SEARCH

SEARCH BY CITATION

Keywords:

  • Reprogramming;
  • Induced pluripotency;
  • iPS;
  • Pluripotent stem cells;
  • Progenitor cells;
  • Hepatocyte differentiation;
  • Cardiac;
  • Neural induction

Abstract

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

Direct reprogramming of differentiated cells to induced pluripotent stem (iPS) cells by ectopic expression of defined transcription factors (TFs) represents a significant breakthrough towards the use of stem cells in regenerative medicine (Takahashi and Yamanaka Cell 2006;126:663–676). However, the virus-mediated expression of exogenous transcription factors could be potentially harmful and, therefore, represents a barrier to the clinical use of iPS cells. Several approaches, ranging from plasmid-mediated TF expression to introduction of recombinant TFs (Yamanaka Cell 2009;137:13–17; Zhou, Wu, Joo et al. Cell Stem Cell 2009;4:381–384), have been reported to address the risk associated with viral integration. We describe an alternative strategy of reprogramming somatic progenitors entirely through the recruitment of endogenous genes without the introduction of genetic materials or exogenous factors. To this end, we reprogrammed accessible and renewable progenitors from the limbal epithelium of adult rat eye by microenvironment-based induction of endogenous iPS cell genes. Non cell-autonomous reprogramming generates cells that are pluripotent and capable of differentiating into functional neurons, cardiomyocytes, and hepatocytes, which may facilitate autologous cell therapy to treat degenerative diseases. STEM CELLS 2009;27:3053–3060


INTRODUCTION

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

The direct reprogramming of somatic cells to a pluripotent state holds significant implications in treating intractable degenerative diseases by ex vivo cell therapy. In addition, the reprogrammed cells can serve as a model for diseases and discovery of drugs and genes. Reprogramming of somatic cells has been induced by somatic cell nuclear transfer (SCNT), [4, 5] by fusion between embryonic stem (ES) cells, [6, 7] and more recently by ectopic expression of a set of transcription factors (TFs) in somatic cells [1, 8, 9]. The generation of induced pluripotent stem (iPS) cells by exogenous TFs is simple and addresses serious technical and ethical barrier associated with SCNT and cell fusion. This strategy therefore represents a significant breakthrough towards the practical use of stem cells in regenerative medicine.

The generation of first iPS cells from mouse fibroblasts involved retrovirus-mediated expression of 24 ES-cell specific genes that were narrowed down to four, encoding OCT4, KLF4, SOX2, and c-MYC (OKSM) [1]. Since then, iPS cells have been derived from somatic cells of all three germ layers using the same combination of OKSM or that of OCT4, SOX2, LIN28, and NANOG [10]. The reprogramming predominantly involved the expression of exogenous TFs using virus-mediated transduction. The potentially harmful genomic insertion of the virus has been the major barrier to clinical use of iPS cells with this approach. The generation of the mouse iPS cells using plasmid-mediated transient expression of TFs [11, 12] or by the delivery of recombinant TFs [3] has demonstrated that this barrier could be surmounted by alternate TF delivery systems alone or in combination with small molecules-based epigenetic changes (see below).

A more practical approach from the viewpoint of therapeutic use of iPS cells is to recruit the endogenous OKSM genes for reprogramming, thus obviating the need of introducing extraneous genes and factors. The feasibility of this approach is suggested by a remarkable increase in reprogramming efficiency brought about by small molecules that affect chromatin remodeling through acetylation or demethylation [13–15]. Here, we demonstrate a proof of principle for a facile reprogramming of adult limbal progenitors without exogenous TFs by altering the microenvironment. The reprogrammed limbal progenitors are pluripotent (=limbal iPSCs) as shown by three independent lines of evidence of pluripotency that includes embryoid body formation, teratoma formation, and generation of functional cells from three different embryonic lineages in vitro. The non cell-autonomous reprogramming of limbal progenitors to a pluripotent state and their directed differentiation into functional neurons, cardiomyocytes, and hepatocytes provides a novel strategy for autologous cell therapy for degenerative diseases using an accessible adult somatic progenitors.

MATERIALS AND METHODS

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

Mouse ES Cell Culture and Conditioned Medium Collection

Mouse D3 ES cells (ATCC) were cultured in gelatin-coated flasks in the presence of 2000 units/ml of leukemia inhibitory factor (LIF) (Invitrogen, Carlsbad, CA, http://www. invitrogen.com) [16]. Conditioned medium was collected on the third day when cells were 60% confluent. The medium was centrifuged at 1,800 rpm for 5 minutes, filtered through a 45-μm cell strainer (BD Bioscience, San Diego, http://www.bdbiosciences. com), and stored at −80 °C until use.

