Electrophysiological deterioration and resurrection in the scarred heart. Pijnappels, D.A.

(1)

Citation

Pijnappels, D. A. (2009, June 18). Electrophysiological deterioration and resurrection in the scarred heart. Retrieved from

https://hdl.handle.net/1887/13851

Version: Corrected Publisher’s Version

License: Licence agreement concerning inclusion of doctoral thesis in the Institutional Repository of the University of Leiden

Downloaded from: https://hdl.handle.net/1887/13851

Note: To cite this publication please use the final published version (if applicable).

(2)

I V

Chapter

N ormal commitment but impaired maturati- on of cardiomyocytes from induced pluripo- tent stem cells

Sean M. Wu,^1,9,* Matthias Stadtfeld,^2,* Min Zeng,^1,* Daniël A. Pijnappels,^1,4 Esther Yu,¹ Yuko Fujiwara,⁵ Guangwen Wang,⁶ Stuart H. Orkin,^5,7,8,9 Laurie Jackson-Grusby,^6,9 Konrad Hochedlinger^2,3,9

*Authors contributed equally

Cardiovascular Research Center,¹ Division of Cardiology, Department of Medicine, Center for Regenerative Medicine,² Cancer Center,³ Massachusetts General Hospital, Boston, MA 02114

Department of Cardiology,⁴ Leiden University Medical Center, Leiden, the Netherlands Division of Hematology and Oncology,⁵ Department of Pathology,⁶ Howard Hughes Medical Institute,⁷ Children’s Hospital, Boston, MA 02115, USA

Department of Pediatric Oncology,⁸ Dana-Farber Cancer Institute, Boston MA, 02115, USA

Harvard Stem Cell Institute,⁹ Cambridge, MA 02138, USA Submitted for publication

(3)

I V

Chapter

N ormal commitment but impaired maturati- on of cardiomyocytes from induced pluripo- tent stem cells

Sean M. Wu,^1,9,* Matthias Stadtfeld,^2,* Min Zeng,^1,* Daniël A. Pijnappels,^1,4 Esther Yu,¹ Yuko Fujiwara,⁵ Guangwen Wang,⁶ Stuart H. Orkin,^5,7,8,9 Laurie Jackson-Grusby,^6,9 Konrad Hochedlinger^2,3,9

*Authors contributed equally

Cardiovascular Research Center,¹ Division of Cardiology, Department of Medicine, Center for Regenerative Medicine,² Cancer Center,³ Massachusetts General Hospital, Boston, MA 02114

Department of Cardiology,⁴ Leiden University Medical Center, Leiden, the Netherlands Division of Hematology and Oncology,⁵ Department of Pathology,⁶ Howard Hughes Medical Institute,⁷ Children’s Hospital, Boston, MA 02115, USA

Department of Pediatric Oncology,⁸ Dana-Farber Cancer Institute, Boston MA, 02115, USA

Harvard Stem Cell Institute,⁹ Cambridge, MA 02138, USA Submitted for publication

(4)

Abstract

Enforced expression of Oct3/4, Sox2, Klf4, and c-Myc in somatic cells results in developmental reprogramming and acquisition of a pluripotent phenotype. Despite their ability to contribute widely in chimeric mouse assays, induced pluripotent stem (iPS) cells have not been examined rigorously for their biological behavior in vitro. To investigate this, we compared iPS and embryonic stem (ES) cell lines derived from Nkx2.5-eGFP transgenic mice for their capacity to differentiate cell autonomously into cardiac cells. We found that differentiating iPS cells undergo appropriate expression of pluripotency, germ layer, and cardiac progenitor markers. iPS cell-derived cardiac progenitor cells differentiate into immature cardiomyocytes with electrophysiological properties similar to those derived from ES cells. However, upon further differentiation, cardiomyocytes derived from iPS cells exhibit decreased sarcomeric striation and increased apoptosis when compared with ES cell-derived cardiomyocytes. This phenotype is associated with an elevated expression of Delta-like 1 and can be reversed by inhibition of Notch signaling. These data suggest that improvements in the reprogramming process are required to allow proper maturation of cardiac iPS cells for disease modeling and clinical applications.

(5)

Introduction

P

luripotent stem cells hold tremendous promise to unravel key developmental pathways involved in organogenesis and may serve as a tissue source for regenerative therapy. Their ability to self-renew long-term and differentiate into cells of a variety of lineages upon in vitro induction has facilitated discovery of important ge- nes and cell populations responsible for tissue-specific differentiation. Recent studies have identified key transcription factors involved in the induction and maintenance of the pluripotent phenotype including Oct3/4, Nanog, Sox2, Rex1, and others.^2,3,14,30 By direct or indirect interactions, these proteins prevent the activation of germ layer differentiation programs while promoting developmental pluripotency.

The widely held belief that developmental commitment and terminal differentiation are irreversible was recently challenged by the discovery that over-expression of only four factors (Oct3/4, Sox2, Klf4, c-Myc) is sufficient to convert adult mouse-tail fibroblasts into so-called induced pluripotent stem (iPS) cells.²⁸ The mechanisms of reprogramming are under intense investigation, but it has been demonstrated that iPS cells are more ES cell-like when they are selected based on the expression of genes involved in the maintenance of pluripotency such as Nanog or Oct3/4.^16,21,31 Subsequent studies demonstrated that iPS cells can be generated from somatic cells without the use of c-Myc, thereby minimizing the risk for tumor formation, as well as by morphology selection criteria to obviate the need for Nanog or Oct3/4 promoter driven transgenes.^17,22 Such iPS cells contributed widely to a variety of tissues in a chimeric mouse assay including the germline, and mice derived from germline-transmitted iPS cells have been successfully created.²¹ More recently, human iPS cells have been generated using the same or similar factors as those used in the derivation of mouse iPS cells.^23,29,35 As with the discovery of human ES cells nine years ago, the availability of human iPS cells has triggered a rapidly growing interest from basic as well as trans- lational/clinical scientists in the potential of these cells to serve as in vitro disease models for drug discovery and pathway identification, and as genetically-matched cells for regenerative therapies.

Although iPS cells appear similar to ES cells in their global epigenetic state and ability to contribute to the three germ layers in chimeric assays,^16,19 their capacity to maintain appropriate differentiation in vitro has not been critically examined. This is important because chimeric assays select for those pluripotent cells that are euploid and diffe- rentiate appropriately in vivo. In vitro differentiation of ES or iPS cells, on the other hand, does not exert such selection pressure. In fact, cells that are chromosomally abnormal or epigenetically altered tend to out-compete the growth of normal cells in vitro and are prone to generate tumors in vivo.¹³ Furthermore, a chimeric assay re-

(6)

veals developmental deficits primarily in a cell autonomous manner whereas in vitro differentiation can reveal deficits in differentiation by both cell intrinsic and extrinsic mechanisms. Indeed, despite the wide-spread tissue contribution of iPS cells in chimera assays, the derivation of an “all iPS cell” derived live born mice by tetraploid complementation assays has been unsuccessful thus far. Hence, a direct comparison of the in vitro differentiation capacities of iPS and ES cells may reveal differences that are not readily apparent in chimeric experiments in vivo.

