briefly, hMAPCs were generated by flushing the bone fragment and plating the total cell fraction at 0.5 × 106 cells per centimeter square in medium consisting of 60% Dulbecco's modified Eagle's medium (DMEM) low glucose (Gibco, Invitrogen, Carlsbad, CA, www.invitrogen.com), 40% MCDB-201 (Sigma-Aldrich, St Louis, MO, www.sigmaaldrich.com), supplemented with 50 nM dexamethasone, 10−4 M L-ascorbic acid, 1× selenium-insulin-transferrin (ITS), 0.5× linoleic acid-bovine serum albumin (all from Sigma-Aldrich), 1% penicillin/streptomycin (Gibco, Invitrogen), along with 2% Serum Supreme (Lonza BioWhittaker, Basel, Switzerland, www.lonza.com), and human platelet derived growth factor BB (PDGF-BB) (R&D Systems, Minneapolis, MN, www.rndsystems.com) and human EGF (Sigma-Aldrich) (both 10 ng/ml). Human MAPC cultures were maintained under hypoxic conditions (5% O2) in a 5.5% CO2 humidified incubator at a density of 400 cells per centimeter square and were passaged every 2–3 days. Clonal populations were obtained by plating 5 cells per well in a 96-well or 48-well plate between passages 2 and 12.

Human MSCs were generated by flushing the bone fragment and plating the mononuclear fraction, obtained after Ficoll density gradient centrifugation, at 0.5 × 106 cells per centimeter square in MSCs growth medium (Lonza). Human MSC cultures were maintained at 5,000 cells per centimeter square, at normal oxygen levels in a 5% CO2 humidified incubator and were passaged every 4–7 days. MSCs were also purchased from Lonza and cultured under the same conditions as the other MSCs. In our experiments, commercial MSCs were always used along with at least 2 MSC lines which were isolated in our laboratory.

To isolate hMab, the muscle fragment was rinsed in phosphate-buffered saline (PBS) (w/o Ca2+Mg2+), cut into very small pieces (1–2 mm diameter) and transferred to a Petri dish coated with type I collagen (Sigma-Aldrich). The medium consisted of MegaCell DMEM (Sigma-Aldrich) supplemented with 5% fetal bovine serum (FBS) (Lonza BioWhittaker), 5 ng/ml basic fibroblast growth factor (bFGF) (R&D Systems), 2 mM L-glutamine, 0.1 mM β-mercaptoethanol, 1% nonessential amino acids, and 1% penicillin/streptomycin (all from Gibco, Invitrogen). The tissue fragments were cultured for 7–10 days and after the initial outgrowth of fibroblast-like cells, small round, and refractile cells could be observed. Human Mab cultures were maintained at 5,000 cells per centimeter square, under hypoxic conditions (5% O2) in a 5.5% CO2 humidified incubator and were split every 2–3 days.

Population doublings (PDs) were calculated according to the number of cells initially seeded (C0) to the number of cells harvested (C1) using the following equation: PDnew = PDinitial + [log (C0/C1)]/log 2.

Cell Lines

The human embryonic stem cell (hESC) line H9 (purchased from WiCell, Madison, WI, www.wicell.org) was obtained with approval from the Ethical Committee of the K.U.Leuven. Human MSCs were purchased from Lonza (Basel, Switzerland). A detailed description of the maintenance of hESCs is available in the extended methods section in Supporting Information.

Flow Cytometry and Immunofluorescence

A detailed description of the protocols and antibodies used for flow cytometry and immunofluorescence is available in Supporting Information Table S1–S2.

RNA Isolation and Reverse-Transcriptase Quantitative Polymerase Chain Reaction (RT-qPCR)

A detailed description of the protocol and primers used for RNA isolation and RT-qPCR is available in Supporting Information Table S3.

In Vitro Differentiation and Functional Assays

Adipogenic, osteogenic, and chondrogenic differentiation assays were performed using the hMSC functional identification kit (R&D Systems) according to the manufacturer's instructions. Smooth muscle, skeletal muscle, and endothelial differentiation assays, as well as the in vitro and in vivo Matrigel assays were performed as previously described [14, 23, 27]. A detailed protocol of all differentiation and functional assays can be found in Supporting Information.

Microarrays

Human MSCs (n = 5), hMab (n = 3), hMAPCs (n = 7), and hESCs (n = 2, biological replicates) were used for microarray comparison. Human MSCs (n = 3, including one technical replicate for one line), hMab (n = 2), hMAPCs (n = 5, including one technical replicate for one line), MAPC_to_MSC (n = 4), MAPC_to_Mab (n = 2), MSC_to_MAPC (n = 2), and Mab_to_MAPC (n = 2) were used for microarray analysis to assess the influence of the culture conditions. Hybridization method and analysis are described in Supporting Information.