Isolation and Induction of Limbal Progenitors

Dissection and enrichment of limbal epithelium progenitors was done as previously described [17]. Briefly, adult Sprague-Dawley rats (Sasco) eyes were enucleated in Hank's balanced salt solution. The entire cornea was removed by circular incision below the limbus. The limbal region was dissected away from the peripheral cornea and cut into small pieces. Tissues were incubated first in 0.05% trypsin (Sigma, St. Louis, http://www.sigmaaldrich.com) and then in 78 U/ml of collagenase (Sigma) and 38 U/ml of hyaluronidase (Sigma) at 37°C for 27 and 30 minutes, respectively. Tissues were dissociated into single cells by trituration. Dissociated cells were cultured in DMEM: F12 supplemented with one x N2 supplement (Gibco, Grand Island, NY, http://www.invitrogen.com), 20 ng/ml of EGF (R&D Systems, Minneapolis, http://www. rndsystems.com), 10 ng/ml of FGF2 (R&D Systems, Minneapolis, http://www.rndsystems.com), and 100 ng/ml of Noggin (R&D Systems, Minneapolis, http://www.rndsystems.com) at a density of 105 cells/cm2. After 7 days, resulting neurospheres were cultured in mouse ES cell conditioned medium mixed with equal volume of RCM with 1% FBS (1:1) for 10-14 days until colony formation in the feeder-free conditions. The medium was changed every other day. For the collection of embryonic stem cell conditioned medium (ESCM), D3 ES cells were plated at the density of 1 × 106 cells on gelatin-coated T75 flask. Upon 60% confluency, the medium was collected, filtered through 40 μ cell strainers (BD Biosciences), and stored at −80 °C until use. Colonies were collected for polymerase chain reaction (PCR) and immunocytochemical analyses. Alkaline phosphatase staining was done using ALP kit (ATCC) as per manufacturer's instructions. In some instances, LE cells were infected with Nanog-green fluorescent protein (GFP) lentivirus (System Biosciences, Mountain View, CA, http://www.systembio.com) as per vendor instructions to estimate the efficiency of reprogramming, based on number of Nanog-GFP+ colonies formed per cells seeded, with the efficiency of MOI for lentivirus at 55%.

PCR Analysis

PCR analysis was done as previously described [17]. Total RNA was extracted from cells using the MiniRNeasy Kit (Qiagen, Hilden, Germany, http://www1.qiagen.com) according to the manufacturer's instructions. Complementary DNA synthesis was carried out on 5 μg of total RNA/sample using the SuperscriptIII RT kit (Invitrogen) following the manufacturer's instructions. Transcripts were amplified and their levels quantified using gene-specific primers (supporting information Table 1) and Quantifast SYBR Green PCR kit (Qiagen) on RotorGene 6,000 (Corbett Robotics, San Francisco, CA, http://www.corbettlifescience. com). Measurements were performed in triplicate; a reverse-transcription-negative blank of each sample and a no-template blank served as negative controls. Amplification curves and gene expression were normalized to the housekeeping gene GAPDH, used as an internal standard.

Microarray Analysis

Total RNA was isolated from un-induced and induced LE progenitors and used to synthesize biotin-labeled cRNA probe using Gene Chip 3′ IVT Express kit (Affymetrix, Santa Clara, CA, http://www.affymetrix.com). Fragmented cRNA probes were hybridized to Rat genome 430 2.0 Gene chip arrays (Affymetrix) at 45°C for 16 hours. The arrays were scanned using an Affymetrix GCS3000 7G device, and images were analyzed using the GCOS software. Normalization and expression values were calculated using log scale robust multiarray analysis, implemented in BioConductor [18].

Bisulfite Genomic Sequencing

Bisulfite genomic sequencing was carried out on 0.36 μg of DNA from rat limbal iPSCs and limbal progenitors using EZ DNA Methylation-Direct kit (Zymo Research, Orange, CA, http://www.zymoresearch.com) according to manufacturer's instructions. Bisulfite modified DNA was amplified using gene-specific primers (supporting information Table 2) and cloned into TOPO vector (Invitrogen), and ten randomly selected clones were sequenced.

Karyotype Analysis

G-banding procedure was performed on induced LE cells chromosomes at Human Genetics Laboratories, Munroe-Myer Institute, University of Nebraska Medical Center.

Generation of EBs

Embryoid bodies (EBs) were generated from rat limbal iPSCs and mouse ES cells by hanging drop culture method [19]. Briefly, rat limbal iPSCs and mouse ES cells were suspended in IMDM containing 20% FBS and cultured in 50 μl droplets containing 100 cells on a lid covering sterile 100 mm Petri dish with PBS for 3 days at 37°C.

Generation of Teratomas

For teratoma induction, 2 × 106 rat limbal iPSCs and mouse ES cells were injected subcutaneously into the dorsal flank of severe combined immunodeficiency (SCID) mice. Teratomas were recovered 3–4 weeks post injection, fixed overnight in 10% formalin, paraffin-embedded, and stained with modified periodic acid Schiff with Alcian blue and Masson's trichrome stain. Segments from teratomas were frozen for reverse transcription-PCR (RT-PCR) analysis.

Neuronal Differentiation

Rat limbal iPSCs were differentiated into neurons by a previously described method [20]. Briefly, EBs generated by rat limbal iPSCs were cultured in neural induction medium for 5–7 days at 37°C (induction phase). Resulting colonies were trypsinized and cultured with neural expansion medium [neural induction medium + 20 ng/ml of FGF2 (R&D Systems)] for 7 days (expansion phase). The expanded cells were cultured in neural differentiation medium [DMEM/F12, BDNF of 100 ng/ml of BDNF (R&D systems), 40 ng/ml of GDNF (R&D Systems), and 3 nM Shh (R&D Systems) containing equal volume of conditioned medium from postnatal day 1 rat retinal cells] for 15 days (differentiation phase).