To examine the biological properties of cardiac lineage cells during development, we, along with others, have recently described the identification, isolation, and characterization of multipotent cardiac progenitor cells (CPCs) from the developing mouse embryo and in vitro differentiated ES cells.^20,32 Utilizing the earliest expression of the key cardiac transcription factor Nkx2.5, we showed that CPCs purified from embry- onic mice are capable of in vitro expansion and spontaneous differentiation into car- diomyocytes and smooth muscle cells. The expression of Nkx2.5, marked by transgenic expression of eGFP, is initiated at the CPC stage of development and continues as these cells mature into primitive cardiomyocytes as evidenced by the presence of a spontaneously beating eGFP+ heart tube. Indeed the fate of the cardiac lineage cell population from CPCs to immature cardiomyocytes can be followed in these Nkx2.5- eGFP transgenic mice during embryonic cardiogenesis as well as in pluripotent stem cell differentiation assays in vitro when ES or iPS cells are derived from these mice.

In this study we analyzed the biological properties of cardiac lineage cells derived from ES and iPS cells carrying the Nkx2.5-eGFP transgene. Comparison of these cells at the undifferentiated state revealed similarities in cell morphology, expression of pluripotency markers, growth behavior as well as contribution into all three germ layers in chimera assays. Upon in vitro differentiation, both ES and iPS cells expressed germ layer, cardiac progenitor, and immature cardiomyocyte markers in a developmentally appropriate fashion. Interestingly, despite their ability to beat spontaneously in vitro, iPS cell-derived cardiomyocytes exhibited impaired maturation into diffe- rentiated cardiomyocytes after transplantation into a myogenic environment in vivo.

Analysis of myogenic differentiation pathways in iPS cells revealed an over-expression of Delta-like 1 (Dlk1) that is associated with decreased beating and sarcomeric disorganization, phenotypes that could be reversed by inhibition of Notch signaling.

Taken together, these studies demonstrate that iPS cells are similar to ES cells but further improvements of the reprogramming procedure will be required to ensure the appropriate maturation of cardiac iPS cells for clinical applications.

(7)

Materials and Methods

Derivation of Nanog-P-I-G/Nkx2.5-eGFP and Nkx2.5-eGFP/ROSA-LacZ (NNG) iPS Cell Lines

Mice carrying either Nanog-G-I-P¹⁶ or constitutively β-galactosidase expressing alle- les (ROSA26-LacZ)⁹ were mated with Nkx2.5-eGFP transgenic mice to generate mice that are compound heterozygous for these reporter genes. Tail tip fibroblasts were derived as previously described and either infected with retroviral supernatant or concentrated, doxyxcycline-inducible lentiviruses encoding c-mycT58A, Klf4, Oct4 and Sox2 together with a constitutive lentivirus expressing the reverse tetracycline- dependent transactivator.^16,27 Colonies were selected on morphological criteria 3-4 weeks after infection and stable iPS lines established as previously described.^16,27 Derivation of Nkx2.5-eGFP (NK) and Rosa26-LacZ/Nkx2.5-eGFP (RN) ES cell lines Nkx2.5-eGFP transgenic mice were generated as previously described.³² These mice were either interbred or crossed with Rosa26-LacZ mice (Jackson Lab, Bar Harbor, ME). Embryonic day 3 blastocysts were isolated and cultured in Dulbecco’s Modi- fied Eagle Medium (DMEM) supplemented with 20% knock-out serum replacement (Invitrogen, Carlsbad, CA), 200 mM L-Glutamine, 5,000 i.u./mL penicillin/strepto- mycin, 5 mM non-essential amino acid, and 10,000 units/mL leukemia inhibitory factor after removal of zona pellucida with 100 mM acidic Tyrodes’s solution. Follo- wing 10 days of culture, each single ES cell-like colony that emerged from the plated blastocyst (~60-80% efficiency) were trypsinized and cultured in the same medium.

Multiple ES cell –like colonies emerged after an additional ~2-5 days and were subsequently expanded in the ES culturing medium.

Cell Culture and In Vitro Differentiation of ES and iPS cells

ES cells and iPS cells were maintained on growth arrested embryonic fibroblasts according to a previously published protocol.⁸ These cells were then adapted to gela- tin-coated dishes in the presence of leukemia inhibitory factor for two days prior to differentiation, On the day of differentiation, cells were digested with trypsin/EDTA for 5 min into single cells and plated as hanging droplets in suspension (~500 cells/

droplet) in differentiation media (DM) containing Iscov’s Modified Eagle Medium with high glucose, 20% FCS, 200 mM L-glutamine, 1-thioglycerol (1.5x10^-4M), and ascorbic acid (50 μg/mL) until the indicated time point. In experiments where cells were differentiated until day 18, media was replaced every 3 days starting from day 10. In experiments involving DAPT treatment, ES and iPS cells were re-plated into gelatinized glass chambers at day 8 of differentiation and incubated with DAPT at the indicated concentrations until day 16 when they were studied.

(8)

Isolation of Nkx2.5+ CPCs from iPS Cells

NkxZ iPS cells were in vitro differentiated as described above. On day 6 of differentiation EBs were harvested and digested with collagenase solution (collagenase A (10 mg/mL) and B (10 mg/mL) (Roche Diagnostics, Indianapolis, IN) in 10 mM Hepes buffed saline (HBS) and 20% FCS) for 60 min at 37⁰C with gentle triturition every 15 min.

Single cells were obtained by passing the cell preparation through a 40 μM strainer.

eGFP+ live cells (as defined by the lack of propidium iodine (PI) staining) were isolated using a FACSAria (BD Biosciences, San Jose, CA) sorter and cultured in DM.

Flow cytometry data was acquired by CellQuest™ v3.3 (BD Biosciences, San Jose, CA) and processed using FlowJo® v4.6.2 (Tree Star, Ashland, OR) software.

Immunohistochemical and Immunofluorescence Analysis of ES and iPS cells in Culture and after Hindlimb Transplantation

Undifferentiated ES and iPS cells and FACS-purified eGFP+ CPCs were fixed in 4%

paraformaldehyde in phosphate buffered saline (PBS) for 1 minute and stained with antibodies for either Oct3/4, Sox2 (Santa Cruz Biotechnology, Santa Cruz, CA), Nanog (CosmoBio Co., Tokyo, Japan), SSEA-1 (Developmental Studies Hybridoma Bank, Ames, IA), cardiac troponin T (Neomarkers, Fremont, CA), smooth muscle α-actin (Dako, Carpenteria, CA), smooth muscle myosin heavy chain (Biomedical Technologies Inc., Brockton, MA), PECAM (CD31) (BD Biosciences, San Jose, CA), or α-sarcomeric actinin (Sigma, St. Louis, MO) as required in each experiment and according to manufacturer’s recommended protocols. Frozen sections from CPC transplanted hindlimbs were stained overnight with X-gal (1 mg/mL) and counterstained with Nuclear Fast Red. Antibodies used for cardiac troponin T, smooth muscle α-actin, and smooth muscle myosin heavy chain staining in these frozen sections are identical to those described above.

Qualitative and Quantitative Analysis of Gene Expression by PCR

Undifferentiated ES or iPS cells or FACS-purified and cultured cells derived from these were isolated at various time points during in vitro differentiation and added to Trizol® reagent (Invitrogen, Carlsbad, CA) and stored at -20⁰C until processed. Total RNA from each sample was purified from the Trizol® prep using the SV Total RNA kit (Promega, Madison, WI) according to manufacturer’s suggested protocol. Conven- tional and real-time PCR were performed on cDNA made from reverse transcribed RNA using the I-script® cDNA synthesis kit (BioRad, Hercules, CA). Conventional PCR was performed using the Hotstar® PCR reagent (Qiagen, Valencia, CA) within the linear range of amplification (25-30 cycles) for each primer. Real-time PCR was performed using the Mastercycler® (Eppendorf, Hamburg, Germany) with SYBR®

(9)

Green substrate (BioRad, Hercules, CA) for 40 cycles. Primers sequences are availa- ble on request.

