Stem Cell Research (2012) 9, 87–100

Ava i l ab l e on l i ne a t www.sc i enced i r ec t . com

www.e l sev i e r . com/ loca te / sc r

REGULAR ARTICLE

One-step derivation of cardiomyocytes andmesenchymal stem cells from humanpluripotent stem cellsHeming Wei a, Grace Tan a, Manasi a, Suhua Qiu a, Geraldine Kong a,Pearly Yong a, Caihong Koh a, Ting Huay Ooi a, Sze Yun Lim a, Philip Wong a, c,Shu Uin Gan b, Winston Shim a, c,⁎

a Research and Development Unit, National Heart Centre Singapore, Singaporeb Department of Surgery, Yong Loo Lin School of Medicine, National University of Singapore, Singaporec Graduate Medical School, DUKE-NUS, Singapore

Received 19 November 2011; received in revised form 14 April 2012; accepted 16 April 2012Available online 24 April 2012

Abstract Cardiomyocytes (CMs) and mesenchymal stem cells (MSCs) are important cell types for cardiac repair postmyocardial infarction. Here we proved that both CMs and MSCs can be simultaneously generated from human inducedpluripotent stem cells (hiPSCs) via a pro-mesoderm differentiation strategy. Two hiPSC lines, hiPSC (1) and hiPSC (2) weregenerated from human dermal fibroblasts using OCT-4, SOX-2, KLF-4, c-Myc via retroviral-based reprogramming. H9 humanembryonic stem cells (hESCs) served as control. CMs and MSCs were co-generated from hiPSCs and hESCs via embryoid body-dependent cardiac differentiation protocol involving a serum-free and insulin-depleted medium containing a p38 MAPKinhibitor, SB 203580. Comparing to bone marrow and umbilical cord blood-derived MSCs, hiPSC-derived MSCs (iMSCs) expressedcommon MSC markers and were capable of adipogenesis, osteogenesis and chondrogenesis. Moreover, iMSCs continuouslyproliferated for more than 32 population doublings without cellular senescence and showed superior pro-angiogenic and woundhealing properties. In summary, we generated a large number of homogenous MSCs in conjunction with CMs in a low-cost andefficient one step manner. Functionally competent CMs and MSCs co-generated from hiPSCs may be useful for autologouscardiac repair.

© 2012 Elsevier B.V. All rights reserved.

Abbreviations: hPSCs, human pluripotent stem cells; hiPSCs, human induced pluripotent stem cells; hESCs, human embryonic stem cells;CMs, cardiomyocytes; hiPSC-CMs, human induced pluripotent stem cell-derived cardiomyocytes; MSCs, mesenchymal stem cells; iMSCs, hiPSC-derived MSCs; EBs, Embryoid bodies; CARM, Cardiomyogenic medium.⁎ Corresponding author at: Research and Development Unit, National Heart Centre, 9 Hospital Drive, School of Nursing, #03-02, Block C,

SingHealth Research Facilities, 169612, Singapore. Fax: +65 62263972.E-mail address: winston.shim.s.n@nhcs.com.sg (W. Shim).

1873-5061/$ - see front matter © 2012 Elsevier B.V. All rights reserved.doi:10.1016/j.scr.2012.04.003

88 H. Wei et al.

Cardiac cell therapy requires not only functional cardiomyo-cytes (CMs), but also pro-angiogenic cell lineages such asmesenchymal stem/stromal cells (MSCs) for effective cardiacrepair. However, clinical application of CMs has been hamperedby the lack of autologous and large quantity of functionalhuman CMs. On the other hand, MSCs derived from adult tissuessuch as bonemarrow have shown therapeutic potential (Amadoet al., 2005; Schuleri et al., 2009). However, clinicalapplication with adult MSCs from the elderly (Fan et al., 2010;Kasper et al., 2009; Yu et al., 2011) and patients (Heeschen etal., 2004; Nie et al., 2010) are limited due to their poor self-renewal potential and functional competency. Therefore, analternative source of MSCs is imperative for successfultherapeutic applications together with CMs in autologouscardiac cell therapy.

Human pluripotent stem cells (hPSCs), including humanembryonic stem cells (hESCs) and human induced pluripotentstem cells (hiPSCs) are capable of differentiating into both CMsand MSCs. However, although hESCs are capable of generatingfunctional CMs and MSCs, their wide-spread clinical applica-tions have been hampered by immunoreactivity and ethicalissues. The introduction of iPSCs by Yamanaka and colleagues(Takahashi and Yamanaka, 2006) may provide a more tenablesolution. Patient-specific iPSCs have been generated fromsomatic cells of patients with various ages and diseases (Park etal., 2008; Takahashi et al., 2007; Yu et al., 2007). Like ES cells,iPSCs can be preserved and repeatedly used to generate manycell lineages such as CMs (Xu et al., 2002; Zhang et al., 2009)and MSCs (Lian et al., 2007, 2010) without reduced capacity asnormally expected in adult tissue derived cells.

Currently, hPSC-CMs and hPSC-MSCs are separately derivedfrom disparate mesodermal differentiation protocols whichare tedious and time-consuming.We hypothesize that CMs andMSCs, both of mesoderm origin; can be generated concurrent-ly from hPSCs via embryoid body (EB) formation with a pro-mesodermal condition. As cardiac differentiation indicates amesodermal diversion from pluripotency of hiPSCs, MSCs co-obtained could be less of tumorigenic concern, which makesthis MSC generation procedure simple, low-cost and efficient.

In this study, with a modified pro-mesodermal differentia-tion protocol, we repeatedly co-generated high quality of CMsand MSCs from individual EBs derived from hiPSCs. Both celltypes showed functional characteristics that are promising forautologous application in cardiac cell therapy.

Results

Generation and characterization of hiPSCs

Concordance with our previous report of virus-free iPSC(Mehta et al., 2011), the current viral approach ofreprogramming dermal fibroblasts gave rise to iPSC colonieswith typical ESC-like morphology in approximately 18 dayspost retroviral transduction. Two hiPSC clones, hiPSC (1) andhiPSC (2) were selected for this study. Their doubling rateswere comparable to the similarly maintained H9 hESC cellline. Further, immunocytochemistry staining confirmed thatall hiPSC clones expressed pluripotent stem cell markers,OCT4, SSEA4, TRA1-60 and TRA-1-81 (Fig. 1a). Moreover,their endogenous expression of OCT-4, SOX-2, KLF-4 and c-Myc were validated by RT-PCR (Fig. 1b). Those hiPSC clones

were confirmed to carry the genome incorporation of thetransgenes in retroviral vectors but do not express those 4transgenes (Supplementary Fig. 1). Further, the pluripo-tency of hiPSCs were confirmed by teratoma formation invivo with phenotypes of three germ layers consisting gutepithelium (endoderm), smooth muscle (mesoderm) andneuroepithelial rosette (ectoderm) following renal capsuleimplant in SCID mice at week 12 (Fig. 1c). Both hiPSC (1) andhiPSC (2) showed normal karyotype that was similar to thewell characterized H9 hESC line (Fig. 1d).