Cardiomyocyte Differentiation

Rat limbal iPSCs were differentiated into cardiomyocytes by a previously described method [19]. Briefly, rat limbal iPSCs were plated on gelatin-coated dish and cultured in IMDM + 20% FBS for 48 hours (induction phase) before shifting to a suspension culture for another 48 hours at 37°C (expansion phase). The medium was changed to IMDM + 0.2% FBS and culture was continued for 10-15 days (differentiation phase). The beating clusters formed at the end of 4 days in differentiation phase.

Hepatocyte Differentiation

Rat limbal iPSCs were differentiated into hepatocytes by a previously described method [21]. Briefly, EBs generated by rat limbal iPSCs were cultured in matrigel (BD Bioscience)-coated dish in differentiation medium I [DMEM/F12, 1% FBS, 1% nonessential amino acids, 1% nucleosides, 1% penicillin + streptomycin, 1% glutamic acid, 3% BSA, 100 ng/ml of FGF2, and 100 ng/ml of Activin A (R&D Systems)] for 3 days at 37°C (induction phase). The medium was changed to differentiation medium II [DMEM/ F12, 15% FBS, 1% nonessential amino acids, 1% nucleosides, 1% penicillin + streptomycin, 1% glutamic acid, 10 ng/ml of HGF (R&D Systems)] and culture was continued for 8 days at 37°C (expansion phase). The medium was changed to differentiation medium III [differentiation medium II + 10−7 M of dexamethasone (Sigma)] and cells were cultured for 10 days (differentiation phase).

Albumin Secretion

The functionality of the differentiated cells was analyzed by the amount of albumin secreted at the end of 10 days of differentiation. The culture medium samples were collected and stored at −20 °C for analysis of albumin. Albumin estimation was done according to the manufacturer's protocol using mouse albumin ELISA kit (Immunology Consultants Laboratory, Inc. Newberg, OR, http://www.icllab.com)

Factor VIIa Assay

Factor VII(a) activity was measured using a modification of the coupled chromogenic assay of factor X activation [22, 23]. Rat primary, rat limbal iPSC, and mouse ES cell derived hepatocytes were treated with vitamin K in the last two days of culture, and media was collected for factor VII activity. Controls included the undifferentiated cells from the rat iPSC and mouse ES cells cultured in the presence and absence of vitamin K. Spent medium was used as the source of factor VII (diluted rat plasma was used as a positive control for factor VIIa activity). Acetone powder of human brain (1/100) was used as the source of tissue factor [24]. Human recombinant factor VIIa was from Novo-Nordisk, and factor X was purified from human plasma [25]. Hirudin (0.1U in 10 ul; Sigma) was added to 20 ul of each sample. Samples with hirudin were combined with 20 ul of factor X (approximately 100 nM in the reaction), 10 ul of spectrozyme factor Xa (American Diagnostica, Stamford, CT, http://www.americandiagnostica.com; final 310 uM), 20 ul of the tissue factor reagent, and 50 ul of 25 mM CaCl2. Absorbance at 405 nm was monitored in a Molecular Devices TMax plate reader. The equation y = VII*S/(k1/2+S) was fit to the data using the direct-plot method [26].

Electrophysiological Analysis

For electrophysiological studies, cells were plated on coverslips, placed in a chamber, and perfused on the stage of an upright, fixed-stage microscope (Olympus BHWI) with oxygenated Ames' medium. Experiments were performed at room temperature. Patch pipettes were pulled on a Narishige PB-7 vertical puller from borosilicate glass pipettes (1.2 mm o.d., 0.95 mm I.D.) and had tips of 1-2 μm O.D. with tip resistances of 6-12 MΩ. Pipettes were filled with a solution containing (in mM): KCH3SO4, 98; KCl, 44; NaCl, 3; HEPES, 5; EGTA, 3; MgCl2, 3; CaCl2, 1; glucose, 2; Mg-ATP, 1; GTP, 1 (pH 7.2). Electrophysiological recordings were obtained using an Axopatch 200B or Multiclamp amplifer (Axon Instruments, Molecular Devices, Sunnyvale, CA, http://www.moleculardevices.com), and responses were acquired using a Digidata 1,322 interface and PClamp 9.2 software (Axon Instruments). Cells were voltage clamped at a steady membrane potential of −70 mV. Capacitative and leak currents were subtracted using a P/8 protocol.

RESULTS

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

The cornea—a transparent and refractive epithelium covering the front of the eye—is continuously regenerated. The progenitors that sustain corneal regeneration and repair are harbored in limbal epithelium, a circular epithelium between cornea and conjunctiva [27–29]. These cells, when removed from their niche and cultured in the presence of Noggin, generate neural progenitors, which differentiate into functional neurons under the influence of extracellular cues [16, 17]. That cells in neurospheres were limbal progenitor derivatives and not of neural crest origin was demonstrated as follows (supporting information Fig. 1): (1) Cells in neurospheres expressed immunoreactivities corresponding to α isoform of ΔP63, which identifies a subset of holoclone-generating progenitors cells in the basal layer of the limbal epithelium [30]. (2) RT-PCR analysis of cells in neurospheres showed the absence of transcripts corresponding to neural crest cell markers, SLUG, BARX1, and DGRC6 [31, 32] and presence of those corresponding to limbal progenitor markers (ΔP63α, α-ENOLASE, K15, and Mesothelin) [33, 34].

