SEARCH

SEARCH BY CITATION

Keywords:

  • hESC;
  • iPS cells;
  • MSCs;
  • SMAD-2/3;
  • Inflammatory bowel disease

Abstract

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

Human ESCs provide access to the earliest stages of human development and may serve as an unlimited source of functional cells for future cell therapies. The optimization of methods directing the differentiation of human embryonic stem cells (hESCs) into tissue-specific precursors becomes crucial. We report an efficient enrichment of mesenchymal stem cells (MSCs) from hESCs through specific inhibition of SMAD-2/3 signaling. Human ESC-derived MSCs (hESC-MSCs) emerged as a population of fibroblastoid cells expressing a MSC phenotype: CD73+ CD90+ CD105+ CD44+ CD166+ CD45− CD34− CD14− CD19− human leucocyte antigen-DR (HLA-DR)−. After 28 days of SMAD-2/3 inhibition, hESC cultures were enriched (>42%) in multipotent MSCs. CD73+CD90+ hESC-MSCs were fluorescence activated cell sorting (FACS)-isolated and long-term cultures were established and maintained for many passages displaying a faster growth than somatic tissue-derived MSCs while maintaining MSC morphology and phenotype. They displayed osteogenic, adipogenic, and chondrocytic differentiation potential and exhibited potent immunosuppressive and anti-inflammatory properties in vitro and in vivo, where hESC-MSCs were capable of protecting against an experimental model of inflammatory bowel disease. Interestingly, the efficient enrichment of hESCs into MSCs through inhibition of SMAD-2/3 signaling was not reproducible with distinct induced pluripotent stem cell lines. Our findings provide mechanistic insights into the differentiation of hESCs into immunosuppressive and anti-inflammatory multipotent MSCs with potential future clinical applications. STEM CELLS 2011;29:251–262


INTRODUCTION

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

Human embryonic stem cells (hESCs) offer great promise for cell replacement strategies. Recent advances in somatic cell reprogramming to induced pluripotent stem (iPS) cells have opened new avenues to generate donor-specific cells for regenerative medicine and disease modeling [1]. Human ESCs also provide access to the earliest stages of human development and may serve as an unlimited source of functional cells for future cell therapies [2, 3]. However, reproducible directed differentiation of hESCs and iPS cells into specific cell types poses a formidable challenge. Thus, optimization of methods aimed at directing the differentiation of hESCs into tissue-specific precursors becomes crucial.

Mesenchymal stem cells (MSCs) may be isolated from a variety of adult somatic tissues and they may differentiate into multiple mesodermal tissues [4–7]. Their multilineage differentiation potential (bone, cartilage, fat) coupled to their immunoprivileged properties is being exploited worldwide for both autologous and allogeneic cell replacement strategies [8–12]. Recently, MSCs have been derived from hESC through coculturing with the OP9 murine bone marrow stromal cell line [13, 14]. However, there is no information available about the mechanistic insights involved in the specific differentiation of hESCs into functional MSCs. Furthermore, whether hESCs and iPS cells are equally capable of generating MSCs remain to be assessed. Also, little is known about the immunological properties of hESC-derived MSCs. Their immunotolerance properties in vitro have been partially reported [14] but no information is available about their potential immunotolerance and anti-inflammatory properties in vivo.

The transforming growth factor (TGF-β) signaling through SMAD-2/3 downstream effectors has an increasing interest in regenerative medicine and lineage specification during human embryonic development [15, 16]. In contrast to specific bone morphogenic proteins (BMPs), which are known to play an important role in directing cell fate decisions toward mesoderm and further differentiation into MSCs, the mesodermal effects of TGF-β signaling in human embryonic development is controversial [15, 16]. We previously reported that specific inhibition of TGF-β signals through SMAD-2/3 results in loss of the hESC phenotype and pluripotency through induction of hESC differentiation, while not affecting survival or self-renewal of hESCs [17]. Furthermore, TGF-β is considered a potent inhibitor of the hematopoiesis, a mesodermal tissue [18–20]. In fact, TGF-β1−/− mice display an enhanced myelopoiesis and an increased stem/progenitor cell (HSPC) compartment. These HSPCs from TGF-β1−/− mice have a competitive repopulation advantage in vivo [18–20]. We, therefore, hypothesized that similar to what occurs in the hematopoietic system, TGF-β signaling through SMAD-2/3 might be a negative regulator in the MSC generation from hESCs. We, thus, studied the role of SMAD-2/3 inhibition in the potential enrichment of functional and multipotent MSCs from hESCs and iPS cells. We report a robust and efficient enrichment of MSCs from hESCs through specific inhibition of the SMAD-2/3 pathway. These hESC-derived MSCs display multilineage differentiation potential and exhibited potent immunosuppressive and anti-inflammatory properties in vitro and in vivo, which are capable to protect against experimental inflammatory bowel disease. Finally, our data reveals possible mechanistic differences in MSC generation from hESCs versus iPS cells.