Co-generation of CMs and MSCs from hiPSCs

EBs were created from undifferentiated colonies of hiPSCs (1)and (2) and H9 hESCs and differentiated towards cardiomyo-cytes following a modified cardiac differentiation protocolwhich involves cardiomyogenic medium (CARM) plus a specificp38-MAPK inhibitor SB 203580 (hereby named CARM/SBprotocol). After 15 days of cardiac differentiation, ~50% ofEBs survived and were plated between 15 and 21 days. Both ofthe contracting (containing functional cardiomyocytes) andnon-contracting EBs gave raise to homogenous outgrowth ofMSC-like cell population that rapidly migrated from theattached EBs. With the contracting EBs, the beating cellclusters were microdissected out to isolate cardiomyocytes.Fig. 2A presents a graphical demonstration for generating CMs(Fig. 2Ae) and MSCs (Fig. 2Af) following the CARM/SB protocol.To test parallel presence of both CMs and MSCs in EB and theirspatial relationship, immunoflurescent (IF) staining of Day 15EBs showed that MSCs present in all EBs (Figs. 2Ba and Bb)while CMs were clearly presented only in the contracting EBs(Fig. 2Ba) but not non-contracting EBs. CMs and MSCs do notco-localize (Fig. 2Ba).

In addition, the percentages of CMs and MSCs in contractingEBs were quantified (Fig. 2C). First, and the percentage of CMsand MSCs in contracting EBs were estimated by immunofluo-rescent staining dissociated cells with cardiac specificmarkers(Myosin light chain: MLC) and MSC marker (CD44), counter-stained with 4′,6-diamidino-2-phenylindole (DAPI) (Figs. 2a,b). The percentage of CMs were 22.03% while MSCs werearound 45.72%. Second, the percentage of MSCs weremeasured by Flow Cytometry assay of cells dissociated fromcontracting and non-contracting EBs. The markers varied from59% (CD105) to 78% (CD90) and 81% (CD29). The variation inthe proportion of different MSC surface markers determinedwith cells dissociated from EBs may reflect the heterogeneityof cell lineages in EBs as some markers may express in non-MSCs.

Following the CARM/SB protocol, 3 lines of iMSCs weregenerated from each line of hPSCs in three independentdifferentiation experiments.

hiPSC-CMs showed structural and functionalcharacteristics of cardiomyocytes

All cardiomyocytes tested in this study were ~5 weeks postdifferentiation. Cardiomyocytes dissociated from contractingcell clusters showed specific cardiacmarkers including cardiacsarcomeric α-actinin, β-myosin heavy chain (β-MHC) and Titin(Fig. 3A). There is no appreciable difference in proportion ofthree subtypes of cardiomyocytes including ventricular-like

A

C

D

B

hESC (H9) hiPSCs (2)hiPSCs (1)

Mesoderm EctodermEndoderm

1 32

H9 i1 i2 HDF WC

OCT-4

SOX-2

c-Myc

KLF-4

β-actin

i1: hiPSC (1), i2: hiPSC (2).HDF: human dermal fibroblasts

Oct-4 DAPI Overlay

SSEA-4 DAPI Overlay

Tra-1-60 DAPI Overlay

Tra-1-81 DAPI Overlay

Figure 1 Characterization of hiPSCs. A: Immunofluorescent staining of iPSC (1) with antibodies against Oct-4, SSEA-4 (63×objective, scale bars: 20 µm), Tra-1-60, and Tra-1-81 (40× objective, scale bars: 100 µm). B: RT-PCR for the expression of endogenousgenes OCT-4, SCX-2, c-Myc and KLF-4 in iPSC (1) and hiPSC (2) with H9 hESCs and the parental neonatal dermal fibroblast (FB) ascontrol. C: Hematoxylin and eosin staining of teratoma derived from iPSC (1) in NOD-SCID mice after 12 weeks. Characteristic tissuestructures representing 3-germ layers (indicated by arrows) are: glandular structure for endoderm (C1), smooth muscle tissue formesoderm (C2) and neural epithelia for ectoderm (C3), scale bar: 200 µM. D: Karyotype of H9 hESC, hiPSC (1) and hiPSC (2).

89One-step derivation of cardiomyocytes and mesenchymal stem cells from human pluripotent stem cells

(V), atrial-like (A), and nodal-like (N) cardiomyocytes betweenhiPSC-CMs and hESC-CMs (Fig. 3C) determined by character-istic action potential traces (Fig. 3B). It was also noted, asshown in Fig. 3D, that the cardiac differentiation efficiency(measured as the % of contracting EBs) for hiPSC (1) and (2)(20–30%) was lower than hESCs (50%).

iMSCs meet the criteria for human MSCs with robustself renewal potential and superior biological function

MSC lines derived from hPSCs were subjected to charac-terizations. MSCs derived from the contracting EBs of hiPSC(1) and hiPSC (2) and hESC (H9) are named as iMSC (1),

90 H. Wei et al.

iMSC (2) and H9 hESC-MSCs respectively. All experimentswere repeated at least for three times. The data of iMSC(1) is presented in most of the figures. Human bone marrowMSCs (BM-MSCs) and umbilical cord blood MSCs (CB-MSCs)were adopted as control in some of the characterizationstudies.

Figure 2 Co-generation of CMs and MSCs from hiPSC by a modifiefrom hiPSC by a modified CARM/SB method. Top– flow chart of thegraphical demonstration of the process for co-generating hiPSC-CMs(b): EBs formed in suspension culture on Day 2 after differentiation. ((d): Contracting/non-contracting EB-aggregates plated after Day 15–contracting and non-contracting EB. Scale bar for a–d, e1 and f: 200from contracting (a) and non-contracting (b) EBs derived from hi(positive for myosin light chain or MLC) and arrows indicate MSCs (poEBs. (a) CMs in dissociated EBs were quantified by immunofluoresquantified by immunofluorescent staining against CD44. (c) hiPSC–M

iMSCs expressed typical MSC surface markersIn contrast to the parental hPSC lines which showed SSEA-4,TRA-1-60 and TRA-1-80 surface pluripotent marker expres-sion, EB-derived MSCs expressed no SSEA-4, TRA-1-60 andTRA-1-80 right after being isolated from EBs (passage 2 orP2) (Fig. 4A). Expectedly, from early (P2) to late passages

d CARM/SB method. A: Diagram for co-generating CMs and MSCsprocess for co-generating hiPSC-CMs and hiPSC-MSCs. Bottom–

and hiPSC-MSCs. (a): Undifferentiated colonies of hiPSCs on MEF.c): EB-aggregates in suspension 15–21 days after differentiation.21. (e) CMs derived from contracting EBs. (f): MSCs derived fromµm. Scale bar for e2: 100 µm. B: Identification of CMs and MSCsPSCs. Magnification: 20×. Arrowheads indicate cardiomyocytessitive for CD44). C: Quantification of CMs and MSCs in dissociatedcent staining against β-MHC. (b) MSCs in dissociated EBs wereSCs in dissociated EBs quantified by flow cytometry assay.