The malleability of cells in neurospheres in culture and the observation that these cells express transcripts corresponding to three out of four iPS cell genes, that is, Klf4, Sox2, and c-Myc (KSM) (Fig. 1D, lane 2), suggested that their plasticity could be extended beyond ectodermal lineages by further manipulation of the extracellular environment. We tested the hypothesis by culturing cells from rat LE-derived neurospheres in the presence of mouse embryonic stem (ES) cell conditioned medium (CM) (ESCM) for 10 days (Fig. 1A). We used rat limbal cells because, as an experimental model of diseases, rats are more relevant than mice for their wealth of physiological and pharmacological information related to a wide range of disorders [35]. We used ES cells as inducers because they are known to influence the pluripotent state non cell-autonomously, because when cultured in low density they adopt a neural phenotype [36]. We observed that cells in neurospheres cultured in ESCM generated colonies by 5DIV, whereasthose cultured in the presence of fibroblast CM did not (Fig. 1B). That these colonies were derived from rat cells and did not represent mouse ES cells that might be present in CM as contaminants was demonstrated by karyotype analysis of induced colonies that revealed normal chromosome complement of a rat (n = 21) and not a mouse (n = 20). (Fig. 1C). In addition, induction of limbal epithelial cells (green) from GFP-mice by CM from D3 ES cells (non-green) generated colonies consisting exclusively of green cells., and no colonies consisting of non-green cells were detected (supporting information Fig. 2). The resulting colonies can be passaged every 3–4 days when plated at a density of 105 cells per 24-well plate. Analysis of the colonies at 10 DIV identified Oct4 and Nanog transcripts (Fig. 1D, lane three) that were not present in LE progenitors in un-induced conditions (Fig. 1D, lane 2). Examination of the colonies revealed cells possessing alkaline phosphatase (AP) activities (Fig. 1E) and expressing stage-specific embryonic antigen-1 (SSEA-1), (Fig. 1G) two definitive markers of mouse ES cells [Fig. 1F (AP), 1H (SSEA1)] [37]. Cells in rat limbal iPSCs colonies expressed immunoreactivities corresponding to OCT4 (Fig. 1I) and NANOG (Fig. 1K) like those derived from mouse ES cells [Fig. 1J (OCT4), 1L (NANOG)].

thumbnail image

Figure 1. Non cell-autonomous induction of iPSC phenotype in rat limbal progenitors. A schematic representation of non cell-autonomous induction protocol (A). Noggin- and FGF2-treated rat limbal progenitors generated colonies (B2) in the presence of CM derived from mouse ES cells (ESCM) and not (B3) fibroblasts (MEFCM) (B). An ES cell-derived colony is represented in B1. The induced colonies, when subjected to a karyotype analysis, revealed a normal rat chromosome complement demonstrating the origin from rat cells (C). Examination of the colonies revealed the expression of Oct4 and Nanog (lane 3), previously undetected in Noggin- and FGF2-treated limbal progenitors that expressed Klf4, Sox2, and c-Myc (lane two) (D). Lane one represents minus-reverse transcriptase controls. Lane four represents transcripts corresponding to OKSM genes and Nanog in mouse ES cells. Cells in rat limbal iPSC colonies expressed alkaline phosphatase activities (Es) like those in mouse ES cell colonies (Fs). Immunocytochemical analyses revealed that cells in rat limbal iPSC colonies expressed immunoreactivities corresponding to SSEA1 (Gs), OCT4 (Is), and NANOG (Ks) like cells in ES cell colonies [SSEA1 (Hs), OCT4 (Js), and NANOG (Ks)]. Insets in I4, J4, K4, and L4 represent higher magnification of cells in colonies for better resolution and localization of immunoreactivities. Abbreviations: ES, embryonic stem; ESCM, embryonic stem cell conditioned medium; iPSCs, induced pluripotent stem cells.

Download figure to PowerPoint

Next, we carried out a temporal analysis of OKSM gene expression to ascertain the inductive influence of ESCM (Fig. 2A). A remarkable increase in the levels of Oct4 transcripts was observed between 0 and 6DIV, with the surge being steepest during the first two days in culture. The expression of Nanog was detected at 2DIV and increased thereafter. Transcripts corresponding to Klf4, Sox2, and c-Myc were detected at 0DIV, and their levels increased with time, particularly in the first six days in culture (data not shown). The fact that the expression of Oct4 preceded the induction of Nanog suggests the recapitulation of the core transcriptional network of pluripotency, where Oct4 and Sox2 are upstream of Nanog [38]. That the reprogramming included reciprocal inhibition of limbal-specific genes was demonstrated by a decrease in the levels of transcripts corresponding to p63 and α-enolase, two putative markers of LE progenitors [39]. Bisulfite sequencing of Oct 4 and Nanog promoters demonstrated a general decrease in the number of methylated CpGs dinucleotides (more pronounced in SP1 (Oct4)- and IK (Nanog)-binding sites), in rat limbal iPSCs compared to controls, suggesting that Oct4 and Nanog promoters are more active in the former relative to the latter (Fig. 2B). A comparative global gene expression analysis revealed that the expression of 14 out of 18 (77.7%) key genes underlying the regulatory network of pluriopotency [40, 41] was upregulated rat limbal iPSCs, relative to limbal progenitor controls (Fig. 2C). In the absence of microarray chips that allowed a comparison of mouse and rat global gene expression, we compared the expression of OKSM, Nanog, and Lin28 between rat limbal iPSCs and mouse ES cells by Q-PCR (Fig. 2D). The levels of transcripts corresponding to Oct4, Sox2, Klf4, cMyc, Nanog, and Lin28 in rat limbal iPSCs were significantly lower than those in mouse ES cells, suggesting that the iPSCs were similar but not identical to mouse ES cells.

