Sox2 Suppression by miR-21 Governs Human Mesenchymal Stem Cell Properties

The miRNA profile of mesenchymal stem cells (MSCs) derived from amniotic fluid, bone marrow (BM), and umbilical cord blood was analyzed. Initially, 67 different miRNAs were identified that were expressed in all three types of MSCs but at different levels, depending on the source. A more detailed analysis revealed that miR-21 was expressed at higher levels in RS-AF-MSCs and BM-MSCs compared with SS-AF-MSCs. Findings suggest that miR-21 might function by regulating Sox2 expression in human MSCs and might also act as a key molecule determining MSC proliferation and differentiation.

Abstract

MicroRNAs (miRNAs) have recently been shown to act as regulatory signals for maintaining stemness and for determining the fate of adult and fetal stem cells, such as human mesenchymal stem cells (hMSCs). hMSCs constitute a population of multipotent stem cells that can be expanded easily in culture and are able to differentiate into many lineages. We have isolated two subpopulations of fetal mesenchymal stem cells (MSCs) from amniotic fluid (AF) known as spindle-shaped (SS) and round-shaped (RS) cells and characterized them on the basis of their phenotypes, pluripotency, proliferation rates, and differentiation potentials. In this study, we analyzed the miRNA profile of MSCs derived from AF, bone marrow (BM), and umbilical cord blood (UCB). We initially identified 67 different miRNAs that were expressed in all three types of MSCs but at different levels, depending on the source. A more detailed analysis revealed that miR-21 was expressed at higher levels in RS-AF-MSCs and BM-MSCs compared with SS-AF-MSCs. We further demonstrated for the first time a direct interaction between miR-21 and the pluripotency marker Sox2. The induction of miR-21 strongly inhibited Sox2 expression in SS-AF-MSCs, resulting in reduced clonogenic and proliferative potential and cell cycle arrest. Strikingly, the opposite effect was observed upon miR-21 inhibition in RS-AF-MSCs and BM-MSCs, which led to an enhanced proliferation rate. Finally, miR-21 induction accelerated osteogenesis and impaired adipogenesis and chondrogenesis in SS-AF-MSCs. Therefore, these findings suggest that miR-21 might specifically function by regulating Sox2 expression in human MSCs and might also act as a key molecule determining MSC proliferation and differentiation.

Introduction

Human mesenchymal stem cells (MSCs) represent a multipotent stem cell population that is able to self-renew and differentiate into multiple cell lineages [1]. Adult MSCs derived from bone marrow (BM) have been widely studied; however, recent attention has focused on MSCs derived from fetal sources, such as amniotic fluid (AF) [2–7] or umbilical cord blood (UCB) [8, 9]. Our group has recently identified MSCs from human second trimester AF obtained during routine amniocenteses for prenatal diagnosis [4, 5]. Two morphologically distinct adherent cell types were isolated, which were termed spindle-shaped (SS) and round-shaped (RS) cells. The two cell types exhibited differences in phenotype, pluripotency, proliferation rate, differentiation potential, and proteomic profile [4]. Both subpopulations expressed mesenchymal stem cell markers such as CD73, CD105, CD166, and integrins CD29 and CD49e at similar levels. However, SS-AF-MSCs expressed higher levels of CD90 and CD44 antigens compared with RS-AF-MSCs [4, 10]. SS-AF-MSCs express pluripotency markers such as Sox2 (SRY sex determination SRY region Y-box2), Oct4 (octamer-binding transcription factor 4) and the homeobox transcription factor Nanog [4, 5, 11]. SS-AF-MSCs also exhibit a high proliferation rate in culture and differentiate in vitro not only into mesoderm-derived cell types but also into endoderm-derived cells, such as hepatocytes [4–6]. Indeed, we have previously demonstrated that SS-AF-MSCs can be expanded in culture at higher levels than BM-MSCs, and this occurs without karyotypic changes and with the maintenance of their capacity to differentiate into osteogenic, adipogenic, and chondrogenic cells [2, 4, 5].

MicroRNAs (miRNAs) are single-stranded noncoding RNA sequences of 19–23 nucleotides that act as post-transcriptional regulators of gene expression by base pairing to partially complementary sequences in the 3′ untranslated region (UTR) of multiple target mRNAs, resulting in silencing of the mRNA [12–14]. Approximately 40%–90% of human protein-coding genes are predicted to be regulated by miRNAs at the translational level [15]. Recent studies have suggested that miRNAs are involved in embryonic and adult stem cell fate by regulating biological processes such as cell proliferation [16, 17], differentiation [18, 19], cell cycle [20, 21], and apoptosis [22].

Because stem cell populations derived from different sources manifest specific miRNA signatures, in this study we attempted to identify and compare the miRNA profiles of AF-MSCs, BM-MSCs, and UCB-MSCs to systematically clarify the post-transcriptional regulation of MSCs from various sources. In particular, we wanted to obtain further insight into the miRNA patterns that are characteristic of AF-MSC subpopulations and to identify specific miRNAs that may play an important role in their molecular identity. More importantly, we provide functional evidence that miR-21 has a key role in both subpopulations of AF-MSCs and is involved in the suppression of the transcription factor Sox2. In addition, our findings suggest that the induction of miR-21 expression inhibits the expression of other pluripotency genes and alters the proliferation rate, the cell cycle, and the differentiation properties of AF-MSCs.