Figure 2 (continued)

Figure 3 Characterization of hiPSC-CMs. A: Cardiomyocytes immunofluorescently stained with cardiac markers. Scale bar: 30 µM.B: Cardiac action potential (AP) recorded on single hiPSC-CM by whole cell patch-clamp assay. All three subtypes of cardiomyocyteswere identified by unique AP patterns. C: Proportion of each subtype of cardiomyocytes in the hPSC lines examined. D: Cardiacdifferentiation efficiency (% contracting EB /total EB aggregates) determined among three hPSC lines. * pb0.05. hiPSC vs. H9 hESCs.

91One-step derivation of cardiomyocytes and mesenchymal stem cells from human pluripotent stem cells

Figure 4 Characterization of iMSCs by flow cytometry assay. Overlay images by flow cytometry assay. A: Pluripotent markersdetected with hiPSC (1) and passage 2 iMSC (1). B: typical MSC markers determined with iMSC (1) at passages 2, 5, and 19. The bargraft presents the quantitative expression of above MSC-markers on iMSC (1) with different passage numbers. C: HLA-ABC and HLA-DLmarkers detected with passage 3 iMSC (1). D: CD31, CD34 and CD45 detected with passage 3 iMSC (1).

92 H. Wei et al.

(P19) hPSC-derived MSCs expressed typical MSC markersCD29, CD44, CD73, CD90, and CD105 that typified adult MSCs(Fig. 4B). Moreover, CD146, a pro-angiogenic marker of MSCswas found to maintain its expression until late passages

(Fig. 4B) even though its expression was known to diminish inadult MSCs after successive passages in culture Moreover,low MHC-I marker (HLA-ABC) and no MHC-II marker (HLA-DL)were detected (Fig. 4C). Furthermore, markers representing

93One-step derivation of cardiomyocytes and mesenchymal stem cells from human pluripotent stem cells

other cell lineages such as CD34 (hematopoietic marker),CD45 (late leukocyte marker) and CD31 (endothelial marker)were all undetectable with iMSCs (Fig. 4D) (Covas et al.,2008). Fig. 4. presents representative data from iMSC (1)which was derived from the contracting EBs of hiPSC (1). Thesurface marker profile of MSCs derived from non-contractingEBs was shown in Supplementary Fig. 2.

iMSCs were capable of lineage-specific differentiationConsistent with known MSC characteristics, our hPSC-MSCsshowed adipogenic, osteogenic and chondrogenic differen-tiation capacity. Fatty droplets and mineralized calciumdeposits were confirmed by Oil red-O (Fig. 5A) and Alizarinred-S (Fig. 5B) staining respectively. In addition, iMSCsformed pellets under chondrogenic differentiation condi-tions and they stained positive for Alcian Blue (Fig. 5C).

iMSCs showed a robust growth potentialSimilar to hESC-MSCs, the iMSC lines showed robust growthkinetic as demonstrated by over 30 cumulative populationdoublings in continuous culture without signs of cellularsenescence (Fig. 6A). In comparison, the two BM-MSC lines

Figure 5 Multi-potent differentiation of iMSCs. A: hiPSC-MSC (1)Oil-Red-O Staining. Scale bar: 100 µm. B: hiPSC-MSC (1) at passage 7Staining. Scale bar: 100 µm. C: Chondrogenesis of iMSC verified by A

could not be expanded to more than passage 9 with a drasticslowing of growth rates and loss of fibroblastic cellularmorphology. On the other hand, cord blood (CB)-MSCs showeda self renewal potential closer to iMSCs. Consistently, the robustproliferation rates of iMSCs coincided with a relatively sustainedtelomerase activity which was similar to HeLa tumor cell line orthe pluripotent H9 hESC line (Fig. 6B). In contrast, the parentalline (for hiPSCs) of human dermal fibroblasts (HDF, passage 3)and BM-MSC (2) showed a diminished telomerase activity.Furthermore, iMSCs and CB-MSCs displayed a more spindle-shaped morphology as compared to BM-MSCs (Fig. 6C).

iMSCs were safe from tumorigenesis andcytogenetic abnormalitiesIn the same teratoma formation experiment which confirmedthe pluripotency of hiPSC (1) and hiPSC (2), passage 2 and 3iMSCs failed to form teratoma as no tumors were formed andimplanted cells were undetectable by histology assay whentransplanted into renal capsule of SCIDmice (data not shown).In addition, normal karyotype of iMSCs at young (P10) or later(P16) passages was confirmed (Fig. 6D).

at passage 7 differentiated into adipocyte-like cells verified bydifferentiated into osteocyte-like cells verified by Alizarin Red Slcian blue staining. Scale bar: 50 µm.

Figure 6 Growth kinetics, telomerase activity and karyotype of iMSCs. A: Cumulative population doubling levels (CPDL) from iMSCs,hESC-MSCs, BM-MSC (1) (from healthy donor) and BM-MSC (2) (from a myocardial infarction patient), and cord blood (CB)-MSC wererecorded and compared. B: Telomerase activity in iMSC (1) of different passages was detected. Human dermal fibroblast (HDF) atpassage 2 and BM-MSCs were used as negative controls while HeLa cell and H9 ES cell were used as positive controls. The arrows onthe right indicate the bands of TRAP products and internal control. C: Morphological comparison of iMSCs, CB-MSCs and BM-MSCs.D: Normal karyotype observed with iMSC (1) at passages 10 and 16.