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 CONFLICTS OF INTEREST
  10. References

Human ESC and iPS Cell Culture

Human ESC lines (H9, AND-1, AND-2, and SHEF-1) and iPS cells (iPS-CB-CD34#2, iAND-4, iMSUH001) were maintained undifferentiated in a feeder-free culture as previously described [21–25]. These iPS cell lines were derived from fetal human fibroblasts using a lentiviral construct that express OCT4, SOX2, KLF4, and c-MYC. Briefly, hESCs were cultured in Matrigel (Becton Dickinson, San Jose, CA)-coated T25 flasks in conditioned medium supplemented with 8 ng/ml of basic fibroblast growth factor (bFGF) (Miltenyi, Madrid, Spain). Media was changed daily, and the cells were split weekly by dissociation with 200 U/ml of collagenase IV (Invitrogen, Edinburgh, Scotland). Distinct hESC and iPS cultures were treated daily with 10 ng of TGF-β1 (R&D Systems, London, U.K.), 10 μM of SB-431542 (Sigma-Aldrich, St Louise, MO), a potent and selective chemical inhibitor of ALK4, ALK5, and ALK7 receptors inducing shutdown of TGF-β signaling through SMAD-2/3, or dimethylsulfoxide (DMSO) as vehicle for 28 (hESCs) or 35 (iPS cells) days. Human ESC and iPS cell cultures were visualized daily by phase-contrast microscopy. Approval from the Spanish National Embryo Ethical Committee was obtained to work with pluripotent cells.

Flow Cytometry Characterization of hESC and iPS-Derived MSCs

Flow cytometry analysis was carried out weekly (day 7, 14, 21, 28, and 35) to analyze the potential emergence of CD73+CD90+CD34− MSCs [6, 26]. Briefly, hESC cultures were dissociated with trypsin-EDTA and the single cell suspension was stained at a concentration of 2–5 × 105 cells per milliliter with a fluorescein isothyocianate (FITC)-conjugated anti-CD90 and phyco-erithrin (PE)-conjugated anti-CD73 monoclonal antibodies (Becton Dickinson, San Jose, CA). Then, the cells were washed and stained with 7-actino (7-AAD) for 15 minutes at room temperature. Live cells identified by 7-AAD exclusion were analyzed for coexpression of CD90 and CD73 using a FACSCanto II flow cytometer equipped with the fluoresence activated cell sorting (FACS) Diva analysis software (Becton Dickinson). In parallel, the expression of the ESC-specific markers Tra-1-60-FITC (Chemicon, San Diego, CA) and Oct3/4-PE (Pharmingen, San Jose, CA) and the hematoendothelial marker CD34 (Becton Dickinson) was determined by flow cytometry. Irrelevant IgG isotype-matched antibodies were consistently used [27].

Western Blotting

The potent and specific effect of the SB-431542 chemical inhibitor in blocking signaling through SMAD2/3 but not via SMAD1/5/8 was confirmed as described [28]. Briefly, hESCs and iPS cells were treated for 60 minutes with or without SB-431542 or TGF-β1 and cell lysates (40 μg) were prepared from equivalent cell numbers. Cells were washed and lysed using the nuclear extraction kit with protease and phosphatase inhibitors (Active Motif). Samples were then run on 10% SDS-polyacrylamide gel (PAGE) and transferred to polivinylidene fluoride (PVDF) membranes using a semidry transfer apparatus at 15 V for 1 hour. Membranes were blocked with 5% bovine serum albumin (BSA) in tris buffer saline tween-20 (TBST) and incubated with the following primary antibodies: rabbit anti-phospho-SMAD2/3, rabbit anti-phospho-SMAD1/5/8 or rabbit anti-total SMAD-2/3, and rabbit anti-total SMAD-1/5/8 (all from Cell Signaling Technologies, Danvers, MA 1:10,000). All secondary antibodies were conjugated to horse radish peroxidase (HRP) and consequently detected using the enhanced chemiluminescence system (ECL; Pierce, Rockford, IL). The ECL signal was detected using an imaging and detection station (Alpha-Innotech Corp, San Diego, CA).

Cell Sorting and Establishment of hESC-MSC Cultures

Differentiating hESCs were incubated with CD73-PE and CD90-FITC and the double-positive CD73+CD90+ population was purified with the FACSAria cell sorter (Becton Dickinson) using the automatic cell deposition unit. A total of 2–3 × 105 sorted cells were immediately plated back into fibronectin-coated plates to facilitate adherence. After 24 hours, nonadherent cells were washed away and fresh prewarmed medium was added. Human ESC-derived CD73+CD90+ cells were cultured in Advance-Dulbecco's modified Eagle's medium (DMEM) with 10% fetal calf serum (FCS), 1% Glutamax, and 1% penicillin/streptomycin (all from Gibco, Edinburgh, Scotland) and were allowed to expand and reach nearly 100% confluence.

Phenotypic Characterization of hESC-MSCs

The immunophenotype of cultured hESC-MSCs was determined by flow cytometry as previously described [29]. Briefly, hESC-MSCs were trypsinized, washed, and suspended in PBS + 1% BSA. A total of 2 × 105 cells were incubated for 30 minutes in the darkness with the following FITC- or PE-conjugated monoclonal antibodies: CD73, CD90, CD105, CD44, CD166, CD106, CD45, CD34, CD14, CD19, human leucocyte antigen-DR (HLA-DR), CD40, CD80 (Becton Dickinson), and SSEA-4 (Chemicon, Billerica, MA). Cells were then washed in PBS + 1% BSA and analyzed in a FACSCanto II flow cytometer.

In Vitro Differentiation of hESC-MSCs

Human ESC-MSCs were seeded at 1 × 104 cells per centimeter square in Advance-DMEM with 10% FCS, 1% Glutamax, and 1% Pen/Strep and were allowed to expand and reach nearly 100% confluence. Culture medium was then replaced with specific differentiation inductive medium. For adipogenic differentiation, cells were cultured in Adipogenic MSCs Differentiation Bullet Kit (Lonza, Basel, Switzerland) for 2 weeks. Differentiated cell cultures were stained with Oil Red (Sigma, Madrid, Spain). For osteogenic differentiation, cells were cultured in the Osteogenic MSCs Differentiation Bullet Kit (Lonza) for 2 weeks. Differentiated cultures were stained with Alizarin Red (Sigma) [29].