Mouse Hindlimb Cell Transplantation Assay

eGFP+ cells from day 6 of in vitro differentiated ES (NK5-2) and iPS (NkxZ35) cells were isolated by FACS and immediately injected into the hindlimbs of anesthesized isogenic mice (2.5x10⁵ cells per injection). At 28 days (N=8 each) after transplantation, mice were sacrificed and their hindlimbs harvested and immediately fixed in 2% paraformaldehyde and 30% sucrose overnight before frozen section and X-Gal staining with Nuclear Fast Red counterstaining. All animal experiments described in this paper were approved by the Subcommittee on Research Animal Care (SRAC) of the Massachusetts General Hospital.

Electrophysiological Studies of Differentiated CPCs

In vitro differentiated ES and iPS cells were dispersed into single cells on day 9 by collagenase treatment as described above and plated onto No. 1 round coverslips in 20-well plate in DM. The following day, eGFP+ cells on coverslips were assessed for their spontaneous action potentials as previously described.³²

Bisulfite Sequencing Analysis of Dlk1 3’-IG-DMR

Genomic DNA was isolated from similar passage ES and iPS cells by lysis in Tris (0.1 M, pH 8.5), NaCl (0.2 M), EDTA (5 mM), 0.2% SDS, proteinase K (0.2 mg/ml) overnight at 520C, extraction with phenol-chloroform and isopropanol precipitation.

Bisulfite treatment was performed on 1 ug DNA from each sample using the CpGe- nome Fast DNA Modification Kit (Chemicon, Billerica, MA) according to the manufacturer’s instruction. The converted DNA was subjected to two rounds of PCR as described with primers: IG-Out-For: 5’-GTGTTAAGGTATATTATGTTAGTG- TTAGG-3’; IG-In-For: 5’-ATATTATGTTAGTGTTAGGAAGGATTGT-3’ and IG- Rev: 5’-TACAACCCTTCCCTCACTCCAAAAATT-3’.¹² The amplified fragments were subcloned into the pGEMT-Easy vector (Promega, Madison, WI). Plasmid DNA from individual clones for each fragment was isolated with QIAprep Spin Mini- prep Kit (Qiagen, Valencia, CA) and sequenced using SP6 primer. Dnmt1 knock-out (Dnmt1 KO) ES cells serves as negative control for methylation in this DMR.

Statistical Analysis

Statistical analysis for all paired-sample comparisons described has been performed using the 2-tailed Student’s t-test with unequal variance. Analysis of the relationship between Dlk1 expression and the frequency of beating embryoid bodies was performed using linear regression.

(10)

Results

Derivation of Nkx2.5-eGFP Transgenic ES and iPS Cell Lines

To derive pluripotent iPS cell lines in which cardiac differentiation can be readily vi- sualized, we generated compound heterozygous mice by cross breeding the previously described Nanog-eGFP-IRES-Puro mice,¹⁶ referred to here as Nanog-G-I-P, with the cardiac-specific Nkx2.5-eGFP mice (Figure 1A). Adult tail fibroblasts from these mice were isolated, expanded in vitro and transduced with retroviruses expressing Oct3/4, Sox2, Klf4, and c-MycT58A (the constitutively active T58A mutant). Three to four weeks after infection,²⁰ puromycin-resistant colonies (from 1.0x10⁵ fibroblasts) emerged; 9 clones were expanded into lines that exhibited ES-like morphology and growth characteristics. The two Nanog-G-I-P/Nkx2.5-eGFP (NNG) iPS cell lines that exhibited the most robust eGFP expression (driven by Nanog) were further investi- gated. We also generated iPS cell lines carrying both the constitutive β-galactosidase expressing Rosa26-LacZ allele and the Nkx2.5-eGFP transgene (Figure 1A). This al- lowed us to be able to isolate CPCs from iPS cells based on their eGFP expression and track the fates of these cells in transplant recipients by detecting the presence of β-galactosidase activity. To minimize the possibility of reactivation of the four factors (particularly c-Myc) in the transplant setting, tetracycline inducible lentiviral expression vectors were used.²⁷ These Rosa26-LacZ/Nkx2.5-eGFP (NkxZ) iPS cells were selected based on their ES cell-like morphology without the use of a drug-resistance transgene as previously described.^16,17 We derived a total of 14 NkxZ iPS cell lines and 8 were further selected for expansion. To derive Nkx2.5-eGFP (NK) and Rosa26- LacZ/Nkx2.5-eGFP (RN) ES cell lines, day 3.5 blastocysts from either Nkx2.5-eGFP or double heterozygous Rosa26-LacZ/Nkx2.5-eGFP transgenic mice were isolated and cultured in the presence of leukemia inhibitory factor (Figure 1B). Fifteen NK and RN ES cell lines were generated and five lines were selected for further studies.

As expected, undifferentiated NK ES and NkxZ iPS cells are eGFP- at baseline since Nkx2.5 expression is silent in pluripotent stem cells (Figure 1C). NNG iPS cells, on the other hand, are eGFP+ at baseline due to the expression of Nanog in these cells (Figure 1C). To confirm that the selected iPS cell lines are appropriately reprogram- med, we first show that the expression of pluripotency markers such as Oct3/4, Sox2, and SSEA1 are similar between ES and iPS cells (Figure 1D). Furthermore, to directly demonstrate their developmental pluripotency, we introduced NkxZ iPS cells into mouse embryonic day 3.5 blastocysts and isolated chimeric embryos at embryonic day 10.5 (e10.5). X-Gals staining of frozen sections from these embryos demonstrated the contribution of iPS cells to tissues encompassing all three germ layers including brain, heart, skeletal muscle, bone, liver, and intestines (Figure S1B).

(11)

Characterization of In Vitro Cardiac Lineage Differentiation by Nkx2.5-eGFP iPS and ES Cells

Since NNG iPS cell lines express eGFP both as undifferentiated cells (indicating Na- nog expression) and when differentiated into CPCs (indicating Nkx2.5 expression), we examined a number of surface markers that might allow us to distinguish between these two cell populations and found that SSEA-1 alone could serve this purpose. As shown in Figure 2A, undifferentiated (day 0) NK ES cells are nearly homogenously SSEA-1+ but eGFP-. On the other hand, at day 7 of differentiation eGFP+ cells are uniformly SSEA-1-. Consistent with this observation, eGFP+ cells from NNG iPS cells are predominately SSEA-1+ at the onset of differentiation (day 0), but become SSEA-1- when committed to the cardiac lineage (day 7). The presence of eGFP- cells within SSEA-1+ cells at day 0 is most likely due to the cyclical pattern of Nanog-

Figure 1. Derivation of Nkx2.5-eGFP iPS and ES cell lines. (A) Schematic diagram for the mouse bree- ding, fibroblast derivation, and viral infection to derive NNG and NkxZ iPS cell lines. (B) Schematic diagram for the derivation of NK and RN ES cell lines. (C) Brightfield and fluorescence microscopic analysis of eGFP expression in representative NNG and NkxZ iPS cell lines and an NK ES cell line.

(D Immunofluorescence analysis of NNG and NkxZ iPS cell lines and an NK ES cell line for their expression of pluripotency genes such as Oct3/4 (left panels), Sox2 (middle panels), SSEA1 (right panels).