thumbnail image

Figure 2. Non cell-autonomous reprogramming of rat limbal progenitors. Rat limbal progenitors were examined for the expression of Oct4 and Nanog at different time points during ESCM-mediated induction (A). Transcripts corresponding to Oct4 were detected at 4DIV and their levels increased with the time. Transcripts corresponding to Nanog were detected at 6DIV, and their levels increased temporally but remained significantly lower than that of Oct4. In contrast, transcripts corresponding to limbal progenitor markers, p63 and α-enolase, decreased with the time and were undetectable at 14DIV under ESCM-mediated induction. Bisulfite genomic sequencing, carried out on genomic DNA derived from FACS-sorted Nanog-GFP+ limbal iPSCs at 14 DIV, revealed a decrease in the number of methylated CpG dinucleotides in the Oct4 and Nanog promoters, compared to those in the uninduced limbal progenitors (B). The asterisks in Oct4 and Nanog promoters identify CpG dinucleotides at SP1 and IR1/IR3 binding sites, respectively, which are significantly demethylated in rat limbal iPSCs. A comparative global gene expression analysis of rat limbal iPSCs and rat limbal progenitors at 14 DIV revealed upregulation of key pluripotency network genes in the former relative to the latter (C). Red lines indicate the linear equivalent and two-fold changes in gene expression levels between the samples. Red dots indicate the 18 regulatory genes. A comparative analysis of Oct4, Klf4, Sox2, cMyc, Nanog, and Lin28 expression by quantitative polymerase chain reaction revealed that they are expressed at significantly lower levels in rat limbal iPSCs, compared to mouse ES cells, suggesting that reprogramming of limbal progenitors is similar but not identical to mouse ES cells (D). The reprogramming efficiency was estimated by counting the GFP+ colonies generated from Nanog-GFP lentivirus-transduced rat limbal progenitors plated and cultured under the influence of ESCM (E, F, and G). The reprogramming efficiency significantly increased in the presence of VPA (G). Abbreviations: ES, embryonic stem; ESCM, embryonic stem cell conditioned medium; iPSCs, induced pluripotent stem cells; VPA, valproic acid.

Download figure to PowerPoint

Furthermore, to examine the specificity of Nanog induction by ESCM, limbal progenitors were transduced with Nanog-GFP lentivirus before induction (Fig. 2E, 2F). GFP+ cells were observed only in culture containing ESCM and none in the sister wells without ESCM. The activation of exogenous Nanog promoter-driven GFP expression was used as a measure to determine the efficiency of reprogramming (Fig. 2G). The programming efficiency was 0.0025%, better than three (0.001%) and two (0.001%) factor retroviral transduction of human fibroblasts but lower than the efficiency of reprogramming obtained by four (0.1%) and three (0.01%) factor retroviral transduction of mouse embryonic fibroblasts [42] and references therein. The efficiency was also lower (0.014%) than that obtained by single-factor retroviral transduction of adult neural stem cells [43]. Next, we tested whether or not small molecules could augment reprogramming of this easily accessible adult stem cell types. Evidence has emerged recently that epigenetic changes brought about by small molecules such as valproic acid (VPA) significantly influence reprogramming efficiency [13]. When we cultured limbal progenitors in ESCM in the presence of VPA, the reprogramming efficiency increased ∼4 fold, compared to controls (0.0078 ± 0.0002% vs. 0.0025 ± 0.0007%; p < .005) (Fig. 2G). These observations suggested that limbal progenitors can be reprogrammed non cell-autonomously to express OKSM and other genes underlying pluripotency under the influence of activities elaborated by ES cells and that the efficiency of reprogramming can be amplified by small molecule-mediated epigenetic changes.