Assessment of T-Lymphocyte Proliferation Response to MSCs and Allogeneic Stimulation

Mixed lymphocyte cultures (MLCs) were performed in 96-well round-bottom plates by stimulating 105 responder peripheral blood mononuclear cells (PBMCs) from donor A with 105 allogeneic HLA-mismatched mitomycin C-treated stimulator PBMCs from donor B in 200 μl complete medium in the presence or absence of different numbers of adipose-derived MSCs (Ad-MSCs), cord blood-derived MSCs (CB-MSCs), or hESC-MSCs. PBMCs were isolated from buffy coat preparations derived from the whole blood of healthy volunteers by density sedimentation on Ficoll-Hypaque gradients (20 minutes, 600g). Cells recovered from the gradient interface were washed twice in RPMI medium and immediately used for culture. Human Ad-MSCs and CB-MSCs were purchased from InbioBank (Inbiomed, San Sebastian, Spain) [26, 30]. These MSCs were expanded in DMEM containing 10% fetal bovine serum (FBS), 2 mM glutamine and 1% pen/strept at 37°C, and 5% CO2.

Cells were pulsed with 5 microcurie per well [3H]-thymidine for the last 12 hours of the culture, harvested onto membranes, and proliferation was determined by measuring [3H]-thymidine uptake in a liquid scintillation counter. After 48 hours, cytokine determinations for interleukin (IL)-2, IL-4, TGF-β, tumor necrosis factor-α (TNF-α), and γ-interpheron (IFN-γ) in the supernatants were determined by enzime-linked immuno sorbent assay (ELISA) using capture/biotinylated detection antibodies from BD Pharmingen. In similar experiments, responder PBMCs were labeled with 2.5 μM carboxybluorescein diacetate succinimidyl ester (CFSE; Molecular Probes, Carlsbad, CA) prior to setting up cocultures. After culture, cells were labeled with PerCP-labeled anti-CD4 antibody, fixed with 1% paraformaldehyde and proliferating cells were determined by CFSE dilution in the CD4+ population on a FACScalibur cytometer (Becton Dickinson). The number of cycling cells was calculated as the percentage of CFSEmild/low cells that had divided × the total number of cells.

Anti-Inflammatory Studies

Fibroblast-like synoviocytes (FLSs) cultures were established in 10% FBS/DMEM from synovial tissue obtained from two unrelated patients with active rheumatoid arthritis (RA) at time of knee replacement surgery. FLS cultures were conducted in complete medium consisting of RPMI supplemented with heat-inactivated human pooled serum (8%), L-glutamine (20 mM), sodium pyruvate (1%), nonessential amino acids (1%), and pen/strept (1%) in a 5% CO2 humidified atmosphere at 37°C. A total of 2 × 105 FLSs were stimulated with either lipopolysaccharide (LPS; 1 μg/ml, Sigma) or TNF-α (20 ng/ml) in the presence or absence of 105 Ad-MSCs, CB-MSCs, or hESC-MSCs. After 24–48 hours, culture supernatants were assayed for cytokine contents (IL-6 and TNF-α) and collagenase activity. Collagenase activity was determined using the EnzChek gelatinase/collagenase assay kit (Molecular Probes), which is a fibril degradation assay that uses self-quenched fluorescein-conjugated type I collagen, to determine the collagenase activity in cell-free supernatants.

Induction of Experimental In Vivo Colitis

To induce in vivo colitis, 3 mg of 2,4,6-trinitrobenzene sulfonic acid (TNBS; Sigma) in 50% ethanol (100 μl) was administered intrarectally in 7-week-old BALB/c male mice. Control mice received 50% ethanol alone. Animals were treated i.p. with medium or with 106 cells of hESC-MSCs, CB-MSCs, or Ad-MSCs 12 hours after TNBS instillation. Animals were monitored for the appearance of diarrhea, body weight loss, and survival. Colons were removed from the caecum to the anus, and colon length and weight were measured as indirect markers of inflammation. Colons were evaluated for macroscopic damage (graded on a scale 0–10) based on criteria reflecting inflammation (i.e., hyperemia, bowel thickening, and ulceration extent). Scores for stool consistency and rectal bleeding were assessed as described [31]. For histopathology analysis, a colon specimen from the middle part was fixed in 10% buffered formalin phosphate, embedded in paraffin, sectioned, and stained with H&E. Inflammation was graded from 0 to 4 as described [31]. Neutrophil infiltration in the colon was monitored by measuring myeloperoxidase (MPO) activity as described [31]. The animal care committee of the University o f Granada-CSIC approved all mice protocols.

Analysis of Biodistribution of hESC-MSCs

To trace the injected cells in vivo, hESC-MSCs were labeled with CFSE in PBS/0.1% BSA for 10 minutes at 37°C, and after extensive washing in DMEM medium/10% FBS, the cells were injected i.p. in naive and TNBS-treated animals. Two days after injection, single cell suspensions were isolated from spleen, mesenteric lymph nodes, various colon segments (inflamed and noninflamed), stomach, intestine, liver, lung, salivary glands, skeletal muscle, brain, and kidney following digestion with DNase I (0.1 mg/ml, Sigma) and collagenase IV (400 U/ml, Sigma) for 20 minutes at 37°C in continuous shaking. After 40-μm filtration, cells were stained with PE-anti-human HLA-ABC and PerCP-anti-mouse CD11b monoclonal antibodies (Pharmingen, 5 μg/ml) at 4°C for 1 hour. After extensive washing, the percentages of CFSE+ HLA-ABC+ hESC-MSCs were determined by flow cytometry on a FACScalibur (Becton Dickinson) and expressed as the number of CFSE+ cells per 100 mg of tissue. To rule out the possibility of detection of potential hESC-MSCs phagocytosed by host macrophages, HLA-ABC+ and/or CFSE+ cell analysis was determined in the CD11b-negative cell population.