(12)

eGFP expression in undifferentiated ES cells.⁵ The reciprocal relationship between the percentage of SSEA-1+/eGFP+ (i.e. undifferentiated) and SSEA-1-/eGFP+ cells (i.e. cardiac cells) in NNG iPS cells during in vitro differentiation is consistent with developmentally-regulated expression of these genes (Figure 2B). Differentiating iPS and ES cells robustly express eGFP on day 6 (Figure 2C) and subsequently develop into beating immature cardiomyocytes by day 8 (Figure 2C).

Figure S1. Characterization of germ layer contribution by NkxZ iPS cell lines. (A) Selected NkxZ iPS cell lines were cultured and stained with X-Gal to confirm the expression of β-galactosidase in these cells. (B) Chimeric analysis of germ layer contribution by NkxZ35 iPS cells. A schematic diagram of iPS cell injection into embryonic day 3.5 mouse blastocysts is shown (top panel). Bright field microscopy of three X-gal stained e10.5 chimeric embryos shows variable degree of chimerism in each embryo.

Histological sections of one of the LacZ+ embryos demonstrate iPS cell contribution into the bone, gut, heart, liver, brain, and skeletal muscles (bottom panels).

(13)

To examine the electrophysiological properties of cardiomyocytes derived from NNG iPS cells, we compared their spontaneous action potentials at day 10 of in vitro dif- ferentiation with those from NK ES cells (Figure 2D). Both iPS and ES cell-derived cardiomyocytes display spontaneous action potentials resembling atrial and ventricular cardiomyocytes.

Figure 2. In vitro differentiation of NNG iPS and NK ES cells. (B) Flow cytometric analysis of SSEA-1 and eGFP expression in undifferentiated (Day 0) (top left panel) and day 7 in vitro differentiated (top right panel) NK5-2 ES cells. Flow cytometric analysis of SSEA-1 and eGFP expression in undifferentiation (Day 0) (bottom left panel) and day 7 in vitro differentiated (bottom right panel) NNG 1-3 iPS cells. (B) Quantitative analysis of SSEA-1+eGFP+ and SSEA-1-eGFP+ cells from NNG 1-3 iPS cells during in vitro differentiation. (C) Expression of eGFP (within black dotted lines) in day 6 (left panels) and appearance of beating cells (within white dashed lines) in day 8 (right panels) in vitro differentiated NK5-2 ES (top pa- nels) and NNG1-3 iPS (bottom panels) cell lines.(D) Electrophysiological analysis of eGFP+ beating cells from day 10 in vitro differentiated NK5-2 ES (left panel) and NNG 1-3 iPS (right panel) cells.

(14)

In addition, some ES cells produce electrical profile consistent with sinoatrial node (SA node), whereas iPS cells differentiate into atrial ventricular node (AVN) cells.

This difference, however, is likely due to limited sampling of these rare cell populations rather than reflect biological differences between ES and iPS cells.

To assess the temporal patterns of gene expression between iPS and ES cells during in vitro differentiation, we measured the expression of pluripotent (Oct3/4 and Nanog), germ layer (Brachyury T), and cardiac (Nkx2.5, cardiac actin and MLC2v) markers in embryoid bodies isolated from different days of in vitro differentiation. As shown in Figure S2, except for some variations in the level of expression, these genes ap- pear with a similar kinetics during in vitro differentiation (Figure S2). Together, these results demonstrate that both ES and iPS cells undergo normal commitment into CPCs.

In Vitro Differentiation of FACS-purified CPCs from iPS Cells

We next sought to assess the developmental capacity of iPS cell-derived CPCs to differentiate spontaneously into the main cell lineages in the heart (e.g. cardiomyocytes, smooth muscle cells, endothelial cells). To this end, we isolated eGFP+ cells from in vitro differentiated NkxZ iPS cells on day 6 of differentiation and cultured these

Figure S2. Comparison of gene expression in differentiating NK ES and NNG iPS cells. Quantitative PCR analysis of pluripotency (Oct3/4, Nanog), mesoderm (Brachyury T), cardiac progenitor (Nkx2.5), and myocardial (Card. Actin, MLC2v) gene expression in NK5-2 ES (black bar), NNG 1-3 iPS (gray bar), and NNG 3-3 iPS (white bar) cells during in vitro differentiation.

(15)

cells for 7 additional days in differentiation media. As shown in Figure 3A, the FACS- purified cell population is highly enriched (>96%) for eGFP+ cells.

Figure 3. Isolation and characterization of CPCs derived from NkxZ iPS cells. (A) representative flow cytometry plot for the purification of eGFP+ CPCs from day 6 in vitro differentiated NkxZ35 iPS cells (left panel). Isolated cells were re-analyzed by flow cytometry to assess their purity (right panel). (B) Brightfield microscopy of sorted cells after plating. Note the high nuclear to cytoplasmic ratio of the purified cells (inset). (C)Immunoflurescence analysis of cultured eGFP+ CPCs for the expression of cardiac troponin T (cTnT) and smooth muscle myosin heavy chain (SM-MHC) (a). Panels (b) and (c) represent magnified images of the areas indicated in (a). A rare cell expressing the endothelial marker PECAM (arrowhead) was also identified among many smooth muscle actin-α expressing cells (arrow) (d). Scale bar – 100 uM in (a); 50 uM in (b,c,d). (D) Quantitative analysis of spontaneous cardiomyocyte (CMC) and smooth muscle cell (SMC) differentiation by FACS-purified CPCs from NkxZ35 iPS cells. eGFP+

cells were cultured for 7 days after sorting. The presence of differentiated CMC and SMC were identified based on immunostaining for cTnT and SM-MHC, respectively. Progenitor cells are identified by the absence of cTnT and SM-MHC staining. (E) RT-PCR analysis of cardiomyocyte (Card. Actin, MLC2v), smooth muscle (SM-MHC), and endothelial cell (CD31) marker expression in undifferentiated ES (ES0) and iPS (iP0) cells compared with eGFP+ CPCs that have been differentiated to day 13 (ES13 and iP13).

Std – mol. weight standard. H20 – water control. L – adult liver. H - adult heart. Mef – mouse embryonic fibroblast.

(16)

Morphologically, these cells appear immature with a low ratio of cytoplasm to nu- cleus (Figure 3B, inset). After 7 days of culture, some of these cells differentiated spon- taneously into cardiac troponin T (cTnT) positive cardiomyocytes while others differentiated into smooth muscle myosin heavy chain (SM-MHC)-expressing smooth muscle cells (Figure 3C, panels a-c). Interestingly, a rare number of PECAM positive cells (Figure 3C, panel d, arrowhead) were identified among smooth muscle α-actin positive cells (Figure 3C, panel d, arrows) suggesting residual developmental potency for the vascular lineages.

Figure 4. In vivo differentiation of purified CPCs from NkxZ iPS cells. Schematic diagram of cell sorting, injection, and harvesting of hindlimbs that have received FACS-purified CPCs from in vitro differentiated NkxZ35 iPS cells and NK5-2 ES cells. (B) Bright field microscopy of X-gal stained and Nuclear Fast Red counterstained sections of mouse hindlimbs at 28 days following cell transplantation. Transplanted cells that have differentiated into smooth muscle cells are shown in higher magnification (arrowheads).

(C) Serial sections from hindlimbs of transplanted NkxZ35 iPS cells and NK5-2 ES cells. Immunohisto- chemical analysis detected the expression of cardiac troponin T (cTnT) in ES cell-derived CPCs but not in NkxZ35 iPS cells. IgG – control antibody.