94 H. Wei et al.

iMSCs showed a marked pro-angiogenic potentialConsistent with previous report of enhanced pro-angiogenicactivity of hiPSC-derived MSCs in comparison to BM-MSCs (Lianet al., 2010), our iMSCs demonstrated superior angiogenic

activity. When co-cultured with HUVEC, CD146+ sorted iMSCspromoted more extensive vascular network (Figs. 7A–D) andmore mature tubule formation in length (Fig. 7L) and caliberof vascular structures as compared to HUVEC alone or co-

Figure 7 Endothelial tube formation assay. HUVECs at passage 3, iMSC (1) at passage 11 and BM-MSC (2) at passage 3 were involved inthis study. A, F: HUVECs only (magnification of 4× and 10×). B, G: HUVECs + iMSC (1) (4× and 10×). C, H: HUVECs + CD44+/CD146+ iMSC (1)(4× and 10×). D, I: HUVECs + BM-MSC (2) (4× and 10×). E: iMSC (1) only. J: BM-MSCs (2) only. K: Phase contrast and fluorescent images ofHUVECs (CM-Dil labeled, Red) and iMSCs (GFP labelled, green) and overlay. Nuclei were stainedwith DAPI (blue). L: Quantification of tubelength. M: Quantification of tube thickness. * pb0.05, ** pb0.01, MSC+HUVEC vs. HUVEC alone (control). Scale bars: A–E and J, 200 µM;F–I and K, 100 µM.

95One-step derivation of cardiomyocytes and mesenchymal stem cells from human pluripotent stem cells

cultured with unsorted iMSCs. In contrast, BM-MSCs failed toshow either effect. Moreover, there was marked co-localization of iMSCs as peri-vascular cells in the HUVECsconstituted vascular network (Fig. 7K).

iMSCs showed a marked tissue healing potentialConditioned medium derived from MSCs is known to containimportant growth factors such as FGF and VEGF (Kinnaird etal., 2004a, 2004b) that are important for therapeutic effectof MSCs. Consistent with augmented reparative property,conditioned medium harvested from our iMSCs induced amore robust migration of fibroblasts in an in vitro woundhealing model as compared to conditioned media from BM-MSC or fibroblast alone. There was earlier migratory

response and superior ingrowth of fibroblasts into thecreated tracks (Fig. 8). Moreover, such effect of iMSC wassustained even in late passage in culture. The effect of iMSC(1) conditioned medium on fibroblast wound healing wasshown in Fig. 8.

Discussion

Large quantities of CMs and supporting parenchymal cells suchas endothelial cells and smooth muscle cells are needed formeaningful cardiac cell therapy in post myocardial infarction.Paracrine effect derived from MSCs play an important role incardiac repair by promoting endothelial angiogenesis. Similar

Figure 8 Wound-healing assay. A–D: Human neonatal dermal fibroblasts (passage 8), BM-MSC (2) (passage 4), CB-MSCs (passage 6)and iMSC (1) (passage 8) were adopted in fibroblast wound-healing assay. Conditioned media containing 2% FBS were collected fromeach cell lines after 24 h of culture. Gap area was measured 48-hours after application of conditioned media. E: Quantification of thegap area. * pb0.05, ** pb0.01, *** pb0.001, BM-MSC, CB-MSCs and iMSC-derived conditioned media vs. fibroblast-derived conditionedmedium as control. ┼ pb0.05, ╫ pb0.01, BM-MSC and CB-MSC derived conditioned media vs. iMSC-derived conditioned medium.Images were taken by a Leica stereomicroscope.

96 H. Wei et al.

to pluripotent hESCs, patient-derived iPSCs can be cryopre-served and repeatedly used to generate specific cell lineagessuch as CMs and MSCs. Given the supportive role of MSCs in celltherapy, a collective administration of both CMs and MSCsmaybe important for cardiogenesis and angiogenesis for cardiacrepair (Kreutziger et al., 2011).

Currently, hiPSC-CMs and iPSC-MSCs are separately derivedthrough a laborious process. While cardiac differentiation wasmainly achieved via 3D EBs formation, hESCs and hiPSCswere 2Ddifferentiated to MSCs, FACS sorted and may even clonallyderived through single cell dilution to yield homogenouspopulation of MSCs (Lian et al., 2007). In this study, wedemonstrated that using a CARM/SB cardiac differentiationprotocol (a serum-free insulin-free cardiac differentiationmedium supplemented with a specific inhibitor for the p38MAPK, SB203580 (Graichen et al., 2008), CMs and MSCs weresimultaneously derived from a single cardiac differentiationexperiment. While CMs can be obtained from contracting EBs,MSCs can be derived from both contracting and non contractingEBs. Comparing to the CARM/SB protocol which consistentlyyielded homogenous iMSCs from the outgrowth of EB aggregates

plated, another well-adopted serum containing cardiac differ-entiation protocol which uses 20% serum and the EB aggregateswere plated early after EB formation, hereby named EB-20(Kehat et al., 2001) method, yielded marginally lower cardiacdifferentiation efficiency, but failed to generate homogenousMSCs even after repeated passages in culture (data not shown).

Thus we speculate that the successful generation ofhomogenous MSCs by CARM/SB protocol could be due toseveral factors: 1) insulin‐free differentiation medium knownto favor the formation of mesoderm (Freund et al., 2008), 2)serum‐free differentiation medium may prevent excessiveoutgrow of mixed cell lineages from three germ layers, and 3)prolonged suspension culturing of EB aggregates (till 15 days)may withhold excessive growth of various cell lineages.

Our iMSCs meet the criteria for human MSCs set by theMesenchymal and Tissue Stem Cell Committee of theInternational Society for Cellular Therapy (Dominici et al.,2006). Our iMSCs were multipotent and expressed typical MSCmarkers (CD29, CD44, CD73, CD90, and CD105), but express noendothelial marker CD31 neither hematopoietic markers CD45and CD34 associated with bone marrow-derived cells.

97One-step derivation of cardiomyocytes and mesenchymal stem cells from human pluripotent stem cells

Large quantities of transplantable cells are critical for celltherapy since cell loss during and after cell transplantation issubstantial. However, due to the limited self-renewal poten-tial and lower functional potency associated with aging andchronic diseases, only human MSCs with a passage number lessthan 6 might retain sufficient reparative capacity (Briquet etal., 2010; Li et al., 2008). Furthermore, clinical trials indicatedthat as much as 107–108 or 5×106/kg MSCs would have to beprepared for clinical application (Chen et al., 2004; Macmillanet al., 2009). In addition, repeated generation of largenumbers of young and functional-specific subpopulation ofMSCs may be needed to sustain therapeutic efficacy in repeatinjections. However, single aspiration of bone marrow mayonly yield sufficient MSCs for one time cell delivery and suchyield is greatly insufficient for isolation of functional sub-populations. In contrast, large quantities of iMSCs (ascompared to BM-MSCs) could be generated from hiPSCs andour iMSCs showed robust self-renewal potential that isconsistent with previously reported hESC-MSCs and hiPSC-MSCs (Hwang et al., 2008; Lian et al., 2007).