Statistical Analysis

All data are expressed as mean ± SD of the mean. Statistical comparisons between experimental groups were performed with either a paired Student's t test or Duncan's multiple range test after two-way analysis of variance. Statistical significance was defined as a p value < .05.

RESULTS

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

Inhibition of SMAD-2/3 Signaling Promotes Enrichment of Human ESC-Derived MSCs

To determine the role of SMAD-2/3 inhibition on the potential generation of MSCs from hESCs, we used SB-431542, a robust chemical inhibitor of ALK4, ALK5, and ALK7 receptors inducing shutdown of TGF-β signaling through SMAD-2/3 downstream effectors [17]. All hESC lines analyzed displayed very similar trend so that data from the distinct lines (H9, SHEF1, AND-1, AND-2) was pooled. Undifferentiated hESC cultures readily present a basal percentage (5%–10%) of cells with a MSC-like phenotype (CD73+CD90+CD34−) throughout a 28-day period (vehicle-treated hESCs; Fig. 1A). However, treatment of undifferentiated hESC cultures with 10 μM of SB-431542 gradually induced the emergence of CD73+CD90+CD34− MSC-like cells (Fig. 1A). After 28 days, 42% (range: 37%–48%) of the cells present in the SB-431542-treated cultures were CD73+CD90+CD34− (Fig. 1A). Addition of TGF-β1 did not alter either culture homeostasis or impair the emergence of MSC-like cells due to the high levels of basal TGF-β1 already present in the conditioned media [17, 23]. As expected, on SB-431542 treatment, the emergence of MSC-like cells occurred in parallel to the gradual loss of pluripotency-associated markers such as Oct4 (Fig. 1B) and Tra-1-60 (Fig. 1C). Figure 1D shows a representative flow cytometry analysis depicting the expression of CD73+CD90+ MSC-like cell population and the expression of Oct4 and Tra-1-60 under the different culture conditions. Morphologically, after 28 days of SB-431542 treatment, hESC cultures looked fully differentiated and composed of fibroblastoid cells, whereas hESC cultures treated with either DMSO or TGF-β1 displayed an undifferentiated morphology, being comprised by many round typical undifferentiated ESC colonies surrounded by some differentiated cells (Fig. 1E). After treatment with SB-431542, hESC cultures displayed a dephosphorylation (inhibition) of SMAD-2/3 but not of SMAD-1/5/8, confirming the potent and specific effect of the SB-431542 chemical inhibitor in blocking signaling through SMAD-2/3 but not via SMAD-1/5/8 (BMP signaling effectors; Fig. 1F). Together, these data reveal an efficient differentiation of hESCs into MSC-like cells through inhibition of the SMAD-2/3 signaling pathway.

thumbnail image

Figure 1. Kinetics of the emergence of mesenchymal stem cells (MSCs) from SB-431542-treated human ESCs (hESCs). (A): Emergence of MSC-like cells (CD73+ CD90+ CD34−) from four distinct SB-431542-treated hESCs over a 28-day period. Concurrent loss of the ESC-associated markers Oct-4 (B) and Tra-1-60 (C) in SB-431542-treated hESCs over a 28-day period. Continuous lines represent SB-431542 treatment. Dotted lines represent TGF-β1 treatment. Dashed lines represent vehicle treatment. Data from the distinct hESC lines was pooled as very similar trend was obtained for each specific hESC line. (D): Representative flow cytometry analyses for CD73 and CD90 in hESC cultures treated with SB-431542 or TGF-β1 or DMSO. Expression of Oct4 and Tra-1-60 is also shown. An irrelevant isotype-matched antibody was always used (right panels). (E): Phase-contrast morphology of hESC cultures untreated (top panel) or treated with either SB-431542 (middle panel) or TGF-β1 (bottom panel). (F): Western blot analysis confirming dephosphorylation of SMAD-2/3 in hESCs treated with SB-431542 (left panel). Addition of TGF-β1 to the hESC cultures does not enhance expression of SMAD-2/3. The SB-431542 inhibitor is highly specific for the SMAD2/3 signaling pathway, not altering the bone morphogenic proteins-linked SMAD-1/5/8 pathway. Abbreviations: 7-ADD, 7-actinomycin D; FITC, fluorescein isothyocianate; PE, phyco erythrin; TGF-β1, transforming growth factor-β1.

Download figure to PowerPoint

SB-431542 ALK1/5/7 Inhibitor Fails to Block SMAD-2/3 Signaling Pathway in hiPS Cells Which, In Turn, Do Not Generate MSCs on SB-431542 Treatment

Little information exists about how close hESC and iPS cell cultures are in terms of cellular and molecular mechanisms regulating self-renewal versus lineage-specific differentiation [32]. We next wanted to assess whether the inhibition of the SMAD-2/3 signaling pathway by means of SB-451342 treatment also facilitates the enrichment of MSCs from distinct iPS cell lines. Surprisingly, in contrast to our findings in hESCs, treatment with the SB-431542 inhibitor did not exert a similar effect in the generation of MSCs from iPS cell lines. As shown in Figure 2A, SB-431542 treatment did not promote the emergence of CD73+CD90+CD34− MSCs from iPS cells, even over an extended period of up to 35 days of SB-431542 treatment. Consistently, the expression of Oct-4 (Fig. 2B) was not decreased in SB-431542-treated iPS cultures, whereas the expression levels of Tra-1-60 slightly decreased overtime (Fig. 2C).

thumbnail image

Figure 2. Inhibition of the SMAD-2/3 signaling pathway does not enrich mesenchymal stem cells (MSCs) from iPS cell cultures. (A): In contrast to human ESCs, two distinct iPS cell lines failed to differentiate into MSC-like cell on SB-431542-treatment over a 35-day period. Expression of the pluripotent-associated markers Oct-4 (B) and Tra-1-60 (C) in SB-431542-treated iPS cells over a 35-day period. (D): Western blot analysis showing that the SB-431542 treatment (used at different concentrations and time points) fails to inhibit SMAD-2/3 signaling pathway in human iPS cells. Abbreviations: CB, cord blood; iPS, induced pluripotent stem; SMAD, •TGF-β, transforming growth factor-β.