Next, we examined whether or not rat limbal iPSCs possess the potential to differentiate along three different embryonic lineages in vitro and in vivo. First, we subjected rat limbal iPSCs and mouse ES cells to hanging drop culture method to generate embryoid bodies (EBs). We observed that the induced rat LE cells generated EBs in the same time frame (2DIV) as mouse ES cells when LIF was removed and substituted by serum (Fig. 3A, 3E). The resulting EBs were indistinguishable from ES cell-generated EBs; both expressed immunoreactivities (Fig. 3B–3D, 3F–3H) and transcripts (Fig. 3I) corresponding to markers of embryonic ectoderm (e.g., OTX2), endoderm (e.g., SOX17), and Mesoderm (e.g., Brachyury). Second, we compared the ability of induced rat limbal iPSCs and mouse ES cells to form teratomas when injected subcutaneously in NOD SCID mice. Mice in both groups developed teratomas within three weeks of injection (Fig. 4A–4D). Except for the size, which was smaller in the case of induced rat limbal iPSCs, morphologically they were similar. Teratomas were not detected when uninduced cells were injected. Histological examination of teratomas revealed the presence of squamous (ectoderm), glandular (endoderm), and cartilagenous (mesoderm) tissues (Fig. 4E–4J). Transcripts corresponding to tissues of different embryonic lineages were identified in teratomas; Otx2, Musashi1, and Map2 (ectoderm); Brachyury, PECAM, and VE cadherin (mesoderm); FoxA2, GATA1, and AFP (endoderm) (Fig. 4K). Third, using established methods for directed differentiation of mouse ES cells, we examined whether or not rat limbal iPSc are amenable to differentiation into functional neurons [44], cardiomyocytes [19], and hepatocytes [21] as described below (Fig. 5).

thumbnail image

Figure 3. Pluripotent potential of rat limbal iPSCs (generation of embryoid bodies): Limbal iPSCs when cultured in the absence of LIF formed EBs (A–D) like mouse ES cells (E–H). Cells in EBs derived from rat limbal iPSCs expressed immunoreactivities (A–H) and transcripts (I) corresponding markers of embryonic ectoderm, mesoderm, and endoderm like those derived from mouse ES cells. Lane M = marker; lane 1 = mouse ES cells; lane 2 = rat limbal iPSCs. Abbreviations: EBs, embryoid bodies; ES, embryonic stem; iPSCs, induced pluripotent stem cells.

Download figure to PowerPoint

thumbnail image

Figure 4. Pluripotent potential of rat limbal iPSCs (generation of teratomas): Rat limbal iPSCs when injected in SCID mice formed teratomas (A, B) like mouse ES cells (C, D). Teratomas generated by rat limbal iPSCs, like those formed by mouse ES cells, contained tissues of all three embryonic lineages; ectoderm [keratinized squamous epithelium (arrow)] (E, H), mesoderm [cartilageneous tissues with chondrocytes (arrow)] (F, I), and endoderm [glandular columnar epithelium (arrow)] (G, J). Examination of teratomas by reverse transcription--polymerase chain reaction analysis revealed the presence of transcripts corresponding to markers of embryonic ectoderm, endoderm, and mesoderm (K). Lane M = marker; lane 1 = mouse ES cells; lane 2 = rat limbal iPSCs. Abbreviations: ES, embryonic stem; iPSCs, induced pluripotent stem cells.

Download figure to PowerPoint

thumbnail image

Figure 5. Pluripotent potential of rat limbal iPSCs (directed differentiation into cells of three embryonic lineages): Rat limbal iPSCs were subjected to a neuronal differentiation protocol established for mouse ES cells (A). Temporal Q-polymerase chain reaction (Q-PCR) analysis of gene expression revealed that limbal iPSCs expressed Sox2 and Mash1 during the induction phase. As cells entered the differentiation phase, expression of these transcripts decreased and those corresponding to neuronal marker Nfl increased (B). Limbal iPSCs at the end of differentiation phase expressed immunoreactivities corresponding to MAP2 (C) and NFL (D). Whole-cell recordings of these cells revealed fast-acting inward currents due to voltage-gated sodium channel (E), blocked by TTX (F), and AMPA receptor mediated inward currents (G), characteristics of neurons. Rat limbal iPSCs were subjected to a cardiomyocyte differentiation protocol established for mouse ES cells (H). Temporal Q-PCR analysis of gene expression revealed that limbal iPSCs expressed Brachyury and GATA4 during the induction phase. As cells entered the differentiation phase, expression of Brachyury decreased and those corresponding to GATA4 and ANP increased (I). Limbal iPSCs at the end of cardiomyocyte differentiation phase expressed immunoreactivities corresponding to α-actinin (J) and cTnT (K). Whole-cell recordings of beating cardiomyocytes revealed the presence of L-type calcium currents (L) blocked by nifedipine (M) and action potentials characteristic (N) of ventricular cardiomyocytes. Rat limbal iPSCs were subjected to a hypotocyte differentiation protocol established for mouse ES cells (O). Temporal Q-PCR analysis of gene expression revealed that limbal iPSCs expressed FoxA2 and Sox17 during the induction phase (P). As cells entered the differentiation phase, expression of these genes attenuated and that of albumin increased. Limbal iPSCs at the end of hepatocyte differentiation phase expressed immunoreactivities corresponding to Cyp7a1 (Q) and albumin (R). The differentiated rat limbal iPSCs, like differentiated mouse ES cells, elaborated albumin in the culture medium, albeit at lower levels (S). A comparison of K1/2, app from the curves fit to the data obtained from a chromogenic assay of factor VII activity revealed that rat limbal iPSC- and mouse ES cell-derived hepatocytes secreted 2- to 3-fold more factor VII than primary rat hepatocytes, compared to undifferentiated control cells (T). The closed symbols resemble the saturation level of the assay. Abbreviations: EBs, embryoid bodies.