Statistical Analysis

Normal distribution of the data was verified with the Kolmogorov-Smirnov test (GraphPad Prism 5, La Jolla, CA, www.graphpad.com/prism). Comparisons between two groups were analyzed using an unpaired 2-tailed Student's t test or Mann-Whitney U test. For in vitro differentiation assays, comparison of gene expression between different time points was analyzed using a paired 2-tailed Student's t test or a Wilcoxon signed-rank test. p values < .05 were considered significant. Data are shown as mean ± SEM.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

Human MSCs, hMab, and hMAPCs Differ in Their Morphology, Expansion Rate, and Cell Surface Marker Expression

Isolations were done from bone and skeletal muscle fragments of children undergoing orthopedic surgery. Several lines of each cell type were derived from different donors (5–7) to overcome bias due to genetic variation. All experiments were performed using at least two or three different donors per cell type and all cells were used in the exponential proliferation phase. An overview of the different culture conditions in which the three cell types were maintained is given in Figure 1A. Human MSC and hMAPC cultures, both derived from bone fragments, had significant morphological differences: hMSCs were large cells with an elongated shape, whereas hMAPCs were smaller cells with a ruffled border. Human Mab, derived from skeletal muscle fragments, were small, triangular cells, and half of the cells appeared nonadherent (Fig. 1B). These cell populations also differed in their expansion capacity: hMSCs could be expanded for 12–18 passages (∼20–25 PDs) and hMab for 20–25 passages (∼30–35 PDs) before starting to senescence. However, hMAPCs could be maintained for more than 30 passages (>70 PDs) (Fig. 1C).

thumbnail image

Figure 1. Human MSCs, hMab, and hMAPCs differ in their morphology, expansion rate, and cell surface phenotype. (A): Overview of the differences in culture conditions among hMSCs, hMab, and hMAPCs. (B): Phase-contrast morphology of hMSCs, hMab, and hMAPCs in their exponential phase of growth (×20, Axiovert 40C, Zeiss). (C): Expansion curve of a representative example of hMSCs, hMab, and hMAPCs (bulk and clonal populations) shown as days in culture (horizontal axis) to the number of PDs over time (vertical axis). (D): Flow cytometric analysis of hMSCs, hMab, and hMAPCs for CD140a, CD140b, ALP, NG2, MHC class I, and MHC class II. The horizontal line on each histogram indicates the percentage of positive cells for each surface protein (representative example of five hMSCs, four hMab, and four hMAPCs isolations). Abbreviations: ALP, alkaline phosphatase; bFGF, basic fibroblast growth factor; EGF, epidermal growth factor; hMab, human mesoangioblasts; hMAPCs, human multipotent adult progenitor cells; hMSCs, human mesenchymal stem cells; MHC, major histocompatibility complex; NG2, neuron/glial type2 antigen; PD, population doubling; PDGF, platelet-derived growth factor.

Download figure to PowerPoint

Human MSCs and hMab expressed higher levels of CD140a and alkaline phosphatase (ALP) compared with hMAPCs. Human Mab expressed the pericyte marker NG2 and low levels of major histocompatibility complex (MHC) class II, hMAPCs expressed lower levels of MHC class I than hMSCs and hMab. CD140b can distinguish between the three cell populations as it is expressed at high levels on hMSCs, at low levels on hMab and not expressed on hMAPCs (Fig. 1D). Human MAPCs, hMSCs, and hMab were negative for hematopoietic (CD34, CD45, and c-kit) and endothelial (kinase insert domain-containing receptor, CD34) cell surface proteins, for CD56 (antigen expressed on satellite cells) and CD271 and dimly positive for CD146. On the other hand, they were all positive for CD44, CD13, CD73, CD90, and CD105 (not shown).

Human MSCs, hMab, and hMAPCs Can Differentiate into Adipocytes, Osteoblasts, Chondrocytes, and Smooth Muscle Cells In Vitro, but Only hMab Can Form Skeletal Myofibers

Human MSCs, hMab, and hMAPCs showed adipogenic differentiation with significant upregulation of the adipogenic-specific transcript PPARγ2 (Supporting Information Figure S1A). On day 14, 31.67% ± 4.25% hMSC-progeny, 14.5% ± 2.63% hMab-progeny, and 48.13% ± 7.67% hMAPC-progeny contained fat vacuoles that stained positive with oil red O solution (Fig. 2A). Upon differentiation toward osteoblasts, a significant increase in ALP mRNA levels was seen in all three populations and an increase in osteocalcin expression in hMSC and hMAPC progeny, which was significant for hMSCs (Supporting Information Figure S1B). Calcium deposits, stained with an Alizarin red solution on day 28 of differentiation, were quantified following extraction of the dye. Differentiated progeny of hMSCs, hMab, and hMAPCs contained similar quantities of the dye (Fig. 2B, 2C). Chondrogenic differentiation was induced through micropellet cultures and by plating the cells as micromasses in medium containing transforming growth factor β3 (TGF-β3). Proteoglycans and glycosaminoglycans, stained with Alcian blue on day 21 of differentiation, were quantified following extraction of the dye. Similar quantities of the dye were seen in all three cell types (Fig. 2D). A significant increase of the chondrogenic transcripts aggrecan (ACAN), COLL II, and SOX9 was detected in differentiated progeny of hMab and hMAPCs. Significant increased levels of COLL II and Sox9 were also seen in hMSCs that, however, already expressed high levels of ACAN mRNA at day 0 (Supporting Information Figure S1C). Furthermore, pellets of all three populations, at day 21 of chondrogenic differentiation, stained positive for COLL II (Fig. 2E).