Download figure to PowerPoint

We, thus, wanted to analyze whether the SB-431542 inhibitor is equally functional in iPS cells as compared with hESCs. Human iPS cells were treated with SB-451342 or DMSO as negative control and the capacity of the SB-431542 inhibitor of blocking signaling through SMAD-2/3 was determined by western blot under the same conditions used for hESCs. Surprisingly, in contrast to hESCs, 60 minutes treatment with 10 μM of SB-451342 was unable to dephosphorylate SMAD-2/3. To confirm this data, we subsequently tested different treatment conditions (15 minutes vs. 60 minutes and 10 μM vs. 20 μM). SB-451342 consistently failed to inhibit the signaling through SMAD-2/3. This data suggests that iPS cells are resistant to SMAD-2/3 inhibition via this specific chemical inhibitor (Fig. 2D). Together, this data seems to suggest the existence of potential differences in the way different developmental signaling pathways should be modulated in hESCs versus iPS cells.

hESC-Derived MSCs Display Bona Fide MSC Phenotype, Growth Properties, and Multilineage Differentiation Potential

To fully characterize these hESC-derived MSC-like cells, the CD73+CD90+ and CD73−CD90+ cell subsets were FACS-isolated (Fig. 3A). CD73−CD90+ cells did not establish a stable culture. Indeed, sorted cells barely attached to the tissue culture plates and survived under mesenchymal culture conditions in suspension, eventually dying off after less than a week. In contrast, the CD73+CD90+ FACS-isolated population rapidly attached to the tissue culture plastic allowing the rapid establishment of long-term cultures of hESC-MSCs. These hESC-MSCs displayed a typical MSC phenotype: CD73+ CD90+ CD105+ CD44+ CD166+ CD45− CD34− CD14− CD19− HLA-DR− (Fig. 3B). These hESC-MSCs could be easily maintained for many passages displaying a faster growth than bone marrow-derived (BM-MSCs) but similar to CB-MSCs (Fig. 3C). More importantly, these hESC-MSCs displayed osteogenic (alizarin red staining), adipogenic (oil red staining), and chondrocytic (alcian blue staining) differentiation potential, even a month after establishment of hESC-MSCs cultures (Fig. 3D). Together, this characterization confirms that the MSCs derived from hESCs display bona fide MSC phenotype, cell growth and multipotent differentiation features.

thumbnail image

Figure 3. Morphological, phenotypic, and functional characterization of hESC-derived MSCs. (A): Cell sorting isolation of CD73+CD90+ and CD73− CD90+ cell subsets. Purity was consistently >85%. (B): Phenotypic profile of established cultures of hESC-derived MSCs. (C): In vitro cell growth, measured as cumulative population doublings, of hESC-MSCs, CB-derived MSCs, and BM-derived MSCs. (D): Osteogenic, adipogenic, and chondrogenic differentiation potential of hESC-MSCs. Abbreviations: BM, bone marrow; CB, cord blood; FITC, fluorescein isothyocyanate; hESC, human ESC; MSC, mesenchymal stem cell.

Download figure to PowerPoint

hESC-MSCs Exert Potent Immunosuppressive and Anti-Inflammatory Properties In Vitro

The potential of hMSCs in regenerative medicine is in part due to their immunomodulatory properties, thus, emerging as attractive candidates for the treatment of immune disorders. We, therefore, investigated the ability of hESC-MSCs to inactivate T-cell responses and to inhibit inflammatory responses. The potential immunomodulatory activity of hESC-MSCs was compared with those of Ad-MSCs and CB-MSCs, which have been previously characterized as potent immunosuppressor cells [6, 33, 34]. The addition of hESC-MSCs to MLCs of PBMCs from different donors significantly reduced the number of total cells in the culture and specifically decreased the number of cycling CD4 T-cells (Fig. 4A). hESC-MSCs were very efficient inhibiting the proliferative response of activated T cells, showing a dose-dependent effect with significant suppressive responses at a ratio as low as 1 hESC-MSC for every 10 PBMCs (Fig. 4B). As they lack class II MHC (HLA-DR, Fig. 3B) and costimulatory molecules (CD80 and CD40; Fig. 4C), hESC-MSCs did not stimulate the proliferation of allogeneic PBMCs (data not shown), supporting their “immuneprivilege” status. Moreover, hESC-MSCs significantly inhibited the production of the Th1-cytokines IL-2, IFNγ, and TNFα on the allogeneic MLCs (Fig. 4D). However, neither IL-4 (a Th2-cytokine) nor TGF-β1 (an immunosuppressive cytokine) was significantly affected by hESC-MSCs (Fig. 4D). Interestingly, the immunosuppressive effects of hESC-MSCs were only observed on allogeneic reactions, but not on syngeneic cultures (Fig. 4A, 4D), indicating the requirement of activated immune responses in such action. Moreover, the effects of hESC-MSCs were not a consequence of induced cell mortality, as they did not affect the apoptosis/survival of PBMCs on the MLCs (data not shown). Overall, the immunesuppressive effect of hESC-MSCs on T cells was similar than that shown by CB-MSCs, but a bit lower than that observed for Ad-MSCs (Fig. 4A, 4B, 4D), confirming that they are ontogenically closer to CB-MSCs than to adult adipose tissue.