(17)

Overall, 13% of the sorted eGFP+ cells were able to spontaneously differentiate into cardiomyocytes (CMC) and 7% into smooth muscle cells (SMC) by immunostaining for cTnT and SM-MHC, respectively (Figure 3D). These percentages are similar to those obtained from differentiation of ES cell-derived CPCs.³² Analysis of gene expression in these sorted and subsequently cultured cells further confirmed their mul- ti-lineage differentiation as evident by the presence of myocardial (MLC2v), smooth muscle (SM-MHC), and endothelial (CD31) cell markers (Figure 3E).

Differentiation of iPS Cell-derived CPCs In Vivo

To examine the capacity of iPS cell-derived CPCs to survive, engraft, and differentiate in a myogenic environment in vivo, we isolated eGFP+ cells from day 6 in vitro dif- ferentiated NkxZ iPS and NK ES cells by FACS and transplanted these freshly derived cells into the hindlimbs of isogenic mice (2.5x10⁵ cells/limb) (Figure 4A). At 28 days after transplantation (N=8), these hindlimbs were harvested for analysis. The presence of engrafted iPS cells was readily apparent by X-gal staining as blue cells localized between skeletal myofibrils in 5 of 8 hindlimbs. iPS cell-derived CPCs that have differentiated into smooth muscle cells can be easily identified in the perivascular space (Figure 4B). Surprisingly, and in contrast to ES cell-derived CPCs, the engrafted CPCs derived from iPS cells rarely expressed cTnT, a marker of differentiated cardiomyo- cyes (Figure 4C), despite their apparent ability to do so in short term culture in vitro (Figure 3C).

Characterization of Cardiomyocyte Maturation in iPS Cells

Since the CPC transplantation studies assessed the cardiac phenotype of engrafted cells after 34 days from the onset of differentiation whereas our in vitro culture stu- dies lasted only 10-13 days, we speculated that a cardiomyocyte differentiation phe- notype may also be detectable at later time points in vitro. To determine the onset of divergent phenotypes between iPS and ES cells, we examined each stage of cardiac differentiation from pluripotent stem cells to mature cardiomyocytes. We found that despite the absence of an overall difference between various iPS and ES lines to differentiate into immature cardiomyocytes (as assessed by the frequency of beating cell clusters formed at day 8), there was a uniform decrease in the beating frequency in iPS cells by day 18 (Figure 5A and 5B). We asked whether this is due to a decreased number of cardiac cells in iPS cells and measured the percentage of eGFP+ cells at each stage of development. As shown in Figure 5C, the percentage of eGFP+ cells ap- peared similar between ES and iPS cells during the progenitor stage of differentiation at day 6 as assessed by flow cytometry. However, at days 10 and 18 the percentages of eGFP+ cells have decreased in iPS cells compared with ES cells. Since the intensity

(18)

of eGFP expression corresponds to the state of cardiac cell differentiation (i.e. eGFP low cells represent cardiac progenitors and eGFP high cells represent differentiated cardiomyocytes),³² we showed that the reduction in iPS cell-derived cardiac cells pre-

Figure 5. Impaired maturation of iPS cell-derived cardiomyocytes. (A) Comparison of beating cluster formation in embryoid bodies derived from multiple ES and iPS cell lines at day 8 (black bar) and day 18 (white bar) of in vitro differentiation. Each bar represents two independent experiments from each individual ES or iPS cell line. (B) Flow cytometry analysis of the percentage of eGFP+ cells from in vitro differentiated NK5-2 ES (top row) and NkxZ35 iPS (bottom row) cells on days 6 (left panel), 10 (middle panel), and 18 (right panel) of differentiation. (C) The mean and SEM of percentage of beating EBs at day 18 compared with day 8 for the six ES and six iPS cell lines studied in (A). (D) Quantitation of the percentage of eGFP+ progenitors (P) and cardiomyocytes (M) present in differentiating ES and iPS cells on days 6, 10, and 18. The values presented are mean and SEM of three different ES (NK1-35, NK5-1, NK5-2) and iPS (NNG1-3, NkxZ4, NkxZ35) cell lines from two independent experiments. (E) Bright field and fluores- cence microscopy of cardiomyocyte clusters from representative EBs from day 18 in vitro differentiated NK5-2 ES (top row) and NkxZ35 iPS (bottom row) cells.

E

D A

C

B

(19)

ferentially affects eGFP high cardiomyocytes and not eGFP low progenitor cells (Fi- gure 5D). Consistent with this, beating areas from iPS cell-derived embryoid bodies exhibited decreased eGFP signal intensity (Figure 5E). Furthermore, the expression of markers for differentiated cardiomyocytes, MLC2v and cTnT, was significantly redu- ced in iPS cells by day 18 (Figure S3), whereas the expression of a marker for imma- ture cardiomyocytes, cardiac α-actin, was only slightly decreased.

To address specifically whether cardiac iPS cells have an impaired ability to mature into well-differentiated cardiomyocytes, we assessed the pattern of sarcomeric striation in these cells. Sarcomeric striation has been shown to correlate precisely with the state of cardiomyocyte maturation in a developing heart and is disrupted in a failing heart.²⁶ In embryonic (e10.5) and well-differentiated cardiomyocytes from ES or iPS cells, immunostaining for α-sarcomeric actinin results in a fully striated pattern (Fi-

Figure S3. Comparison of gene expression in differentiating NK ES and NkxZ iPS cells. NK5-2 ES (black bar) and NkxZ35 iPS (white bar) cell lines were differentiated from day 0 to 18. At the indicated time points, embryoid bodies were harvested and their RNA purified for cDNA synthesis and quantitative PCR analysis for genes representing pluripotent (Oct3/4, Nanog), mesoderm (Bry T), and cardiac (eGFP, Nkx2.5, Card. Actin, cTnT, MLC2v) cells.

(20)

gure 6A). Although ES and iPS cells showed similar percentage of striated cells at day 8 (51±3.2% vs 46±3.9% of total α-sarcomeric actinin+ cells), by day 18, the percentage of striated cells in ES cells has increased to 72±0.3% while the percentage in iPS cells has declined to 25±10% (Figure 6B). This decline in the percentage of striated car- diomyocytes is consistent with the decreased beating frequency in day 18 embryoid bodies from iPS cells (Figure 5A and 5B) and is not due to a culturing artifact since cardiomyocytes from e10.5 heart showed no decline in striation from day 8 to 18 (Figure 6B). The observed phenotypic defect in iPS cell-derived cardiac cells might be due to a number of causes. We did not find a consistent link between the observed beating cardiac cell phenotype and passage number, sex origin, the type of virus used (retrovirus vs lentivirus), or the genetic background of the mice used to generate ES and iPS cell lines. (Table S1).

Figure 6. Analysis of maturational defect in of cardiac iPS cells. (A) Comparison of sarcomeric stria- tion in mouse embryonic day 10.5 cardiomyocytes and cardiomyocytes derived from ES and iPS cells.

(B) Quantitative analysis of sarcomeric striation in ES and iPS cell-derived cardiomyocytes. Embryonic day 10.5 cardiomyocytes (E10.5 CM) (black bar) were plated concurrently with in vitro differentiated ES (gray bar) and iPS (white bar) cells. These cells were stained with antibodies to sarcomeric α-actinin at days 8 and 18 of differentiation and quantitated by an observer blinded to the identity of the cell line. The data shown represents mean and SEM of three ES and three iPS cell lines. (C) Determination of prolifera- tion rates in differentiated cardiac cells derived from ES and iPS cells. The average rate of cardiac ES and iPS cell proliferation (as assessed by BrdU incorporation in day 10 eGFP+ cells) from six ES and six iPS cell lines is shown. (D) Assessment of apoptosis rates in differentiated cardiac cells derived from ES and iPS cells. The rate of cardiac ES and iPS cell apoptosis (as assessed by Annexin V staining in eGFP+ cells) from six ES and six iPS cell lines was measured at days 8 and 18 of in vitro differentiation.