thumbnail image

Figure 2. Human MSCs, hMab, and hMAPCs differentiate to adipocytes, osteoblasts, chondrocytes, and smooth muscle cells in vitro, but only hMab have skeletal myogenic differentiation capacity. (A): Oil red staining on adipogenic differentiated hMSCs, hMab, and hMAPCs on day 14 to stain for fat vacuoles (×20, Axiovert 40C, Zeiss). (B): Alizarin red staining on osteogenic differentiated hMSCs, hMab, and hMAPCs on day 28 to stain for calcium deposits (×20, Axiovert 40C, Zeiss). (C): Extraction of the Alizarin red dye, measured at 560 nm (Wallac Victor 1420 plate reader, Perkin Elmer) in osteogenic differentiated hMSCs, hMab, and hMAPCs (three MSC, three MAPC, and two Mab donors). (D): Extraction of the Alcian blue dye, measured at 620 nm (Wallac Victor 1420 plate reader, Perkin Elmer) in chondrogenic differentiated hMSCs, hMab, and hMAPCs on day 21 (three MSC, three MAPC, and two Mab donors). (E): Immunofluorescence staining for COLL II on chondrogenic differentiated hMSCs, hMab, and hMAPCs on day 21. Nuclei were stained with Hoechst (×40, AxioImager, Zeiss). (F): Immunofluorescence staining for αSMA, calponin, and SM-MHC in smooth muscle differentiated hMSCs, hMab, and hMAPCs on day 6. Nuclei were stained with Hoechst (×40 oil, AxioImager, Zeiss). (G): Immunofluorescence staining for MF20 (antisarcomeric myosin) in skeletal muscle differentiated hMSCs, hMab, and hMAPCs on day 10 to detect multinucleated myotubes. Nuclei were stained with Hoechst (×20, Nikon Eclipse Ti-S, Nikon Corporation, Tokyo, Japan). Abbreviations: αSMA, α smooth muscle actin; AU, absorbance units; Mab, mesoangioblasts; MAPC, multipotent adult progenitor cell; MSC, mesenchymal stem cell; SM-MHC, smooth muscle myosin heavy chain.

Download figure to PowerPoint

Following culture with TGF-β1 for 6 days, a similar increase in smooth muscle transcripts α smooth muscle actin (ACTA2), calponin (CNN1), and smooth muscle myosin heavy chain (MYH11) was seen in hMSC, hMab, and hMAPC progeny, reaching a maximum expression on day 4 or 6 (Supporting Information Figure S1D). Expression of all three genes on day 6 was confirmed at the protein level (>90% positive in each group) (Fig. 2F).

When cells were cultured under skeletal myoblast differentiation conditions (no FBS, but 2% horse serum [HS]), only hMab differentiated into skeletal myofibers. A significant increase in transcript levels of MYOD1, myogenin, and dystrophin was seen by RT-qPCR (Supporting Information Figure S1E) and the presence of multinucleated myotubes could be detected by MF20 (antisarcomeric myosin) (Fig. 2G). The fusion index (percentage of nuclei into myotubes over the total nuclei) was 19.2% ± 2.7%, similar to what has been published [14]. Human MAPCs and hMSCs did not differentiate toward skeletal muscle cells using the 2% HS protocol as assessed by RT-qPCR (not shown) or myotube formation (Fig. 2G), or when applying previously described protocols including addition of 5-azacytidine (10 μM) or vascular endothelial growth factor (VEGF), bFGF, and insulin-like growth factor 1 (IGF1) (not shown) [28, 29].

Only hMAPCs Differentiate to Endothelium In Vitro and In Vivo

To induce endothelial differentiation, hMSCs, hMab, and hMSCs were cultured in basal medium containing VEGF165 for 14 days. The expression of six endothelial genes (CD31, CD34, VWF, FLK1, FLT1, TIE2) was measured by RT-qPCR. Compared with undifferentiated hMSCs, transcripts for CD34, FLK1, and FLT1 were significantly increased in VEGF165-treated hMSCs, whereas CD31 mRNA levels were significantly decreased. Similar to hMSCs, transcripts for CD34, FLK1 and FLT1 in VEGF165-treated hMab were significantly increased and CD31 mRNA levels significantly decreased compared with undifferentiated hMab. In addition, VWF transcripts were also increased significantly. On the other hand, treatment of hMAPCs with VEGF165 resulted in a significant increase in the expression of all evaluated endothelial transcripts (Fig. 3A).

thumbnail image

Figure 3. Only human MAPC (hMAPC) differentiate to endothelium in vitro and in vivo. (A): Reverse-transcriptase quantitative polymerase chain reaction analysis for the expression of CD31, CD34, VWF, FLK1, FLT1, and TIE2 in endothelial differentiated human MSCs (hMSCs; blue), human Mab (hMab; green), and human MAPCs (hMAPCs; red) up to day 14 (three donors per group in triplicate). GAPDH was used as a housekeeping gene to normalize mRNA levels and results are shown in Delta Ct (Ctgene of interest − CtGAPDH). Smaller Delta Cts mean higher expression levels. (B): In vitro 2D Matrigel assay of endothelial differentiated hMSCs, hMab, and hMAPCs. Tube formation was examined after 24 hours (×20, Axiovert 40C, Zeiss). (C): Undifferentiated hMSCs, hMab, and hMAPCs were mixed with Matrigel and vascular endothelial growth factor (VEGF165)/basic fibroblast growth factor (bFGF) and droplets plated in vitro (3D Matrigel assay). Tube formation was examined at day 5 (×5, Axiovert 40C, Zeiss). (D): Undifferentiated hMSCs, hMab, and hMAPCs were mixed with Matrigel and VEGF165/bFGF and transplanted under the skin of nude mice. After 21 days, Matrigel plugs were removed and examined macroscopically and microscopically (upper and middle row; ×2.5 and ×6.6, respectively; Lumar dissection microscope, Zeiss). Human MSC- and hMab-containing Matrigels featured vessel leakage, indicated with white arrows, which could also be seen on H&E-stained cross sections (bottom row; ×40 AxioImager, Zeiss). (E): The ingrowth of host vessels in the Matrigel plugs of each group presented as the percentage of mouse CD31+ area. (F): Human CD34+ cells (indicated with white arrows) were detected only in hMAPC-containing Matrigel plugs (×40, Axioplan, Zeiss). (G): The hMab-containing Matrigel plugs contained desmin+ cells (indicated with white arrows), suggestive for muscle differentiation (×40, AxioImager, Zeiss). Abbreviations: Mab, mesoangioblasts; MAPC, multipotent adult progenitor cell; MSC, mesenchymal stem cell.