thumbnail image

Figure 4. hESC-MSCs display potent immunosuppressive and anti-inflammatory effects in vitro. (A): MLCs were established by coculturing PBMCs from donor A (105 cells) and carboxybluorescein diacetate succinimidyl ester (CFSE)-labeled responder PBMCs from donor B (105 cells). Cultures of PBMCs from donor B (2 × 105 cells) were used as basal controls (syngeneic). A total of 2 × 104 hESC-MSCs, CB-MSCs, or Ad-MSCs were added to MLCs, and the total number of cells was determined after 4 days of culture (upper panel). The number of cycling (CFSEmild/low) CD4+ cells was determined by flow cytometry (lower panel). Cultures of PBMCs from donor B (2 × 105 cells) were used as basal controls (syngeneic). *, p value < .001 versus MLC without MSCs. (B): Dose-dependent effect of hESC-MSCs on lymphocyte proliferation. hESC-MSCs, CB-MSCs, or Ad-MSCs were added at different ratios to allogeneic MLCs (105 PBMCs from donor A and 105 PBMCs from donor B). Proliferation was determined by measuring [3H]-thymidine incorporation after 96 hours culture. Solid squares represent the basal proliferation of each MSC type. PBMCs from donor A (2 × 105 cells) were used as control of basal proliferation (gray column). *, p value < .001 versus MLC without MSCs. (C): hESC-MSCs do not express the costimulatory molecules CD40 and CD80. (D): hESC-MSCs decrease the production of cytokines by activated lymphocytes. hESC-MSCs, CB-MSCs, or Ad-MSCs (2 × 104) were added to allogeneic MLCs (105 PBMCs from donor A and 105 PBMCs from donor B) or syngeneic cultures (2 × 105 PBMCs from donor B). Cytokine contents in the supernatants were determined by ELISA after 48 hours culture. *, p value < .001 versus MLC without MSCs. (E): hESC-MSCs inhibit the inflammatory response in synovial cells from patients with rheumatoid arthritis (RA). Fibroblast-like synovial cells (2 × 105) isolated from two patients with RA were incubated with medium (unstimulated) or stimulated with lipopolysaccharide (1 ng/ml, for cytokine determination) or TNF-α (20 ng/ml, for collagenase activity assay) in the absence or presence of hESC-MSCs, CB-MSCs, or Ad-MSCs (105). Culture supernatants were assayed for collagenase activity (after 24 hours) or cytokine contents (after 48 hours). *, p value < .001 versus FLS alone. Abbreviations: Ad-MSC, adipose-derived mesenchymal stem cell; CB-MSC, cord blood-derived MSC; FLS, fibroblast-like synoviocyte; hESC, human ESC; IL, interleukin; IFNγ, interpheron γ; MLC, mixed lymphocyte cultures; MSC, mesenchymal stem cell; PBMC, peripheral blood mononuclear cell; TGF-β, transforming growth factor-β; TNFα, tumor necrosis factor α.

Download figure to PowerPoint

We next investigated the capacity of hESC-MSCs to regulate the inflammatory response of resident cells of the synovial membrane in patients with active RA. FLSs isolated from RA patients were cultured in the presence of hESC-MSCs and assayed for the production of inflammatory mediators and matrix-degrading enzymes (collagenase activity). hESC-MSCs treatment of RA FLS cultures decreased the secretion of TNFα but not IL-6 after stimulation with LPS and reduced the collagenase activity after stimulation with TNFα (Fig. 4E). These data reveal that the hESC-MSCs display potent immunosuppressive and anti-inflammatory effects, a key functional immune feature of putative MSCs.

Treatment with hESC-MSCs Protects Against Experimental Inflammatory Bowel Disease In Vivo Confirming hESC-MSCs Immunosuppressive and Anti-Inflammatory Properties In Vivo

On the basis of their in vitro immunosuppressive and anti-inflammatory properties, we next investigated the potential therapeutic action of hESC-MSCs in an in vivo experimental model of inflammatory bowel disease induced by intrarectal infusion of TNBS, which displays clinical, histopathological, and immunological features of human Crohn's disease [35]. In both Crohn's disease and TNBS-induced colitis, activated Th1 and Th17 cells promote an exaggerated macrophage and neutrophil infiltration and activation, giving rise to a prolonged severe transmural inflamed intestinal mucosa, characterized by uncontrolled production of inflammatory cytokines and chemokines [35]. Inflammatory mediators such as cytokines and free radicals, produced by infiltrating cells and resident macrophages, play a critical role in colonic tissue destruction. As shown in Figure 5A, TNBS-treated mice developed a severe illness characterized by bloody diarrhea, rectal prolapse, pancolitis accompanied by extensive wasting syndrome, and profound and sustained weight loss resulting in 60% mortality. Macroscopic examination of colons showed striking hyperemia, inflammation, and necrosis (Fig. 5B). However, mice treated with hESC-MSCs, similar to those treated with CB-MSCs or Ad-MSCs, displayed an increased survival rate, rapidly recovered body weight loss, improved the wasting disease, regained a healthy appearance, and only showed slight signs of colon inflammation, similar to control mice treated with vehicle (50% ethanol; Fig. 5A, 5B). Histological examination of the colons showed that hESC-MSC-treatment reduced the TNBS-induced transmural inflammation, depletion of mucin-producing globet and epithelial cells, disseminated fibrosis, focal loss of crypts, and infiltration of inflammatory cells (Fig. 5C). Moreover, hESC-MSC treatment significantly decreased colonic MPO activity, reflecting less neutrophil infiltration in the lamina propria (Fig. 5D). Together, these data confirm the therapeutic effect in vivo of hESC-MSCs on an experimental model of colitis, demonstrating for the first time that hESC-MSCs exhibit potent immunosuppressive and anti-inflammatory properties not only in vitro but also in vivo.