Download figure to PowerPoint

Rat limbal iPSCs subjected to neural induction (Fig. 5A) expressed transcripts corresponding to neural progenitor markers (e.g., SOX2, NESTIN) and regulators of neurogenesis (e.g., Mash1 and NeuroD) during induction and/or expansion phase (Fig. 5B and supporting information Fig. 3). As cells entered the differentiation phase, expression of these transcripts decreased and those corresponding to neuronal markers (e.g., βtubulin and NFL) increased. At the end of the differentiation phase, limbal iPSCs displayed neuronal morphology and expressed neuronal markers, NFL and MAP2 (Fig. 5C, 5D). Whole-cell patch recording of differentiated limbal iPSCs revealed the presence of sustained outward currents due to outwardly rectifying K+ channels and fast Na+ currents (10/22 cells) (Fig. 5E) blocked by bath application of TTX (1μMB, n = 8 cells) (Fig. 5F). Bath application of the ionotropic glutamate receptor agonist, AMPA (20 μM) to cells exhibiting fast Na+ currents, stimulated inward currents (3/3 cells) (Fig. 5G). In contrast, AMPA failed to evoke inward currents in cells that lacked fast Na+ currents (7/7 cells). These results suggested that rat limbal iPSCs possess the potential to differentiate along neuronal lineage and under neuronal differentiation conditions express voltage- and ligand-gated currents characteristic of functional neurons.

Rat limbal iPSCs that were subjected to directed cardiomyocyte differentiation (Fig. 5H) expressed transcripts corresponding to early mesodermal markers, Brachyury and GATA4, during the induction and expansion phases (Figure 5I and supporting information Fig. 4). The differentiation phase was defined by a temporal increase in transcripts corresponding to mature markers of cardiomyocytes, alpha smooth muscle actin (αSMA), atrial nitriuretic peptide (ANP), alpha myosin heavy chain (αMHC), and beta myosin heavy chain (βMHC). Cells in the spontaneously beating areas (supporting information Fig. 6) expressed cardiac troponin T (cTnT) and α-actinin (Fig. 5J and 5K). Whole-cell recordings from cells in the beating areas revealed that depolarizing steps above −50 mV evoked sustained inward L-type Ca2+ currents (24/65 cells) (Fig. 5L) that could be blocked by bath application of nifedipine (30 μM, n = 10 cells) (Fig. 5M). These cells also displayed lengthy action potentials in response to depolarizing current injection (Fig. 5N). Thus, rat limbal iPSCs under cardiomyocyte differentiation conditions expressed contractile behavior, L-type Ca2+ currents, and depolarizing action potentials characteristic of ventricular myocytes.

Rat limbal iPSCs subjected to directed hepatocytes differentiation (Fig. 5O) expressed transcripts corresponding to early endodermal markers, GATA1, FoxA2, and Sox17, during induction and/or expansion phases (Fig. 5P and supporting information Fig. 4). As cells entered the differentiation phase, expressions of these transcripts decreased and those corresponding to mature hepatocytes markers, HPRT, TTR, ALBUMIN, aldolase B, and villin, increased. Rat limbal iPSCs, at the end of the differentiation phase, expressed markers corresponding to Cyp7a1 and ALBUMIN (Fig. 5Q, 5R). Next, we examined whether or not differentiated limbal iPSCs that display albumin immunoreactivities secreted albumin. Analysis of supernatants from rat limbal iPSCs and mouse ES cell cultures by ELISA revealed the presence of albumin, albeit at different levels (25.5 ± 0.63 vs. 100.8 ± 0.48 ng/ml), suggesting functional differentiation of rat limbal iPSCs along hepatocyte lineage (Fig. 5S). To further examine the functional maturity of rat limbal iPSC-derived hepatocytes, we tested whether or not they secreted functional coagulation factor VII like primary hepatocytes. A chromogenic assay of factor VII activity revealed that the rat iPSC-derived hepatocytes secreted readily measureable amounts of factor VII as primary hepatocytes and similarly derived hepatocytes from mouse ES cells (Fig. 5T). A comparison of K1/2app from the curves fit to the data showed that media from in vitro differentiated rat limbal iPSCs and mouse ES cells secreted 2- to 3-fold (1.9-3.1) more factor VII than primary hepatocytes. We observed that in all instances of cell type-specific differentiation protocols, expression of Oct4 was undetectable or decreased temporally, attesting to reprogramming during differentiation that requires silencing of Oct4 [45] (supporting information Figs. 2-4). Together, these results suggest that rat limbal iPSCs are equally pluripotent as mouse ES cells and capable of generating functional cells of three different lineages in vitro.