Download figure to PowerPoint

When VEGF165-treated hMSCs, hMab, and hMAPCs were plated onto cytokine-depleted Matrigel (2D Matrigel assay), only hMAPCs demonstrated clear tube formation when examined after 24 hours (Fig. 3B). Upon plating undifferentiated cells as 20 μl drops onto cytokine-depleted Matrigel (3D Matrigel assay), we could only observe tube formation after 5 days in hMAPC-containing Matrigel droplets (Fig. 3C). Although hMab also formed cellular networks on day 5, the cells that grew out of the Matrigel droplets stained positive for desmin (a muscle protein), suggesting that not vascular tubes but myotubes were formed (not shown).

An in vivo Matrigel plug assay was performed to study the ability of the cells to attract host vessels and to directly contribute to new vessel formation. Undifferentiated hMSCs, hMab, or hMAPCs were mixed with Matrigel supplemented with VEGF165 and bFGF, and transplanted under the skin of nude mice. As a control, PBS instead of cells was mixed with Matrigel and growth factors. After 3 weeks, the Matrigel plugs were dissected out and analyzed macroscopically and microscopically. Matrigel plugs containing hMAPCs harbored numerous functional vessels, which contained red blood cells, none of which were leaky (0/6 Matrigel plugs). In contrast, Matrigel plugs admixed with hMSCs or hMab harbored leaky vessels (3/6 for MSCs, 4/6 for Mab) as evidenced by the presence of red blood cell pools in the Matrigel structures. These pools could also be seen on H&E-stained sections (Fig. 3D). Quantification of the ingrowth of host vessels in the Matrigel plugs revealed that hMSCs and hMAPCs caused a significantly greater ingrowth of mouse (m)CD31+ vessels compared with the control (p < .001 in both groups), whereas the number of mouse vessels was not increased in hMab containing Matrigel plugs compared with the control (p = .287). In addition, significantly more mCD31+ vessels were found in Matrigel plugs with hMAPCs compared with hMSCs and hMab (p < .05 in both groups) and the number of mCD31+ vessels was also significantly higher in hMSCs- compared with hMab-containing Matrigel plugs (p = .0057) (Fig. 3E). Finally, the presence of human endothelium was assessed by staining for the human-specific endothelial cell surface antigen CD34. Human CD34+ vessels were only detected in hMAPCs-containing Matrigel plugs (Fig. 3F). When the Matrigel plugs were stained for desmin, a positive signal was only detected in hMab-containing Matrigel plugs, showing again their potential to differentiate toward skeletal muscle (Fig. 3G).

Transcriptome Analysis Reveals That hMSCs, hMab, and hMAPCs Are Three Distinct Populations

Given the phenotypic and functional differences between hMSCs, hMab and hMAPCs, we investigated differences in transcriptome using microarrays. Besides these three cell populations, hESCs were also included. An unsupervised analysis was performed using all probes present in at least two of the samples (irrespective of the population). Principal component analysis (PCA) was used to reduce dimensionality of all the summarized present probes and represent it as a linear combination of two main orthogonal variables (principal components, PCs). Plotting samples in the two PC-reduced dimensional space, which captured ∼59% variability in the dataset, indicated that hMSCs, hMab, and hMAPCs cluster apart from pluripotent hESCs. This could be observed in the first PC, which captured the largest variability. In the second PC, distinct clusters could be seen consisting of the hMAPC group and the hMSC-hMab group (Fig. 4A). Pairwise supervised comparison of hESC with each of the other three cell populations yielded 6,250 differentially expressed genes (Supporting Information Table S4). Genes such as OCT4, NANOG, and SOX2, which are core transcription factors for maintaining pluripotency and self-renewal in ESCs [30], were at least 1,000-fold enriched in hESCs compared with hMSCs, hMab, and hMAPCs.

thumbnail image

Figure 4. Transcriptome analysis reveals that hMSCs, hMab, and hMAPCs are three distinct populations and different from pluripotent ESCs. (A): Unsupervised principal component analysis (PCA) on all present probes in ESCs (open circles), hMSCs (triangles), hMab (squares), and hMAPCs (filled circles) with the samples plotted in the first two components' space. (B): Unsupervised PCA on all present probes in hMSCs (triangles), hMab (squares), and hMAPCs (circles) with the samples plotted in the first two components' space. The data discussed in this publication have been deposited in NCBI's Gene Expression Omnibus and are accessible through GEO Series accession number GSE23131. (C): Table with transcripts from over-represented pathways and biological processes in each group. Transcripts shown are at least twofold differentially expressed. Abbreviations: hMab, human mesoangioblasts; hMAPC, human multipotent adult progenitor cell; hMSC, human mesenchymal stem cell.