(21)

The level of expression of the four viral transgenes also did not change significantly with differentiation and we failed to detect a positive correlation between the degree of viral transgene expression and severity of the cardiac phenotype (Figure S4). Next, we assessed the rate of proliferation of eGFP+ cardiac cells but found no statistically significant difference between iPS and ES cell lines (Figure 6C). However, we observed a 42±4% increase in apoptosis in iPS cell-derived cardiac cells on day 18 compared with ES cell-derived cardiac cells (Figure 6D). This difference, when accumulated over time, may account for some of the observed decrease in the number of differentiated cardiomyocytes in iPS cells.

Figure S4. Expression of viral four factor transcripts during iPS cell differentiation in vitro. Embryoid bodies from NK5-2 ES (dark gray), and NNG1-3 (light gray) and NNG3-3 (white) iPS cells were harvested and their RNA purified for cDNA synthesis at the indicated time points. Results for quantitative PCR analysis using primers specific for the viral transcripts are shown. RetroV-MEF (RM) represents fibroblasts that have been infected for 3 days with retroviruses expressing the four factors.

(22)

Reversal of Cardiac iPS Cell Defect by Chemical Inhibition of the Notch Signaling Pathway

To gain further insight into the molecular mechanism of the cardiomyocyte defect in iPS cells, we explored whether the reported developmental anomalies associated with animals cloned by nuclear transfer may play a role here. It has been shown that fetal and placental overgrowth is a common feature of embryos cloned from embryonic or somatic cells, presumably due to the misexpression of imprinted genes.⁷ Fetal overgrowth with skeletal muscle hypertrophy has also been described in mice and sheep that over express Delta-like 1 (Dlk1).^6,15 Interestingly, we found a similar increase in skeletal muscle differentiation in iPS cells as measured by the expression of skeletal muscle α-actin when compared with ES cells (Figure 7A). This increase in skeletal muscle differentiation is correlated with an elevated level of Dlk1 expression in iPS cells (Figure 7B).

Figure S5. Quantitation of 3’ IG-DMR methylation at the DLK1-GTL2 locus. Multiple ES and iPS cell lines were subjected to bisulfite treatment and PCR-based sequencing. The methylated CpG islands are shown as black circles. Unmethylated CpG island are white circles. Dnmt1 KO ES cells serve as negative control for 3’ IG-DMR methylation.

(23)

Since Dlk1 expression is regulated directly by the degree of methylation at the 3’ intergenic differentially methylated region (IG-DMR) of the DLK1-GTL2 locus on mouse chromosome 12,¹² we assessed its methylation status by bisulfite sequencing and found that all of the six iPS cell lines studied exhibited increased 3’-IG-DMR methylation (80-95%) compared with none of the four ES cell lines (65-67%) (Table S1 and Figure S5).

Figure 7. Involvement of Notch pathway in cardiac iPS cell defect. (A) Quantitative PCR analysis of the expression of skeletal actin in three ES (G74, FCNK1, NK5-2) and eight iPS (NNG1-3, NNG3-3, NkxZ4, NkxZ8, NkxZ13, NkxZ14, NkxZ17, NkxZ35) cell lines. The mean values and statistical analysis are shown in the right hand panel. (B) Quantitative PCR analysis of the expression of Dlk1 in the same ES and iPS cell lines as in (A). The mean values and statistical analysis are shown in the right hand panel. (C) Comparison of the expression of Dlk1 and Gtl2 in FACS-purified CPCs from ES and iPS cells. The bar graph represents fold over-expression of Dlk1 and Gtl2 in iPS cells compared with ES cells. (D) Correlation of Dlk1 expression with the reduction of beating embryoid bodies from days 8 to 18. Linear regression analysis was performed to generate the plotted line that best fits the data. (E) Notch pathway inhibition in differentiated cardiac iPS cells. iPS cells from three independent iPS cell lines (NNG1-3, NkxZ4, NkxZ35) were treated with 0 (white bar), 2 µM (gray bar), or 20 µM (black bar) of DAPT for 8 days. The mean changes in the percentage of eGFP+ cells (left panel), percentage of beating embryoid bodies (EBs) (middle panel), or percentage of striated cells (right panel) are shown.

(24)

We next assessed the level of Dlk1 expression specifically in cardiac cells by FACS- purification of eGFP+ cells at the progenitor stage of differentiation (day 6) and after these cells were differentiated into cardiomyocytes and smooth muscle cells (day 13).

We found that Dlk1 expression was elevated in both progenitors (Figure 7C) and dif- ferentiated cardiac cells (Figure S6) derived from iPS cells compared with ES cells.

Consistently, the expression of Gtl2, a well-characterized non-coding transcript lo- cated distal to the Dlk-1 locus on chromosome 12, was down regulated in iPS cells (Figure 7C).

Figure S6. Comparison of the expression levels of Dlk1 and Gtl2 in differentiated cardiac cells from ES and iPS cells. eGFP+ CPCs were sorted on day 6 from in vitro differentiated NK5-2 ES and NkxZ35 iPS cells and cultured for 7 additional days before they were harvested for RNA purification and quantitative and semi-quantitative PCR analysis for the expression of Dlk1 and Gtl2 transcripts. Std – mol. weight standard. H– e13.5 heart tissue. H2O – water control.

Table S1. Characteristics of iPS and ES cell lines studied.

(25)

This reciprocal regulation of Dlk1/Gtl2 expression has previously been described and is attributed to methylation of CpG islands at the 3’ IG-DMR of the DLK1-GTL2 locus.¹² To determine whether the elevated level of Dlk1 expression may be mecha- nistically linked to the cardiomyocyte maturation defect in iPS cells, we performed a linear regression analysis between Dlk1 expression levels and the percentages of reduction of beating embryoid bodies from day 8 to day 18 of differentiation. As shown in Figure 7D, there appears to be a closely associated relationship (R² >0.85) between Dlk1 expression and beating cardiomyocyte phenotype.

Since Dlk1 mediates Notch signaling and Notch pathway inhibition is required for full cardiomyocyte maturation in vivo,^4,24 we assessed whether the defect in cardiac iPS cells can be rescued by disrupting Notch signaling with DAPT, a γ-secretase in- hibitor which blocks the release of Notch intracellular domain (NICD) from plasma membrane thus preventing the ability of NICD to transactivate gene expression in the nucleus.¹¹ As shown in Figure 7E, treatment of iPS cells with DAPT from days 8 to 16 of in vitro differentiation resulted in an increased percentage of eGFP+ cells (left panel), as well as a recovery in the frequency of beating embryoid bodies (middle panel), and sarcomeric striation (right panel). Treatment of differentiating ES cells with DAPT, on the other hand, resulted in minimal change in these parameters (data not shown).

Discussion

Here we performed a detailed comparison of the developmental properties of iPS and ES cell-derived cardiac lineage cells. In particular, we found that early germ layer differentiation to cardiac lineage commitment is similar between iPS and ES cells during

Table S2. Bisulfite analysis of ES and iPS cell lines at the 3’-IG-DMR of Dlk1.