DISCUSSION

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

These observations demonstrate that somatic progenitors can be reprogrammed to pluripotency without forced expression or introduction of exogenous transcription factors by a simple manipulation of the microenvironment. The limbal progenitors, which can be obtained from 2 mm limbal biopsy of a patient without affecting vision, may be reprogrammed in a relatively short time, non cell-autonomously, for autologous cell therapy. The examination of cell specific-markers suggests that the rat iPSCs arise from limbal progenitors and not from stromal cells of neural crest origin. Although a contamination with rare immigrant bone marrow mesenchymal stem cells expressing ES cell markers [46] cannot be entirely excluded, the fact that limbal cells, besides expressing specific markers, are induced to express Oct4 and Nanog suggests against that possibility. The reprogramming involved two stages: initial induction of the expression of KSM under the influence of defined exogenous factors (FGF2 and Noggin) and the activation of Oct4 and Nanog under the influence of ESCM. The unidentified activities elaborated by ES cells effectively influenced the maintenance of the expression of all four iPS cell genes, OKSM. Whether or not these activities can similarly reprogram differentiated somatic cells such as fibroblasts or neural cells is currently being investigated. However, it is likely that progenitors will be more susceptible to non-cell autonomous reprogramming than somatic cells if one considers recent findings that hematopoietic stem cells are reprogrammed by OKSM with a significantly higher efficiency than B- and T-cells [47], KSM expressing neural stem cells can be reprogrammed by exogenous Oct4 alone [43], and that the Oct4-expressing adult germline stem cells can be reprogrammed under culture conditions similar to ours [48]. It is quite likely that the ES cell-derived activities were effective in reprogramming because the limbal progenitors genome was rendered malleable to these activities due to prior expression of KSM. The induction of Oct4 might represent the tipping point where OCT4 and SOX2 initiated a concerted activation of Nanog, and NANOG, in turn, stabilized the core transcriptional network for a pluripotent state [38]. Such sequential induction of genes and stabilization of their expression may involve cooperative interactions between iPS cell TFs and chromatin remodeling enzymes whose function and/or expression is open to extrinsic influences as suggested by a significant increase in reprogramming efficiency of limbal progenitors brought about by VPA.

CONCLUSION

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

Our observations offer a proof-of-principle for a stepwise non cell-autonomous reprogramming strategy where cell-extrinsic influences aided by small molecule-based epigenetic changes may lead somatic cells to a pluripotent state, driven entirely by endogenous genes, thus circumventing potential problems associated with exogenous transcription factor-based reprogramming. This will address a significant barrier to iPSC-based therapy for degenerative diseases.

Acknowledgements

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

Thanks are due to Drs. Graziella Pellegrini for P63α antibody, Judith Christman and David Klinkebiel for bisulfite sequencing, James Eudy for transcription profiling, Greg Bennett for a neural crest cell line, Hesham Basma for help in hepatocytes differentiation, George Rozanski for initial help with electrophysiology, Gordon Todd for image analysis, and Janice Taylor for confical microscopy.

DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST

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

This research was supported by the Lincy Foundation, Nebraska Department of Human and Health Services, the Pearson Foundation, and Nebraska Tobacco Fund for Biomedical Research.

References

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

Supporting Information

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

Additional supporting information available online.

FilenameFormatSizeDescription
STEM_242_sm_suppinfofigure1.tif21814KFigure 1. Limbal specificity of progenitors: Sections of adult rat eye through limbal epithelium reveal that cells in the basal region express immunoreactivities corresponding to limbal progenitors markers, ΔP63α (A-C). Cells in the neurospheres prior to iPSC induction expressed immunoreactivities corresponding to ΔP63α (D-F). RT-PCR analysis carried out on these cells revealed that they express transcripts corresponding to limbal progenitor markers, ΔP63α, α-ENOLASE, CYTOKERTIN 15, and MESOTHELIN and not neural crest markers, SLUG, BARX1, and DGRC6 (G). Lane M=marker; lane 1= limbal epithelium cells; 2=limbal neurospheres; 3= neural crest cell line.
STEM_242_sm_suppinfofigure2.tif6715KFigure 2: LE progenitors derived from GFP-mice when cultured in the presence CM derived from mouse ES cells (D3 line) generated colonies, which consist exclusively of GFP+ cells
STEM_242_sm_suppinfofigure3.tif6628KFigure 3. Neuronal differentiation of rat limbal iPSCs. Rat limbal iPSCs were subjected to neuronal differentiation protocol established for mouse ES cells (A). Temporal PCR analysis of stage- and cell type-specific gene expression in limbal iPSCs undergoing neuronal differentiation (B).
STEM_242_sm_suppinfofigure4.tif4513KFigure 4. Cardiomyocyte differentiation of rat limbal iPSCs. Rat limbal iPSCs were subjected to cardiomyocyte differentiation protocol established for mouse ES cells (A). Temporal PCR analysis of stage- and cell type-specific gene expression in limbal iPSCs undergoing cardiomyocyte differentiation (B).
STEM_242_sm_suppinfofigure5.tif8569KFigure 5. Hepatocyte differentiation of rat limbal iPSCs. Rat limbal iPSCs were subjected to hepatocyte differentiation protocol established for mouse ES cells (A). Temporal PCR analysis of stage- and cell type-specific gene expression in limbal iPSCs undergoing hepatocyte differentiation (B).
STEM_242_sm_suppinfofigure6a.mp43260KFigure 6. Beating colonies of cardiomyocytes. Rat limbal iPSCs when subjected to cardiomyocytes differentiation protocol generated colonies 4DIV with spontaneous beating properties (A).
STEM_242_sm_suppinfofigure6b.mp43177KCells from beating colonies displayed contractile properties upon culture (B).
STEM_242_sm_suppinfotables.doc90KTable 1: Gene-specific primers for RT and Q-PCR. Table 2: Primers for bisulfite genomic sequencing.

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