The safety of iMSCs remains to be a major concern forclinical application since they are derived from pluripotentiPSCs with prolonged culture period. Our iMSCs showed noexpression of SSEA-4, TRA-1-60 and TRA-1-81 and noteratoma formation despite obviating FACS sorting or thederivation of single cell colonies (Lian et al., 2007). Ourdata suggest that such iMSCs derived with the CARM/SBprotocol are free of pluripotent stem cell contaminationand could maintain a normal karyotype after cultureexpansion.

Pro-angiogenic effect of MSCs has been documented inpost-MI cardiac repair (Amado et al., 2005) and ischemiclimbs (Lian et al., 2010). MSCs have been suggested tofunction as pericytes or perivascular cells in supportingendothelial cell to form vascular networks (Covas et al.,2008; Crisan et al., 2008; Schugar et al., 2009) in which MSCsare indispensable for forming functionally competent vas-culature without leakage (Kreutziger et al., 2011). Consis-tently, our results showed that iMSCs acted cooperativelywith HUVECs to form extensive tube-like vascular structuresfurther supporting their perivascular role in angiogenesis.Furthermore, our iMSCs retained stronger pro-angiogeniceffect at passage 12 than BM-MSCs at passage 3. In addition,the CD44+/146+ sorted iMSCs showed a stronger pro-angiogenic effect than unsorted iMSCs. Consistent withpro-wound healing effect of MSCs (Walter et al., 2010; Wuet al., 2007), our iMSCs demonstrated a superb healingeffect in promoting cellular migration and tissue coales-cence than that of BM-MSCs. Collectively, our iMSCs exhibitsubstantial paracrine protective effects that may wellsupport substantial tissue repair when transplanted togeth-er with concurrently generated CMs for autologous celltherapy.

Materials and methods

Generation and characterization of humanpluripotent stem cell lines

Human iPSC lines, hiPSC (1) and hiPSC (2), were generated inour laboratory from human neonatal dermal fibroblasts

(ATCC) via retroviral-based reprogramming with Yamanakafactors (OCT-4, SOX-2, c-Myc and KLF-4) (Takahashi et al.,2007) . In addition, a human embryonic stem cell (hESC) line(H9), acquired from WiCell Institute, Wisconsin, USA, wasadopted as control.

ImmunocytochemistryhiPSCs grown on feeder cells were fixed and permeabilizedand subjected to immunocytochemistry assay with mAbsagainst OCT-4, SSEA-4, TRA 1–60, and TRA 1–81 (Millipore).Laser confocal microscopic images were taken with Carl Zeiss510 microscope.

RT-PCR and genomic DNA PCRRNA was extracted from H9 hESCs, hiPSC(1), hiPSC (2),

hiPSC (1) and (2)-derived EBs and the parental HDF. Next,first-strain cDNA was synthesized using oligo dTs from 1 μg oftotal RNA. Endogenous and transgene expression of OCT-4,SOX-2, c-MYC and KLF-4 genes were amplified by semi-quantitative RT-PCR. Moreover, genomic DNA was extractedfrom H9 hESCs, hiPSC (1), hiPSC (2), iMSC (1), and theparental HDF. OCT-4, SOX-2, c-MYC and KLF-4 transgeneswere amplified by genomic DNA PCR. Two sets of primerswere used. The 1st was for the endogenous gene RT-PCR andthe 2nd was for both of the transgene RT-PCR and genomicDNA PCR. All primers and PCR conditions used in this studywere adopted from a previous report (Takahashi et al.,2007).

Teratoma formationApproximately 2×106 of hiPSC (1) and (2) were injected intothe kidney capsule and hindlimb muscle of female NOD-SCIDmice for 12 weeks. Tumors were explanted, fixed, paraffinembedded and sections were stained with haematoxylin andeosin.

KaryotypingChromosomal studies were carried out in the CytogeneticLab of KK Women and Children Hospital, Singapore. StandardG-banding chromosome analysis was performed with hiPSC(1) and hiPSC (2) using standard protocols for high resolutionG-banding.

Generation of hiPSC-CMs and hiPSC-MSCs

Prior to differentiation, hiPSCs and hESCs were maintained onmitomycin C-treated mouse embryonic fibroblasts (MEF)feeder in hESC medium (final concentration: 80% KnockoutDMEM, 20% Serum replacement, 1% non-essential amino acid,1 mM L-glutamine, 0.1 mM beta-mercaptoethanol and 4 ng/mlbasic Fibroblasts Growth Factor (bFGF), all from Invitrogen).

Cardiac differentiation was performed by the CARM/SBdifferentiation procedure (Graichen et al., 2008): On Day 1, EBswere formed mechanically from hESC/hiPSC colonies andtransferred into 6-well non-adherent dishes and cultured inhESC medium without bFGF for 24 h. On Day 2, the CM/MSCdifferentiation were initiated with a serum-free and insulin-free medium, CARM (DMEM High Glucose 485 ml, L-glutamine5 ml, NEAA 5 ml, Selenium Transferrin 5 ml (Sigma) and 2-mercaptoethanol 3.5 μl) supplemented with SB 203580 (5 μM),a specific p38-MAPK inhibitor (Sigma) and kept for 4 days till

98 H. Wei et al.

Day 6 when the differentiation medium was changed andmaintained till Day 14. EB aggregates were plated on 0.1%gelatin-coated 6-well plate or 3.5 cm petri dishes in DMEMmedium containing 2% FBS and kept for 1 week. Contracting EBaggregates emerged from Days 15 to 21. Next, the contractingCM clumps were dissected out. In addition, H9 hESC, hiPSC (1)and (2) were also differentiated with the EB-20 method (Kehatet al., 2001; Zhang et al., 2009) in which the EB aggregateswere plated on Day 8 before contracting EBs emerged.Contracting and non-contracting Day 15 EBs from the samedifferentiation experiment were fixed with 4% paraformalde-hyde (PFA) and permeabilized with 0.1% Triton X-100 and co-stained with a polyclonal antibody against human cardiacmyosin light chain (MLC), (rabbit anti-human MLC), a specificCMs marker from Abcam (phospho S20) antibody (ab2480))followed by a second antibody (Alexa Fluo® 488 goat anti-rabbit IgG, Invitrogen), and anti-CD44, a MSC marker from BDCat#550392. Followed by a second antibody (Alexa Fluo® 555Donkey anti-mouse IgG, Invitrogen).

EBs dissociation and quantitative assays for CMs and MSCsin EBs

Day 15 EBs were dissociated in 0.1% Trypsin as and seededon glass coverslips with a density of 5×105 cells/cm2 with acytospin approach as previously described (Zweigerdt et al.,2003). Cells were fixed and permeabilized as abovementioned. Next, cells from different slides were incubatedwith the primary antibodies against β-MHC (Alexis Bio-chemicals, FL, USA), or CD44 followed by second antibody(Alexa Fluo® 555 Donkey anti-mouse IgG from Invitrogen).Cells were counter-stained with DAPI for nuclear stainingand examined on a Zeiss LSM 710 Microscope.