thumbnail image

Figure 5. Treatment with hESC-MSCs protects against experimental colitis in vivo. (A): Colitis was induced by intracolonic administration of TNBS. Mice (10 mice per group) were treated i.p. with 106 hESC-MSCs, CB-MSCs, or Ad-MSCs, 12 hours after TNBS injection. Control mice received 50% ethanol (vehicle). Clinical evolution was monitored by body weight changes, colitis score, and survival. (B): Colon length and weight and macroscopic colonic damage score was evaluated at day 2. (C): Histopathology was determined 2 days after transplant (4–6 mice per group) Scale bar = 100 μm. (D): Neutrophil infiltration was assayed by determining MPO activity in the colons isolated at day 2 (4–6 mice per group). (E): CFSE-labeled hESC-MSCs were injected i.p. to colitic mice (3 mice per group) and their presence in various organs determined by flow cytometry 2 days after injection and expressed as the mean number of CFSE+HLA-ABC+CD11b cells per 100 mg tissue. *, p value < .001 versus TNBS-colitic mice. Abbreviations: Ad-MSC, adipose-derived mesenchymal stem cell; CB-MSC, cord blood-derived MSC; CFSE, carboxybluorescein diacetate succinimidyl ester; hESC, human ESC; MLN, mesenteric lymph node; MPO, myeloperoxidase; TNBS, 2,4,6-trinitrobenzene sulfonic acid.

Download figure to PowerPoint

Finally, to better understand the trafficking of the infused hESC-MSCs, we injected CFSE-labeled hESC-MSCs into colitic mice. As previously described for Ad-MSCs [31], we detected the inoculated hESC-MSCs in the draining lymph nodes and spleen, as well as in the inflamed colon of the recipients 2 days postinjection (Fig. 5E). Interestingly, hESC-MSCs barely homed to noninflamed colon and other regions of the gastrointestinal tract (salivary glands, stomach, intestine) or to other organs/tissues (muscle, kidney, or brain). Moreover, similar to MSCs from other sources, hESC-MSCs are significantly retained in the lung and liver (Fig. 5E).

DISCUSSION

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

Human ESCs and iPS cells hold the promise as a potentially unlimited source of functional cell types for regenerative medicine and drug screening [36]. One of the key challenges used to fulfill this potential is the ability to reproducibly direct the differentiation of these pluripotent stem cells to selective cell fates in vitro and in vivo. Because of their biological properties, the generation of a potentially unlimited number of functional MSCs from hESC/iPS cells would be a great step forward in regenerative medicine.

MSCs from adult somatic tissues differentiate in vitro and in vivo into multiple mesodermal tissues including bone, cartilage, adipose tissue, tendon, ligament, or even muscle. They also home to damaged tissues where they exert their therapeutic potential [31, 33]. A striking feature of the MSCs is their immunosuppressive properties in vitro. Their multilineage differentiation potential coupled with their immunoprivileged properties is being exploited worldwide for cell replacement strategies and autoimmune diseases [4, 6]. Despite clear in vivo evidence of mesenchymal specification from hESCs, surprisingly little work has been performed to develop enrichment protocols for MSCs from hESCs/iPS cells [37]. There have been reports on the derivation of specific mesenchymal cell types from hESCs. For instance, hESCs were differentiated into mineralizing bone [38, 39]. Improvements in osteogenic differentiation in vitro were also reported [40] and Xiong et al. [41] showed adipogenic differentiation from hESCs.

However, little work has been performed in the derivation of multipotent MSCs rather than specific mesechymal derivatives from hESCs/iPS cells. Barberi et al. [13] and Trivedi and Hematti [42] were the first to derive MSCs from hESCs through a xeno-coculture of hESC with the OP9 murine stromal cell line. Olivier et al. [43] also reported the derivation of MSCs from hESCs through a feeder free culture system. However, the methodology employed in this study consisted in culturing MSCs from hESCs that were grown into a thick multilayer epithelium. Unfortunately, there is no information available about the mechanistic insights involved in the specific differentiation of hESCs into functional MSCs. Here, we demonstrate for the first time the involvement of the SMAD-2/3 signaling pathway inhibition in the efficient generation/enrichment of MSCs from hESCs. We previously reported that specific inhibition of TGF-β signals through SMAD-2/3 results in loss of the hESC phenotype and pluripotency through induction of hESC differentiation, while not affecting survival or self-renewal of hESCs [17]. Furthermore, TGF-β is considered a potent inhibitor of mesodermal tissue [18–20]. We, therefore, confirmed our initial hypothesis that similar to what occurs in the hematopoietic system, TGF-β signaling through SMAD-2/3 is a negative regulator in the MSC generation from hESCs. Human ESC-MSCs easily established long-term cultures and could be maintained for many passages displaying a faster growth than somatic MSCs while maintaining MSC morphology and phenotype. They also displayed osteogenic, adipogenic, and chondrocytic multipotent differentiation potential and lacked signs of oncogenic transformation [17]. An approximate quantification of our differentiation experiments revealed that about 30% of the hESC-MSCs differentiate into adipogenic lineage and about 70% differentiate into osteogenic or chondrogenic lineages. This data is in line with the data reported for MSCs from many other tissues indicating that the hESC-MSCs seem also hierarchically organized and therefore, functionally heterogeneous within a phenotypically homogeneous culture [44–48].