(26)

in vitro differentiation. Interestingly, we observed an impairment of late cardiomyo- cyte maturation in iPS cells compared with those derived from ES cells. This difference was characterized by an increased apoptosis and decreased sarcomeric striation in cardiac iPS cells compared with those from ES cells. Further analysis revealed that over-expression of Dlk1 is at least partly responsible for the observed phenotype since inhibition of Notch signaling in iPS cells can reverse this maturational defect. Taken together, these data demonstrate a high degree of similarity between ES and iPS cells during early cardiac development in vitro, but a divergence in their behavior during late cardiomyocyte maturation. These data suggest that improvements in reprogramming by using a delivery system that does not require viral integration, or starting with a different cell source, or different factors will be needed to ensure the derivation of iPS cell lines that are capable of proper cardiac cell maturation for clinical and pharmaceutical applications.

Developmental Similarity Between ES and iPS Cells

The derivation of pluripotent stem cells by the introduction of defined factors into differentiated somatic cells constitutes a major breakthrough in our understanding of cell lineage commitment. Specifically, the derivation of iPS cells has demonstrated that epigenetic control mechanisms regulating tissue-specific differentiation can be reversed. Based on the level of pluripotency gene expression and the ability to contribute to tissues from all three germ layers, undifferentiated iPS cells are highly similar to ES cells (Figure 1).^16,21,31 We found in this study that in vitro differentiation of iPS cells into germ layers and subsequently, cardiac progenitor cells is indistinguishable from ES cells. This is consistent with the global remodeling of epigenetic marks during iPS cell derivation from adult fibroblasts,¹⁶ and the observation that iPS cells that are incompletely reprogrammed show impaired ability to contribute to germ layers.²⁸ In addition, using a cardiac lineage reporter, we found that the commitment of iPS cells from mesoderm into cardiac progenitor cells appears similar to those from ES cells.

This finding is of critical importance since the transplantation of lineage committed progenitor cells is likely to be more advantageous than transplanting differentiated cardiomyocytes given that progenitor cells are more likely to expand after engraftment.³³ In addition, lineage committed progenitor cells are less likely to generate teratoma compared with undifferentiated iPS cells when transplanted. Consistent with this, we observed no teratoma formation in hindhimbs injected with FACS-purified Nkx2.5+

progenitor cells (Figure 4). The absence of teratoma formation was also recently repor- ted in iPS cell-derived hematopoietic progenitor cell transplantation where iPS cells were purified according to their expression of hematopoietic markers.¹⁰

(27)

Biological Differences Between ES and iPS Cells

While the biological similarity between ES and iPS cells in early cardiac lineage induction is encouraging, we found that cardiomyocytes derived from iPS cells exhibited impaired maturation compared with those derived from ES cells. This difference was characterized by increased apoptosis and decreased sarcomeric striation in iPS cell-derived cardiomyocytes. These differences contributed to the decreased number of functional cardiomyocytes during late stage differentiation (day 18) and consequently, decreased frequency of beating embryoid bodies (Figure 5A). The me- chanism responsible for the impaired maturation of cardiomyocytes from iPS cells appears distinct from the mechanism responsible for the increased cardiac cell apoptosis. We found that DAPT treatment reversed the sarcomeric striation defect and improved beating embryoid body formation. However, DAPT treatment was unable to correct the apoptosis phenotype (data not shown). This increased apoptosis may be due to a number of causes including the persistent low-level expression of one or several of the viral transgenes.

To explore further the molecular mechanism responsible for the defect observed in cardiac iPS cells, we postulated that developmental abnormalities described in cloned embryos may share similar features with in vitro differentiated iPS cells. Indeed, we found that increased skeletal muscle differentiation is present in day 18 iPS cells but not ES cells. Accordingly, Dlk1 expression, which regulates skeletal muscle maturation, was elevated in most of the iPS cell lines studied. We demonstrated that the CpG islands at the critical 3”-IG-DMR of the DLK1-GTL2 locus are hypermethylated in iPS cells, consistent with their increased Dlk1 expression. Furthermore, we showed that the effect of Dlk1 over expression is mediated by Notch signaling since the inhibition of Notch pathway by DAPT treatment in iPS cells reversed the striation and beating cardiomyocyte phenotypes (Figure 7E).

We currently do not know the exact mechanism responsible for the increased DNA methylation at the Dlk1 locus. iPS cells may be incompletely reprogrammed despite their demonstration of pluripotency in chimeric assays or are more susceptible to epi- genetic modifications during in vitro culturing than ES cells. Indeed, recent studies suggest that hypermethylation is a feature of partially reprogrammed cells as well as cells that have been extensively cultured.^18,19 These differences between iPS and ES cells that we observed underscore the need to study both cell types in parallel if the full biological and therapeutic potentials of pluripotent cells are to be realized.

iPS Cells as In Vitro Model of Cardiac Disease

ES cell in vitro differentiation has served in recent years as a useful platform to inves- tigate early stage organ development in mouse and human. The derivation of human

(28)

iPS cells raises the possibility of generating patient-specific pluripotent stem cells for in vitro disease modeling and for autologous cell-based therapy. However, few studies thus far have demonstrated a faithful recapitulation of the disease phenotype observed in differentiated ES cells with those found in patients, particularly for adult onset de- generative diseases. The impaired cardiac maturation of differentiated iPS cells raises caution as to their application in modeling cardiac disease using patient-specific iPS cells. A difference in developmental potency between ES and iPS cells was previously shown by the inability of iPS cells to give rise to live born mice by tetraploid complementation.³¹ In fact, these mice die generally between embryonic days 9-12 and fail to survive beyond embryonic day 14.5, a developmental failure most commonly attributed to cardiovascular defects. Indeed, congenital heart diseases such as atrial and ventricular septal defects have been observed in iPS cell-derived tetraploid embryos (unpublished data - Fujiwara, Y. and Orkin, S). A detailed comparison of cardiac li- neage differentiation between human ES and iPS cells will be required to determine whether a similar cardiomyocyte maturation defect exists.

Conclusion

Despite the high degree of similarity between undifferentiated iPS and ES cells described thus far, there is evidence for differences between these cells. Specifically, iPS cells undergoing cardiogenic differentiation exhibit impaired maturation into differentiated cardiomyocytes. These results suggest that our current technology for factor-based reprogramming is imperfect and could be improved with different factors or non-viral delivery system. The use of iPS cells in a clinical/translational setting should take place only after these limitations have been eliminated.

Acknowledgments

We thank Patricia Follett for assistance with chimeric mice assays, Laura Prickett-Ri- ce and Katherine Folz for expert FACS operation, Drs. Robert Blaustein and Chuang Du for expert assistance with the electrophysiology studies, and Dr. Joy Wu for ma- nuscript critiques. Financial support was provided by the Interuniversity Cardiology Institute of the Netherlands, the Dutch Heart Foundation and the Leiden University Fund (to D.A.P.), GlaxoSmithKline Research and Education Foundation and NHLBI HL081086 (to S.M.W.), NIH OD003266 (to K.H.), and the Howard Hughes Medical Institute (to S.H.O.).

None of the authors have a conflict of interest related to this work.

(29)

References

1. Aoi, T, Yae, K., Nakagawa, M., Ichisaka, T., Okita, K., Takahashi, K., Yamanaka, S. (2008) Generation of Pluripotent Stem Cells from Adult Mouse Liver and Stomach Cells. Science. 321, 699-702.

2. Boyer, L,A., Lee, T.I., Cole, M.F., et al. (2005) Core transcriptional regulatory circuitry in human em- bryonic stem cells. Cell. 122, 947-956.

3. Boyer, L.A., Plath, K., Zeitlinger, J. et al. (2006) Polycomb complexes repress developmental regula- tors in murine embryonic stem cells. Nature. 441, 349-353.