Isolation and expansion of hiPSC-MSCsAs shown in Fig. 2A, following the CARM/SB method, rapid

outgrowth of MSC-like cells from the attached EB aggregateswere observed immediately after plating. Those MSC-likecells migrated out and covered N80% of a well of 6-well platein two weeks time. After mechanical removing CM clumpsand other non-MSC-like cells, the remaining MSC-like cellswere cultured in MSC medium (DMED supplemented with 10%serum) to full confluency and dissociated with trypsin andtransferred into a T25 flask (Passage 1). Cells were furtherexpanded in MSC medium with a 1:4 split ratio. MSCs werecontinually expanded to passage 20 and above.

Characterization of hiPSC-CMs

Clumps of contracting CMs were dissected out under amicroscope and were enzymatically dissociated (Maltsev etal., 1994). The cardiac marker expression of hiPSC-CMs wasdetermined by immunocytochemistry assay. Dissociatedcardiomyocytes were continually cultured in DMEM+2% FBSfor 2 more weeks before characterization.

Immunofluorescent characterization of cardiac markersThe prepared cardiomyocytes were seeded on glass coverslipsand immunocytochemically stained for cardiac markers withantibodies against cardiac sarcomeric α-actinin (clone EA-35.Sigma) and β-MHC (Alexis Biochemicals, FL, USA), followed byAlexa Fluo® 488 goat anti-rabbit IgG (Invitrogen, CA, USA);

and cardiac titin (1:10) (Sigma-Alrich, MO, USA) followed byAlexa Fluo® 555 donkey anti-rabbit IgG. All cardiomyocytestested in this study were ~5 weeks post differentiation.

Electrophysiological characterization of cardiomyocytesDissociated CMs were seeded on uncoated 3.5 cm petri dishesand were transferred to a recording chamber mounted on thestage of an inverted microscope (TE2000-S, Nikon, Tokyo,Japan). Whole-cell action potential (APs) was recorded with aPatch-Clamp amplifier (Axon 200B, Axon Instruments). Dataacquisition and analysis were performed using Clampex andClampfit software (version 10.0, Axon Instruments). In thecurrent-clampmode, the APs were recorded in normal Tyrode'ssolution. Ventricular-like, atrial-like and nodal-like-CM wereidentified by their characteristic action potential (AP) proper-ties including AP amplitude, AD duration and maximum rate ofdepolarization (dV/dtmax).

Characterization of hiPSC-MSCs

hiPSC (1) and hiPSC (2) derived-MSCs, labeled as iMSC (1) andiMSC (2), respectively, were thoroughly characterized. Twolines of human bone marrow (BM)-derived MSCs were used inthis study as control. BM-MSC (1) was derived from marrowmononuclear cells (DV Biologics, Costa Mesa, CA). BM-MSC(2) and CB-MSCs were derived from the bone marrow of amyocardial infarction patient recruited at National HeartCentre Singapore, and healthy umbilical cord blood respec-tively, following protocols approved by the InstitutionalReview Board of Singapore General Hospital.

Surface markers measurement and cell sortingThe derived iMSC (1), iMSC (2) and hESC-MSCs, BM-MSCs andcord blood-MSC (CD-MSCs) were analyzed by flow cytometrywith a BD Aria II system. PE, APC and FITC congregated mAbs(all from BD Biosciences) were used to determine CD29, CD44,CD73, CD90, CD105, CD146, CD31, CD34 and CD45 expression.CD44+/CD146+ hiPSC-MSCs were sorted out by fluorescent-activated cell sorting (FACS). Moreover, pluripotent stem cellmarkers SSEA-4, TRA-1-60, TRA-1-81 (Primary antibodies fromMillipore) and MHC Class I and II markers, HLA-AC and HLA-DL(BD Biosciences) were determined.

Multi-potent differentiation of iMSCs

Adipogenic and osteogenic differentiation of iMSCs. Passage 6iMSC (1) were differentiated into adipocytes and osteocytes inthe corresponding differentiation media from Stemcell Tech-nologies®. The adipocytes and osteocytes were confirmed byOil-Red-O staining and Alizarin Red S staining respectively.

Chondrogenic differentiation of iMSCs. Passage 3 iMSC (1)wasdifferentiated into chondrocytes using STEMPRO ChondrogenesisDifferentiation Kit (GIBCO). Alcian Blue Staining was conductedon cell pellet 18 days after differentiation.

Growth kinetics of iMSCs. iMSC (1), iMSC (2), H3 hESCs, BM-MSCs and CB-MSCs were serially cultured to Passage 20. AfterPassage 2, iMSCs were expanded with a starting density of1×106 cells per T75 flasks in MSC medium. Cells weresubcultured when they reached 90% confluence. The growth

99One-step derivation of cardiomyocytes and mesenchymal stem cells from human pluripotent stem cells

kinetics of MSCs derived from two bone marrow samples wassimilarly determined. The cumulated population doublingslevel (NPDL) was calculated and plotted.

Telomerase activity assay. The telomerase activity ofcultured cells was explored using TRAPEZE telomerase detectionkit (Millipore) as described (Kim et al., 1994). Cell extracts (froma total of 5×106) was prepared from iMSCs, BM-MSCs, humandermal fibroblasts, HeLa cell and H9 ESCs using 3-[(3-cholami-dopropyl)-dimethylammonio]-1-propanesulfonate (CHAPS) lysisbuffer. Detection of telomerase activity in cells was carried outby the PCR-based telomeric-repeat amplification protocol(TRAP).

Teratoma formation assay. See above section of hiPSCteratoma formation assay. In the same teratoma formationexperiment for hiPSCs, a total of 2×106 of passage 2 and 3iMSCs were injected intramuscularly into the hindlimb andkidney capsule of female NOD-SCID mice for 16 weeks toform teratoma.

Karyotyping. The cytogenetic stability of iMSCs was mon-itored by karyotyping by the Cytogenetic Lab of KK Womenand Children Hospital Singapore following standard protocolsfor high resolution G-banding.