To further determine whether these hESC-MSCs are bona fide MSCs, we analyzed for the first time their immunological properties in vitro and in vivo. Their immunotolerance properties have been previously suggested in vitro [14], but no information is available about the potential therapeutic action of hESC-MSCs in vivo. Here, we demonstrate how hESC-MSCs displayed potent immunosuppressive and anti-inflammatory properties in vitro and in vivo in an experimental model of inflammatory bowel disease or colitis model [31, 33]. In vitro, they were very efficient inhibiting the proliferative response of T cells, they lacked class II HLA and costimulatory molecules, and did not stimulate the proliferation of allogeneic PBMCs. Moreover, hESC-MSCs significantly inhibited the production of the Th1-cytokines IFNγ, IL-2, and TNFα but not the Th2-cytokine IL-4 on MLC assays. Overall, the immunosuppressive effect of hESC-MSCs on T cells was similar to that shown by CB-MSCs, but a bit lower than that observed for Ad-MSCs, confirming that they are ontogenically closer to CB-MSCs than to adult adipose tissue. They also regulated the inflammatory response of resident cells of the synovial membrane. FLS isolated from patients with active RA cultured in the presence of hESC-MSCs produced less inflammatory mediators and matrix-degrading enzymes.

More importantly, in vivo, using an experimental model of inflammatory bowel disease, mice treated with hESC-MSCs similar to those treated with CB-MSCs or Ad-MSCs, displayed an increased survival rate, rapidly recovered body weight loss, improved the wasting disease, regained a healthy appearance, and significantly reduced colon inflammation, similar to control mice treated with vehicle, demonstrating for the first time that hESC-MSCs exhibit potent immunosuppressive and anti-inflammatory properties not only in vitro but also in vivo. This immunosuppressive effect could be exerted locally on the colonic mucosa, or peripherally on lymphoid organs, as suggested by the fact that hESC-MSCs show preferential homing for inflamed tissues and secondary lymphoid organs, in addition to the systemic filter organs (lung and liver).

Finally, whether hESCs and iPS are equally capable of generating MSCs remains to be assessed and little information exists about how close hESC and iPS cultures are in terms of cellular and molecular mechanisms regulating self-renewal versus lineage-specific differentiation [32]. Recently, differences in the differentiation capacity between hESCs and iPS cells were reported [49]. Accordingly, we found that in contrast to that observed in hESCs, treatment with SB-431542 did not exert a similar effect in the generation of MSCs from distinct iPS cell lines, even over an extended period of up to 35 days of SB-431542 treatment. Interestingly, further experiments revealed that the SB-451342 inhibitor consistently failed to inhibit the signaling through SMAD-2/3 in hiPS cells, suggesting that hiPS cells are resistant to SMAD-2/3 inhibition via this specific chemical inhibitor. Whether alternative chemical inhibitors or overexpression of ectopic SMAD-6 may block this pathway in hiPS cells remains to be elucidated. Alternatively, we found that many hiPS cells reactivate the ectopic reprogramming factors during differentiation toward blood and neuroectodem (Ramos-Mejía et al., manuscript in preparation). Whether these ectopic factors may also reactivate their expression during hiPS cell differentiation toward MSCs should also be addressed in future studies to gain insight into whether constitutive expression of ectopic reprogramming factors (Oct-4, Klf-4, c-myc, and Sox-2) may be comparing signaling through TGF-ALKs-SMAD-2/3. Finally, it is worth assessing in future studies that whether iPS cells generated without c-myc or using a different cocktail of ectopic reprogramming factors (i.e., Oct-4, Nanog, Sox-2, and Lin28) are equally resistant to SMAD-2/3 inhibition. This data, although preliminary, suggests potential differences in the cellular and molecular mechanisms underlying the generation of MSCs from hESCs versus iPS cells. These still uncharacterized differences should be kept in mind when attempting to differentiate hESCs or iPS cells into a certain lineage. It is plausible that cellular mechanisms and molecular determinants involved in the stepwise differentiation may differ, and therefore the experimental data gained in one model can not be extrapolated to the other model.

CONCLUSION

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

We show for the first time how specific inhibition of SMAD-2/3 signaling pathway in hESCs facilitates the conversion of hESCs into multipotent MSCs with fully immunosuppresive and anti-inflammatory properties in vivo, opening up new avenues for potential future clinical applications.

Acknowledgements

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

We are indebted to Prof. Jose Cibelli (Michigan State University) for provision of the iPS cell line MSUH001. This work was supported by the CSJA (0030/2006 to P.M. and 0108/2007 to R.R.) and CICE (P08-CTS-3678 to P.M.), de la Junta de Andalucía, the FIS/FEDER to P.M. (PI070026 and PI100449), C.B. (CP07/00059), and P.J.R. (CP09/0063), the MINICC to P.M. (PLE-2009-0111), and the Marie Curie IIF to V.R.-M. (PIIF-GA-2009-236430). R. Rodriguez is supported by a Fellowship from the AECC. J.L.G.-P.'s group is supported by ISCIII-CSJA (EMER07/056), by a Marie Curie IRG action (FP7-PEOPLE-2007-4-3-IRG), by CICE (P09-CTS-4980) from Junta de Andalucía, by Proyectos de Investigación en Salud PI-002 from Junta de Andalucia, and by the FIS/FEDER (CP07/00065 and PI08171). M.D.'s group is supported by FIS/FEDER (PS09-00928) and Junta de Andalucia (P09-CTS-4723).

References

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