Download figure to PowerPoint

An unsupervised analysis of the three cell populations (hMAPCs, hMSCs, and hMab) without hESCs, revealed that hMAPCs clustered separately from hMSCs as well as from hMab (Fig. 4B). The total variance captured by the first 2 PCs was ∼54%. We next identified the genes that were at least twofold differentially expressed in one group compared with any of the other two groups. A total of 7,576 differentially expressed genes were found (including 1,955 unannotated probes). In hMAPCs, 827 genes were more highly expressed and 863 genes were expressed at lower levels compared with hMSCs and hMab. In hMSCs, 665 genes were higher and 341 genes were lower expressed and in hMab, 329 genes were higher and 364 genes were lower expressed compared with the other two cell populations (Supporting Information Table S5). Ingenuity pathway analysis was used to gain insight into biological functions and pathways represented by the differentially expressed transcripts in each group. In hMAPCs, genes that promote differentiation of endothelial cells and increase angiogenesis or blood vessel maturation were over-represented (summarized in Fig. 4C). Genes involved in the differentiation of chondrocytes and osteocytes were over-represented in hMSCs and hMab, as well as transmembrane molecules from the antigen-presenting pathway. In hMSCs, genes involved in the organization of collagen fibrils and filaments were over-represented, as well as genes involved in the development and contraction of smooth muscle and genes important for neovascularization (Fig. 4C). Human Mab expressed higher levels of genes involved in the development of skeletal muscle and genes that promote cardiogenesis (Fig. 4C).

Differences in Phenotype, Proliferation, Differentiation Ability, and Gene Expression Profile of hMSCs, hMab, and hMAPCs Are Partially due to Differences in Culture Conditions

To determine whether the differences in culture conditions (Fig. 1A) may be responsible for the differences seen between hMSCs, hMab, and hMAPCs, we performed an experiment in which cells isolated under one of the three culture conditions were switched to a different culture condition. Isolations were done under normal conditions and hMSCs and hMab were switched to MAPC conditions after 5–10 passages and 10–15 passages, respectively (MSC_to_MAPC and Mab_to_MAPC). Likewise, hMAPCs were cultured under MSC and Mab conditions after 15–25 passages (MAPC_to_MSC and MAPC_to_Mab). The switched cultures were maintained in the new conditions for at least 3–4 additional passages (10–14 days) and cell morphology, proliferation, phenotype, and differentiation capacity were re-examined. In addition, the expressed gene profile of the cells maintained under the switched culture conditions was reassessed.

When hMAPCs were cultured under MSC conditions (MAPC_to_MSC), they became long and elongated, a morphology typically seen for hMSCs, whereas MSC_to_MAPC became small cells with a ruffled border, the same as was seen for hMAPCs (Fig. 5A, 5B). It is noteworthy to mention that change in morphology was observed after 2–3 passages in almost all cells, suggesting change in phenotype rather than selection of a cell subpopulation by the culture switch. Likewise, hMab cultured under MAPC conditions (Mab_to_MAPC) acquired hMAPC morphology. Although some hMAPCs cultured under Mab conditions (MAPC_to_Mab) became triangular, they remained adherent whereas 50% of hMab are commonly nonadherent (Fig. 5C, 5D). The proliferative potential of the cells was affected by the secondary culture conditions. MAPC_to_MSC and MAPC_to_Mab cultures could only be expanded for 10–15 additional PD and did not reach 60–70 PD. By contrast, MSC_to_MAPC and Mab_to_MAPC could now be expanded beyond 60PD (Fig. 5E–5H). There were also phenotypic changes induced by plating the cells under different culture conditions. MAPC_to_MSC became CD140alow, CD140b+, and MHC class I+ and MSC_to_MAPC became CD140a, CD140b and MHC class Ilow (Fig. 5I). Likewise, MAPC_to_Mab became CD140alow, CD140blow and MHC class I+ and Mab_to_MAPC became CD140a, CD140b, and MHC class Ilow (Fig. 5J). NG2 and MHC class II expression was lost in Mab_to_MAPC and acquired on MAPC_to_Mab, although expression of MHC class II was very low on MAPCs cultured under Mab conditions. No changes in ALP expression were seen when culture conditions were switched (Fig. 5I, 5J).

thumbnail image

Figure 5. Switch in culture conditions affects cell morphology, proliferation, and surface phenotype. (A, C): Morphology of hMAPCs after culture under MSC and Mab conditions. (B, D): Morphology of hMSCs and hMab after culture under MAPC conditions (×20, Axiovert 40C, Zeiss). (E–H): Expansion curve for hMAPCs, hMSCs, hMab, MAPC_to_MSC, MAPC_to_Mab, MSC_to_MAPC, and Mab_to_MAPC. (I, J): Flowcytometry for hMAPCs, hMSCs, hMab, MAPC_to_MSC, MAPC_to_Mab, MSC_to_MAPC, and Mab_to_MAPC. Percentage of expression for each antigen is indicated. The isotype control is shown in gray and the specific antigen in black (representative example of three MSC_to_MAPC, three MAPC_to_MSC, two Mab_to_MAPC, and two MAPC_to_Mab cultures). Abbreviations: ALP, alkaline phosphatase; Mab, human mesoangioblasts; MAPC, human multipotent adult progenitor cell; MHC, major histocompatibility complex; MSC, human mesenchymal stem cell; NG2, neuron/glial type2 antigen.