Matrigel-based tube formation assay. Matrigel (BD 8482) of75 μL was added in each well of a 96-well plate. Unsorted andCD44+/CD146+ iPSC-MSCs (P6-P11) and unsorted BM-MSCs(P3), with a total number of 0.6×104, were seeded on Matrigeltogether with 1.25×104 human umbilical vein endothelialcells (HUVECs) (Invitrogen) at P3–4. After 24 h, vessel-likenetwork structureswere examined under lightmicroscopy andimages were taken. Co-localization of HUVECs and iPSC-MSCswere confirmed by fluorescent microscopy of pre-labeledHUVECs (CM-Dil) and iPSC-MSCs (Lentiviral vector expressingGFP). With each experiment, triplicated wells were set fordifferent conditions and three independent experiments wereperformed. The average data of triplicated wells wascalculated and data from three independent experimentswas plotted as bar-graft (mean% ±SD). Tube formation [totaltube length (mm)/high-powered field (mm2)] was quantifiedusing ImageJ software (NIH).

Wound-healing assay. Human neonatal dermal fibroblasts(P8) were cultured to 90% confluence and a 2-mm wide trackwas physically created by a cell scraper. The conditioned-media (CMed) of iMSC, iMSC (P6) and BM-MSC (P4) werecollected after 24 h incubating in serum-free DMEM andadded to the wounded monolayer of fibroblasts andincubated for 24 h. With each experiment, triplicated wellswere set for different conditions and three independentexperiments were performed. The average data of triplicat-ed wells was calculated and data from three independentexperiments was plotted as bar-graft (mean±SD). Imageswere acquired with a light microscope (BX51, Olympus) anddigital camera (Moticam 1000, Moticam). The gap area(mm2) was quantified using ImageJ software (NIH) by scaleset at 300 pixel/mm.

Supplementary data associated with this article can befound, in the online version, at http://dx.doi.org/10.1016/j.scr.2012.04.003.

Statistical analysis

Data were expressed as mean±SD and analyzed by 2-tailedStudent t tests. A p values below 0.05 were considered to bestatistically significant.

Author disclosure statement

No competing financial interests exist.

Acknowledgment

This study is funded by National Research Foundation ofSingapore (NRF2008 NRF-CRP001-068).

References

Amado, L.C., Saliaris, A.P., Schuleri, K.H., St John, M., Xie, J.S.,Cattaneo, S., Durand, D.J., Fitton, T., Kuang, J.Q., Stewart, G.,et al., 2005. Cardiac repair with intramyocardial injection ofallogeneic mesenchymal stem cells after myocardial infarction.Proc. Natl. Acad. Sci. U. S. A. 102, 11474–11479.

Briquet, A., Dubois, S., Bekaert, S., Dolhet, M., Beguin, Y., Gothot, A.,2010. Prolonged ex vivo culture of human bone marrow mesenchy-mal stem cells influences their supportive activity toward NOD/SCID-repopulating cells and committed progenitor cells of Blymphoid and myeloid lineages. Haematologica 95, 47–56.

Chen, S.L., Fang, W.W., Ye, F., Liu, Y.H., Qian, J., Shan, S.J., Zhang,J.J., Chunhua, R.Z., Liao, L.M., Lin, S., et al., 2004. Effect on leftventricular function of intracoronary transplantation of autologousbone marrow mesenchymal stem cell in patients with acutemyocardial infarction. Am. J. Cardiol. 94, 92–95.

Covas, D.T., Panepucci, R.A., Fontes, A.M., Silva Jr., W.A.,Orellana, M.D., Freitas, M.C., Neder, L., Santos, A.R., Peres,L.C., Jamur, M.C., et al., 2008. Multipotent mesenchymalstromal cells obtained from diverse human tissues sharefunctional properties and gene-expression profile with CD146+perivascular cells and fibroblasts. Exp. Hematol. 36, 642–654.

Crisan, M., Yap, S., Casteilla, L., Chen, C.W., Corselli, M., Park, T.S.,Andriolo, G., Sun, B., Zheng, B., Zhang, L., et al., 2008. Aperivascular origin for mesenchymal stem cells in multiple humanorgans. Cell Stem Cell 3, 301–313.

Dominici, M., Le Blanc, K., Mueller, I., Slaper-Cortenbach, I., Marini,F., Krause, D., Deans, R., Keating, A., Prockop, D., Horwitz, E.,2006. Minimal criteria for defining multipotent mesenchymalstromal cells. The International Society for Cellular Therapyposition statement. Cytotherapy 8, 315–317.

Fan, M., Chen, W., Liu, W., Du, G.Q., Jiang, S.L., Tian, W.C., Sun, L.,Li, R.K., Tian, H., 2010. The effect of age on the efficacy of humanmesenchymal stem cell transplantation after a myocardialinfarction. Rejuvenation Res. 13, 429–438.

Freund, C., Ward-van Oostwaard, D., Monshouwer-Kloots, J., van denBrink, S., van Rooijen, M., Xu, X., Zweigerdt, R., Mummery, C.,Passier, R., 2008. Insulin redirects differentiation from cardiogenicmesoderm and endoderm to neuroectoderm in differentiatinghuman embryonic stem cells. Stem Cells 26, 724–733.

Graichen, R., Xu, X., Braam, S.R., Balakrishnan, T., Norfiza, S., Sieh, S.,Soo, S.Y., Tham, S.C., Mummery, C., Colman, A., et al., 2008.Enhanced cardiomyogenesis of human embryonic stem cells by a smallmolecular inhibitor of p38 MAPK. Differentiation 76, 357–370.

Heeschen, C., Lehmann, R., Honold, J., Assmus, B., Aicher, A.,Walter, D.H., Martin, H., Zeiher, A.M., Dimmeler, S., 2004.Profoundly reduced neovascularization capacity of bone marrow

100 H. Wei et al.

mononuclear cells derived from patients with chronic ischemicheart disease. Circulation 109, 1615–1622.

Hwang, N.S., Varghese, S., Lee, H.J., Zhang, Z., Ye, Z., Bae, J., Cheng,L., Elisseeff, J., 2008. In vivo commitment and functional tissueregeneration using human embryonic stem cell-derived mesenchy-mal cells. Proc. Natl. Acad. Sci. U. S. A. 105, 20641–20646.

Kasper, G., Mao, L., Geissler, S., Draycheva, A., Trippens, J., Kuhnisch,J., Tschirschmann, M., Kaspar, K., Perka, C., Duda, G.N., et al.,2009. Insights into mesenchymal stem cell aging: involvement ofantioxidant defense and actin cytoskeleton. Stem Cells 27,1288–1297.

Kehat, I., Kenyagin-Karsenti, D., Snir, M., Segev, H., Amit, M., Gepstein,A., Livne, E., Binah,O., Itskovitz-Eldor, J., Gepstein, L., 2001.Humanembryonic stem cells can differentiate into myocytes with structuraland functional properties of cardiomyocytes. J. Clin. Invest. 108,407–414.