4. Bray, S.J., Takada, S., Harrison, E., Shen, S-C., Ferguson-Smith, A.C. (2008) The atypical mammalian ligand Delta-like homologue 1 (Dlk1) can regulate Notch signaling in Drosophila. BMC Develop- mental Biology. 8, 11.

5. Chambers, I., Silva, J., Colby, D., Nichols, J., Nijmeijer, B., Robertson, M., Vrana, J., Jones, K., Grote- wold, L., Smith. A. (2007) Nanog safeguards pluripotency and mediates germline development.

Nature. 450, 1230-1234.

6. Davis, E., Jensen, C.H., Daa Schroder, H., Farnir, F., Shay-Hadfield, T., Kliem, A., Cockett, N., Georges, M., Charlier, C. (2004) Ectopic expression of Dlk1 protein in skeletal muscle of padumnal heterozygotes causes the callipyge phenotype. Curr. Biol. 14, 1858-1862.

7. Eggan, K., Hidenori, A., Loring, J., Jackson-Grusby, L., Klemm, M., Rideout, W.M. 3rd,

Yanagimachi, R., Jaenisch, R. (2001) Hybrid vigor, fetal overgrowth, and viability of mice derived by nuclear cloning and tetraploid complementation. Proc. Natl. Acad. Sci. USA. 98, 6209-6214.

8. Fehling, H.J., Lacaud, G., Kubo, A., Kennedy, M., Robertson, S., Keller, G., Kouskoff, V. (2003) Trac- king mesoderm induction and its specification to the hemangioblast during embryonic stem cell dif ferentiation. Development. 130, 4217-4227.

9. Friedrich, G., Soriano, P. (1991) Promoter traps in embryonic stem cells: a genetic screen to identify and mutate developmental genes in mice. Genes Dev. 5,1513-23.

10. Hanna, J., Wernig, M., Markoulaki, S. et al. (2007) Treatment of sickle cell anemia mouse model with iPS cells generated from autologous skin. Science. 318, 1920-1923.

11. Hayward, P., Kalmar, T., Arias, A.M. (2008) Wnt/Notch signalling and information processing during development. Development. 135, 411-424.

12. Hiura H, Komiyama J, Shirai M, Obata Y, Ogawa H, Kono T. (2007) DNA methylation imprints on the IG-DMR of the Dlk1-Gtl2 domain in mouse male germline. FEBS Lett. 581, 1255-1260.

13. Hochedlinger, K. and Jaenisch, R. (2006) Nuclear reprogramming and pluripotency.

Nature. 441,1061-1067.

14. Kim, J., Chu, J., Shen, X., Wang, J., Orkin, S.H. (2008) An extended transcriptional network for pluri- potency of embryonic stem cells. Cell. 132, 1049-1061.

15. Lin, S.P., Coan, P., da Rocha, S.T., Seitz, H., Cavaille, J., Teng, P.W., Takada, S., Ferguson-Smith, A.C.

(2007) Differential regulation of imprinting in the murine embryo and placenta by the Dlk1-Dio3 imprinting control region. Development. 134, 417-426.

16. Maherali, N. Sridharan, R., Xie, W. et al (2007) Directly reprogrammed fibroblasts show global epi- genetic remodeling and widespread tissue contribution. Cell Stem Cell. 1, 55-70.

17. Meissner, A., Wernig, M., Jaenisch, R. (2007) Direct reprogramming of genetically unmodified fibro- blasts into pluripotent stem cells. Nat Biotechnol. 25, 1177-1181.

18. Meissner, A., Mikkelsen, T.S., Gu, H., et al (2008) Genome-scale DNA methylation maps of pluripo- tent and differentiated cells. Nature. 454, 766-770.

(30)

19. Mikkelsen, T.S., Hanna, J., Zhang, X., Ku, M., Wernig, M., Schorderet, P., Bernstein, B., Jaenisch, R., Lander, E.S., Meissner, A. (2008) Dissecting direct reprogramming through integrative genomic ana- lysis. Nature. 454, 49-55.

20. Moretti, A., Caron L, Nakano A, et al. (2006) Multipotent embryonic isl1+ progenitor cells lead to cardiac, smooth muscle, and endothelial cell diversification. Cell. 127, 1151-1165.

21. Okita, K., Ichisaka, T., Yamanaka, S. (2007) Generation of germline-competent induced pluripotent stem cells. Nature. 448, 313-317.

22. Nakagawa, M., Koyanagi, M., Tanabe, Kl, Takahashi, K., Ichisaka, T., Aoi, T., Okita, K., Mochiduki, Y., Takizawa, N., Yamanaka, S. (2008) Generation of induced pluripotent stem cells without Myc from mouse and human fibroblasts. Nat Biotechnol. 26, 101-106.

23. Park, I., Zhao, R., West, J.A., Yabuuchi, A., Huo, H., Ince, T.A., Lerou, P.H., Lensch, M.W., Daley, G.Q. (2008) Reprogramming of human somatic cells to pluripotency with defined factors. Nature.

451, 141-146.

24. Schroeder, T., Fraser, S.T., Ogawa, M., Nishikawa, S., Oka, C., Bornkamm, G.W., Nishikawa, S-I., Honjo, T., Just, U. (2003) Recombination signal sequence-binding protein Jκ alters mesodermal cell fate decisions by suppressing cardiomyogenesis. Proc. Natl. Acad. Sci. U.S.A. 100, 4018-4023.

25. Rowland, B.D., Peeper, D.S. (2006) KLF4, p21, and context-dependent opposing forces in cancer.

Nat. Rev. Cancer. 6, 11-23.

26. Seguchi, O., Takashima, S., Yamazaki, S. et al (2007) A cardiac myosin light chain kinase regulates sarcomere assembly in the vertebrate heart. J. Clin. Invest. 117, 2812-2824.

27. Stadtfeld, M., Maherali, N., Breault, D.T., Hochedlinger, K. (2008) Defining molecular cornerstones during fibroblast to iPS cell reprogramming in mouse. Cell Stem Cell. 2, 1-11.

28. Takahashi, Y., and Yamanaka, S. (2006) Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell. 126, 663-676.

29. Takahashi, Y., Tanabe, K., Ohnuki, M., Narita, M., Ichisaka, T., Tomoda, K., Yamanaka, S. (2007) In- duction of pluripotent stem cells from adult human fibroblasts by defined factors. Cell. 131, 861-872.

30. Wang, J.L., Rao, S., Chu, J., Shen, X., Levasseur, D.N., Theunissen, T.W., Orkin, S.H. (2006) A protein interaction network for pluripotency of embryonic stem cells. Nature. 444, 364-368.

31. Wernig, M., Meissner, A., Foreman, R. et al (2007) In vitro reprogramming of fibroblasts into a pluri- potent ES-cell-like state. Nature. 448, 318-324.

32. Wu SM, Fujiwara Y, Scibulsky SM, Clapham DE, Lien C-L, Schultheiss TM, Orkin SH. (2006). Deve- lopmental origin of a bi-potential myocardial and smooth muscle cell precursor in the mammalian heart. Cell. 127,1137-1150.

33. Wu, S.M., Chien, K.R., Mummery, C. (2008) Origins and fates of cardiovascular progenitor cells. Cell. 132, 537-543.

34. Wutz, A., and Barlow, D.P. (1998) Imprinting of the mouse Igf2r gene depends on an intronic CpG island. Mol Cell Endocrinol. 140, 9-14.

35. Yu, J., Vodyanik, M.A., Smuga-Otto, K. et al. (2007) Induced pluripotent stem cell lines derived from human somatic cells. Science. 318,1917-1920.

(31)