Download figure to PowerPoint

The differentiation ability was also affected by the secondary culture conditions. Endothelial differentiation capacity was lost in MAPC_to_MSC as shown by RT-qPCR (only CD34, FLK1, and FLT1 were significantly induced) and lack of vascular tube formation in a 2D Matrigel assay (Fig. 6A, 6F). By contrast, MSC_to_MAPC gained endothelial differentiation ability as a significant upregulation of five endothelial transcripts (CD34, VWF, FLK1, FLT1, TIE2) was seen upon differentiation and MSC_to_MAPC could form vascular tubes (Fig. 6B, 6G). Likewise, a significant increase of five endothelial transcripts (CD34, VWF, FLK1, FLT1, TIE2) was seen in Mab_to_MAPC, as well as the formation of vascular tubes (Fig. 6D, 6K). Interestingly, exposure of hMAPCs to Mab conditions did not affect the endothelial differentiation capacity, as a significant increase in expression of five endothelial transcripts (except for CD31) was seen and cells continued to form vascular tubes (Fig. 6C, 6 J). By contrast, MAPC_to_MSC lost endothelial differentiation capacity (Fig. 6A, 6F). Next, we assessed the myogenic potential of MAPC_to_Mab and Mab_to_MAPC using the serum starvation protocol with 2% HS. After 10 days, there was a clear decrease in myogenic differentiation potential of Mab_to_MAPC, as MYOD1 transcript levels were no longer increased and the presence of multinucleated tubes was significantly reduced, quantified via the fusion index (19.8% ± 2.73% vs. 3.95% ± 0.61%) (Fig. 6M–6P). However, MAPC_to_Mab did not acquire myogenic differentiation potential (not shown).

thumbnail image

Figure 6. Effect of changing culture conditions on the differentiation capacity. (A–D): Reverse-transcriptase quantitative polymerase chain reaction analysis of six endothelial transcripts (CD31, CD34, VWF, FLK1, FLT1, TIE2) on hMAPCs, MAPC_to_MSC, MSC_to_MAPC, hMSCs, MAPC_to_Mab, hMab_to_MAPC, and hMab differentiated toward endothelial cells up to 14 days (two donors in triplicate). GAPDH was used as a housekeeping gene to normalize mRNA levels and results are shown in Delta Ct (Ctgene of interest − CtGAPDH). Smaller Delta Cts mean higher expression levels. (E–L): In vitro 2D Matrigel assay of endothelial differentiated hMSCs, hMab, hMAPCs, and the switched cultures. Tube formation was examined after 24–48 hours (×20, Axiovert 40C, Zeiss). (M): RT-qPCR analysis of the skeletal genes MYOD1, myogenin and dystrophin in hMab and Mab_to_MAPC differentiated toward skeletal muscle cells (two donors in triplicate). (N, O): Immunofluoresence staining for MF20 (antisarcomeric myosin) in skeletal muscle differentiated hMab and Mab_to_MAPC on day 10 to detect multinucleated myotubes. Nuclei were stained with Hoechst (×10, Nikon Eclipse Ti-S). (P): Fusion index for hMab and Mab_to_MAPC, calculated as the percentage of fused nuclei over the total nuclei. Abbreviations: Mab, mesoangioblasts; MAPC, multipotent adult progenitor cell; MSC, mesenchymal stem cell.

Download figure to PowerPoint

Finally, we performed additional transcriptome studies to determine whether the global gene expression pattern was also changed upon switching culture conditions. An unsupervised PCA revealed that MSC_to_MAPC and Mab_to_MAPC became significantly closer to hMAPCs in the first PC, which captured the largest variability (36.3% and 32%, respectively) indicating that their global expressed gene profile became more similar to that of hMAPCs than the original cell type. Likewise, when hMAPCs were switched to either MSC or Mab culture conditions, their expressed gene profile moved away from that of hMAPCs and became more similar to that of hMSCs or hMab, respectively (Fig. 7A, 7B). As expected from the global analysis of gene expression, the lists of differentially expressed genes between the original culture and the switched condition revealed that genes determined as “characteristic” for each cell type (Fig. 4C) became upregulated or downregulated under the switched condition (Fig. 7C, 7D). For example, when comparing hMSCs versus MSC_to_MAPC, the latter acquired expression of ANGPTL4 and F3 endothelial genes, which have been described as characteristic for hMAPCs. Likewise, Mab_to_MAPC acquired expression of ANGPT1, ANGPTL4, EFNB2, IL8, FGF5, CAV1, and CAV2 endothelial genes when compared with hMab. Furthermore, analysis of MAPC_to_Mab versus hMAPCs demonstrated that the former did not acquire expression of skeletal genes MYF5 or MLK2, in line with the observation that they did not form skeletal myofibers in functional assays. The observation that not all genes were reversed in MAPC_to_Mab is also reflected by the differences observed in PC2 between hMab and MAPC_to_Mab conditions (Fig. 7B). The findings from the microarray analysis were corroborated by RT-qPCR analysis for some of the characteristic genes (Supporting Information Figure S2A, S2B).