Kim, N.W., Piatyszek, M.A., Prowse, K.R., Harley, C.B., West, M.D.,Ho, P.L., Coviello, G.M., Wright, W.E., Weinrich, S.L., Shay,J.W., 1994. Specific association of human telomerase activitywith immortal cells and cancer. Science 266, 2011–2015.

Kinnaird, T., Stabile, E., Burnett, M.S., Lee, C.W., Barr, S., Fuchs, S.,Epstein, S.E., 2004a. Marrow-derived stromal cells express genesencoding a broad spectrum of arteriogenic cytokines and promotein vitro and in vivo arteriogenesis through paracrine mechanisms.Circ. Res. 94, 678–685.

Kinnaird, T., Stabile, E., Burnett, M.S., Shou, M., Lee, C.W., Barr,S., Fuchs, S., Epstein, S.E., 2004b. Local delivery of marrow-derived stromal cells augments collateral perfusion throughparacrine mechanisms. Circulation 109, 1543–1549.

Kreutziger, K.L., Muskheli, V., Johnson, P., Braun, K., Wight, T.N.,Murry, C.E., 2011. Developing vasculature and stroma in engineeredhuman myocardium. Tissue Eng. Part A 17, 1219–1228.

Li, W.Y., Choi, Y.J., Lee, P.H., Huh, K., Kang, Y.M., Kim, H.S., Ahn,Y.H., Lee, G., Bang, O.Y., 2008. Mesenchymal stem cells forischemic stroke: changes in effects after ex vivo culturing. CellTransplant. 17, 1045–1059.

Lian, Q., Lye, E., Suan Yeo, K., KhiaWay Tan, E., Salto-Tellez, M., Liu,T.M., Palanisamy, N., El Oakley, R.M., Lee, E.H., Lim, B., et al.,2007. Derivation of clinically compliant MSCs from CD105+, CD24−differentiated human ESCs. Stem Cells 25, 425–436.

Lian, Q., Zhang, Y., Zhang, J., Zhang, H.K., Wu, X., Lam, F.F., Kang,S., Xia, J.C., Lai, W.H., Au, K.W., et al., 2010. Functionalmesenchymal stem cells derived from human induced pluripo-tent stem cells attenuate limb ischemia in mice. Circulation 121,1113–1123.

Macmillan, M.L., Blazar, B.R., DeFor, T.E., Wagner, J.E., 2009.Transplantation of ex-vivo culture-expanded parental haplo-identical mesenchymal stem cells to promote engraftment inpediatric recipients of unrelated donor umbilical cord blood:results of a phase I–II clinical trial. Bone Marrow Transplant. 43,447–454.

Maltsev, V.A., Wobus, A.M., Rohwedel, J., Bader, M., Hescheler, J.,1994. Cardiomyocytes differentiated in vitro from embryonic

stem cells developmentally express cardiac-specific genes andionic currents. Circ. Res. 75, 233–244.

Mehta, A., Chung, Y.Y., Ng, A., Iskandar, F., Atan, S., Wei, H.,Dusting, G., Sun, W., Wong, P., Shim, W., 2011. Pharmacologicalresponse of human cardiomyocytes derived from virus-freeinduced pluripotent stem cells. Cardiovasc. Res. 91, 577–586.

Nie, Y., Lau, C., Lie, A., Chan, G., Mok, M., 2010. Defectivephenotype of mesenchymal stem cells in patients with systemiclupus erythematosus. Lupus 19, 850–859.

Park, I.H., Arora, N., Huo, H., Maherali, N., Ahfeldt, T., Shimamura, A.,Lensch, M.W., Cowan, C., Hochedlinger, K., Daley, G.Q., 2008.Disease-specific induced pluripotent stem cells. Cell 134, 877–886.

Schugar, R.C., Chirieleison, S.M.,Wescoe, K.E., Schmidt, B.T., Askew, Y.,Nance, J.J., Evron, J.M., Peault, B., Deasy, B.M., 2009. High harvestyield, high expansion, and phenotype stability of CD146 mesenchymalstromal cells from whole primitive human umbilical cord tissue. J.Biomed. Biotechnol. 2009, 789526.

Schuleri, K.H., Feigenbaum, G.S., Centola, M., Weiss, E.S., Zimmet,J.M., Turney, J., Kellner, J., Zviman, M.M., Hatzistergos, K.E.,Detrick, B., et al., 2009. Autologous mesenchymal stem cellsproduce reverse remodelling in chronic ischaemic cardiomyop-athy. Eur. Heart J. 30, 2722–2732.

Takahashi, K., Yamanaka, S., 2006. Induction of pluripotent stemcells from mouse embryonic and adult fibroblast cultures bydefined factors. Cell 126, 663–676.

Takahashi, K., Tanabe, K., Ohnuki, M., Narita, M., Ichisaka, T., Tomoda,K., Yamanaka, S., 2007. Induction of pluripotent stem cells fromadult human fibroblasts by defined factors. Cell 131, 861–872.

Walter, M.N., Wright, K.T., Fuller, H.R., MacNeil, S., Johnson, W.E.,2010. Mesenchymal stem cell-conditioned medium acceleratesskin wound healing: an in vitro study of fibroblast andkeratinocyte scratch assays. Exp. Cell Res. 316, 1271–1281.

Wu, Y., Chen, L., Scott, P.G., Tredget, E.E., 2007. Mesenchymalstem cells enhance wound healing through differentiation andangiogenesis. Stem Cells 25, 2648–2659.

Xu, C., Police, S., Rao, N., Carpenter, M.K., 2002. Characterizationand enrichment of cardiomyocytes derived from human embry-onic stem cells. Circ. Res. 91, 501–508.

Yu, J., Vodyanik, M.A., Smuga-Otto, K., Antosiewicz-Bourget, J.,Frane, J.L., Tian, S., Nie, J., Jonsdottir, G.A., Ruotti, V.,Stewart, R., et al., 2007. Induced pluripotent stem cell linesderived from human somatic cells. Science 318, 1917–1920.

Yu, J.M., Wu, X., Gimble, J.M., Guan, X., Freitas, M.A., Bunnell, B.A.,2011. Age-related changes in mesenchymal stem cells derivedfrom rhesus macaque bone marrow. Aging Cell 10, 66–79.

Zhang, J., Wilson, G.F., Soerens, A.G., Koonce, C.H., Yu, J.,Palecek, S.P., Thomson, J.A., Kamp, T.J., 2009. Functionalcardiomyocytes derived from human induced pluripotent stemcells. Circ. Res. 104, e30–e41.

Zweigerdt, R., Burg, M., Willbold, E., Abts, H., Ruediger, M., 2003.Generation of confluent cardiomyocyte monolayers derived fromembryonic stem cells in suspension: a cell source for newtherapies and screening strategies. Cytotherapy 5, 399–413.