thumbnail image

Figure 7. Global transcriptome changes in the switched culture conditions. (A): Unsupervised principal component analysis (PCA) on all present probes in hMSCs (filled triangles), hMAPCs (filled circles), MAPC_to_MSC (open circles), and MSC_to_MAPC (open triangles) with the samples plotted in the first two components' space. A technical replicate was included for MSCs and MAPCs, where the arrow leaves from two data points and ends in one data point. (B): Unsupervised PCA on all present probes in hMab (filled squares), hMAPCs (filled circles), MAPC_to_Mab (open circles), and Mab_to_MAPC (open squares) with the samples plotted in the first two components' space. (C): Table with transcripts higher or lower expressed in MAPC_to_MSC compared with hMAPCs and in MSC_to_MAPC compared with hMSCs. (D): Table with transcripts higher or lower expressed in MAPC_to_Mab compared with hMAPCs and in Mab_to_MAPC compared with hMab. Transcripts shown are at least twofold differentially expressed. Abbreviations: Mab, mesoangioblasts; MAPC, multipotent adult progenitor cell; MSC, mesenchymal stem cell; PC, principal components.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

We compared three populations of adherent postnatal stem cells (skeletal muscle-derived hMab and BM-derived MSCs and MAPCs) with respect to their molecular and functional characteristics. Several independent donors per cell type and uniform protocols were used for the characterization of phenotype and function, allowing direct comparison of the three cell populations. The phenotype and function of hMSCs, hMab and hMAPCs was also assessed following a minimum of 3–4 passages (10–14 days) under different culture conditions (MSC_to_MAPC, Mab_to_MAPC, MAPC_to_MSC and MAPC_to_Mab).

Human MAPCs could be expanded for a significantly longer time than hMSCs and hMab. However, when hMAPCs were cultured under Mab or MSC conditions, cell proliferation was significantly reduced, whereas this was increased for hMab and hMSCs cultured under MAPC conditions, as they could be expanded beyond 60PD. Human MSCs, hMab, and hMAPCs were phenotypically distinct and the cell surface marker CD140b could discriminate among the three cell populations, being highly expressed on hMSCs, expressed at low levels on hMab and not expressed on hMAPCs. The latter is noteworthy as hMAPCs, but not hMSCs and hMab, are maintained in culture with PDGF-BB, which binds to the PDGF receptors. The transcriptome data revealed that levels of CD140b mRNA in hMAPCs are as high as in hMab suggesting that the CD140b receptor might be internalized in hMAPCs. Consistent with this notion, when hMAPCs were cultured under Mab or MSC conditions they expressed CD140b, whereas this expression was lost in hMab and hMSCs cultured under MAPC conditions.

The three cell types differentiated in a similar way toward typical mesenchymal cell types, including adipocytes, osteoblasts, and chondrocytes in vitro. Also smooth muscle differentiation in vitro was observed in all three populations, as described in previous studies [14, 27, 31]. Only hMab could differentiate into skeletal myocytes, whereas hMAPCs and hMSCs could not be committed to the skeletal muscle lineage, even when treated with the demethylating agent 5-azacytidine or in the presence of VEGF, bFGF, and IGF1, which have been shown to induce myogenic differentiation of MSCs and/or MAPCs [28, 29]. Of note, changing hMab to MAPC conditions resulted in a decrease in the ability to form myofibers. However, when hMAPCs were cultured under Mab conditions, no myogenic ability was acquired.

Furthermore, only hMAPCs differentiated to functional endothelium in vitro and in vivo. Even though some studies have suggested that endothelial gene expression and tube formation can be induced from hMSCs in vitro [32], others have demonstrated that MSCs form pericyte-like cells supporting blood vessels, not the endothelial lining itself [33]. Recently, Delorme et al. found that hMSCs may be primed for endothelial differentiation but cannot be induced down this pathway by exposure to VEGF [34]. We show here that when hMSCs and hMab were cultured under MAPC conditions, they acquired the ability to form vascular tubes in a 2D Matrigel assay and a significant increase in endothelial transcripts was seen. By contrast, MAPC_to_MSC lost their endothelial differentiation capacity but this remained when hMAPCs were cultured under Mab conditions.

In an in vivo Matrigel plug assay, functional vessel formation was only seen in the hMAPC-containing Matrigel plugs, whereas Matrigel plugs with hMSCs and hMab harbored leaky vessels. The ingrowth of host vessels was the highest in hMAPC-containing Matrigels and only here could human CD34+ vessels be found, consistent with the in vitro differentiation capacity of hMAPCs and earlier studies where hMAPCs were grafted in ischemic limbs of nude mice [25]. In hMab-containing Matrigel plugs, desmin-positive cells were found, suggesting that like in vitro, hMab may have undergone myogenic differentiation.

The expressed gene profile of hMAPCs, as well as hMSCs and hMab, differed significantly from pluripotent hESCs, consistent with a recent publication by Aranda et al. [35]. We also showed that hMSCs, hMab, and hMAPCs are three distinct populations with different phenotypic and functional characteristics, which is reflected in their transcriptome. Although hMSCs and hMAPCs were isolated from BM and hMab from skeletal muscle, the expressed gene profile of hMSCs was more similar to that of hMab than to that of hMAPCs. This may reflect the recent finding that hMSCs and hMab may both be subpopulations of pericytes in vivo [11, 12, 14, 36, 37]. Even though, some studies have suggested that the tissue of origin does not affect the functional properties of cultured pericytes [11], our studies demonstrate that the functional characteristics of BM-derived hMSCs and muscle-derived hMab differ and that this is also reflected in their transcriptome.

It should be noted that for Mab and MAPC cell lines, differences exist between species. Mab derived from the dorsal aorta in mouse have the ability to generate endothelium, aside from mesenchymal progeny and myocytes, suggesting that mouse Mab represent mesodermal progenitors [38]. By contrast, our studies here show that Mab derived from human muscle are capable of differentiation to mesenchymal and myoblast cells, but only form endothelium when cultured under MAPC conditions.

Likewise, differences exist between hMAPCs and rodent MAPCs. We previously published that rodent MAPCs expressing significant levels of Oct4 (but not Nanog and Sox2) and a number of primitive endoderm genes, had a much more robust ability to generate hepatocyte-like cells [39, 40]. As was shown for rodent cells isolated under MAPC conditions, that express low levels of Oct4, treatment of hMAPCs with the hepatocyte induction protocol resulted in a small increase in hepatic transcripts, to levels similar as described in Schwartz et al. [24] for hMAPCs, but significantly less than when high Oct4 rodent MAPCs are directed to the hepatic lineage [39].

In line with the functional studies described above, the global transcriptome of hMSCs, hMab, and hMAPCs was also affected by the culture conditions. Following a switch in culture conditions, the expressed gene profile shifted from the cell of origin toward that of the cells isolated in the culture conditions to which they were switched to. This finding suggests that well-defined sets of gene transcripts may help to characterize and infer the functional properties of cell populations. How different serum concentrations, growth factors, hypoxic conditions, or seeding densities influence these characteristics will need to be further determined. Several reports showed that when hMSCs are initially plated and cultured at low density, their expansion, and differentiation capacity is greatly enhanced [41, 42]. Likewise, adding bFGF to the medium or culture under hypoxic conditions, increases the proliferation rate of hMSCs [42–46]. How these factors affect the functional capacity of cultured cells will need to be further determined in clonal studies, as we here compared populations of cells that do not necessarily equate with the intrinsic properties of individual cells. Such clonal studies might also clarify if there is a hierarchy among the cells.

CONCLUSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

In conclusion, our study has identified clear phenotypic and in vitro and in vivo functional differences between cultured hMSCs, hMab, and hMAPCs, which are associated with their specific expressed gene profiles. However, this study also demonstrates that the phenotypic and functional properties, as well as the expressed gene profile are partially influenced by changes in the culture conditions.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

We thank Nele Peersman, Lotte Vanbrabant, and Ellen Konings for the excellent technical assistance with cell culture and RT-qPCR. We also thank Kristel Eggermont for the help with fluorescence microscopy. We are thankful to Dr. Anja Van Campenhout for providing us with the bone and skeletal muscle fragments. We thank Kris Van Den Bogaert for her critical review of this article to assure accuracy of the data. This work was supported by the Center of Excellence funding K.U.Leuven, an Odysseus award, research funding from Athersys Inc., and a grant from the European Commission (EC-FP6-STREP-STROKEMAP; to C.M.V.); the Center of Excellence funding K.U.Leuven (EF/05/13) (to A.L.); a CAF-DCF chair for stem cell research (to M.D.), and a OT (ETH-C0420-OT/09/053) and GOA (EME-C2161-GOA/11/012) grant from the K.U.Leuven (to C.M.V., M.S., and A.L.). V.D.R. is funded by a grant from the Institute for the Promotion of Innovation through Science and Technology in Flanders (IWT Vlaanderen). S.C. is a research assistant of the Flemish Fund for Scientific Research (FWO Vlaanderen). S.W.V.G. and M.D. are senior clinical investigators of FWO Vlaanderen.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
STEM_633_sm_suppinfoFigureS1.tif10038KFigure S1: RT-qPCR analysis for hMSC, hMab and hMAPC differentiated towards adipocytes, osteoblasts, chondrocytes, smooth and skeletal muscle cells. (A) Expression of PPAR?2 in adipogenic differentiated hMSC, hMab and hMAPC up to day 14 (3 donors per group in duplicate). GAPDH was used as a housekeeping gene to normalize mRNA levels and results are shown in Delta Ct (Ct gene of interest - Ct GAPDH). Smaller Delta Cts mean higher expression levels (B) Expression of OC and ALP in osteogenic differentiated hMSC, hMab and hMAPC up to day 28 (3 (MSC, MAPC) or 2 (Mab) donors per group in duplicate). (C) Expression of ACAN, COLL II and SOX9 in chondrogenic differentiated hMSC, hMab and hMAPC up to day 21 (3 (MSC) or 2 (Mab, MAPC) donors per group in duplicate). (D) Expression of αSMA, calponin and SM-MHC in smooth muscle differentiated hMSC, hMab and hMAPC up to day 6 (2 donors per group in triplicate). (E) Expression of MYOD1, myogenin and dystrophin in skeletal muscle differentiated hMab on day 10 (2 donors in triplicate).
STEM_633_sm_suppinfoFigureS2.tif5680KFigure S2: Gene expression changes in switched culture conditions. (A) RT-qPCR analysis of ANGPTL4, EFNB1, calponin and GATA6 in hMAPC, hMSC, MAPC_to_MSC and MSC_to_MAPC (2 donors in duplicate). (B) RT-qPCR analysis of ANGPTL4 and MYF5 in hMAPC, hMab, MAPC_to_Mab and Mab_to_MAPC (2 donors in duplicate). GAPDH was used as housekeeping gene to normalize mRNA levels and results are shown in Delta Ct (Ct gene of interest - Ct GAPDH).
STEM_633_sm_suppinfo.doc107KSupporting Information

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.