Materials and Methods

Isolation and Culture of Human MSCs

All protocols involving human subjects were approved by the Ethical Committee of Alexandra Hospital, Athens, Greece; the Bioethics Committee of the School of Medicine of the University of Athens; and the Bioethics Committee of the Biomedical Research Foundation of the Academy of Athens (BRFAA). All samples were collected after informed consent was obtained from each individual. The AF-MSCs were isolated and cultured according to methods described in previous studies [2, 4–6]. SS and RS colonies were selected and subcultured [4]. BM samples from healthy volunteers were obtained from posterior iliac crest aspirates, as described previously [5, 23]. Bone marrow mononuclear cells were plated at a density of 2 × 10⁵ cells per cm² in Dulbecco’s modified Eagle’s medium (DMEM; Sigma-Aldrich, Gillingham, U.K., http://www.sigmaaldrich.com) supplemented with 20% fetal bovine serum (Gibco-BRL, Paisley, U.K., http://www.invitrogen.com) and incubated at 37°C in a 5% (v/v) CO₂ humidified chamber for approximately 20 days, until the first BM-MSC colonies appeared in culture. UCB samples were processed within 24 hours. UCB-MSCs were isolated by Ficoll-Paque density centrifugation (Sigma-Aldrich) and cultured in DMEM (Gibco-BRL) containing 20% fetal bovine serum (Gibco-BRL), according to the method described previously [24].

miRNA Microarray Analysis

Total RNA, including miRNAs, from three AF-MSC, three BM-MSC, and two UCB-MSC samples were extracted using Trizol reagent according to the manufacturer’s protocol (Invitrogen, Paisley, U.K., http://www.invitrogen.com). The samples were labeled using a miRCURY Hy3/Hy5 labeling kit (Exiqon A/S, Vedbaek Denmark, http://www.exiqon.com/ls) and hybridized on an miRCURY LNA Array (v.8.1) (Exiqon A/S), which consists of control probes, mismatch probes, and 427 capture probes that perfectly match all of the human miRNAs registered and annotated in miRBase release 7.1 at the Wellcome Trust Sanger Institute. Analysis of the scanned slides indicated that labeling was successful because all of the capture probes for the control-spiked oligonucleotides produced signals within the expected range. The quantified signals were normalized using the global Lowess (Locally Weighted Scatterplot Smoothing) regression algorithm, which minimizes the intensity-dependent differences between the dyes. The expression matrix contains normalized Hy3/Hy5 ratios (log2 transformed) from all hybridizations. The subset of differentially expressed miRNAs was used to construct an unsupervised hierarchical clustering of the different samples (Exiqon, A/S). For in silico data analysis, we used the GOmir program (http://www.bioacademy.gr/bioinformatics/projects/GOmir/), which integrates miRNA target prediction and functional analysis by combining the predicted target genes from the TargetScan (http://www.targetscan.org/), miRanda (http://www.microrna.org/microrna/home.do), RNAhybrid (http://bibiserv.techfak.uni-bielefeld.de/rnahybrid/), and PicTar (http://pictar.mdc-berlin.de/) computational tools and provides a full gene description and functional analysis for each target gene [25, 26].

Real-Time Quantitative Polymerase Chain Reaction

This experimental approach used a well-established protocol, the details of which are provided in the supplemental online data.

Western Blots

A detailed protocol is provided in the supplemental online data.

Lentiviral Vector Construction and Transduction of AF-MSCs

A four-plasmid lentiviral expression system containing pCCLsin.cPPT.hEF1a.ΔLNGFR.Wpre [27], a kind gift from Dr. L. Naldini (University School of Medicine, Milan, Italy), which includes the elongation factor 1a (EF1a), was used to induce miR-21 expression. To construct the lentiviral vector expressing human pre-miR-21, the pre-miR-21 sequence was amplified by polymerase chain reaction (PCR) from human genomic DNA. We cloned the pre-miR-21 gene into the pCCLsin.cPPT.hEF1a.ΔLNGFR.Wpre plasmid using the following primers: forward 5′-TCGACTCGAGGTTCGATCTTAACAGG-3′ and reverse 5′-TCGAACGCGTACCAGACAGAAGGACC-3′. miR-21 or EF1 (empty vector) viruses were produced by transient transfection of HEK293T cells, as previously described [2, 28], followed by collection with Amicon Ultra Centrifugal Filters-100K Units (Merck, Whitehouse Station, NY, http://www.merck.com). The lentiviral titers were determined by infection of HT1080 cells with serial dilutions of the concentrated viral stock. The plasmid contained a complementary DNA for the low-affinity nerve growth factor receptor (ΔLNGFR) as a positive expression marker. Nerve growth factor receptor (NGFR)-positive cells were identified by fluorescence-activated cell sorting (FACS) analysis using a PE-tagged mouse anti-human CD271 antibody (BD Biosciences, San Diego, CA, http://www.bdbiosciences.com) and a Beckman Coulter Cytomics FC 500 flow cytometer (Beckman Coulter, Palo Alto, CA, http://www.beckmancoulter.com). The titers ranged from 10⁶ to 5 × 10⁶ infectious units/ml. For transduction, 2.5 × 10⁴ SS-AF-MSCs or HEK293T cells per well were seeded into 12-well plates and lentivirus was added at a multiplicity of infection (MOI) of 4. After 7 days, the NGFR-positive cells were identified by flow cytometry and sorted using a FACSAria cell sorter (BD Biosciences). In all cases, cell sorting was performed, and the efficiency was determined to be greater than 98%.

miR-21 Inhibition and sh-Sox2 Transfection

The RS-AF-MSCs and BM-MSCs were transfected with 0.3 μM miR-21 antagonist (Exiqon A/S) and/or 1 μg pLKO.1-sh-Sox2 (Erasmus University Medical Center, Rotterdam, The Netherlands) using Lipofectamine 2000 reagent according to the manufacturer’s protocol (Invitrogen). An equal concentration of scrambled miR-21 antagonist (Exiqon A/S) or pLKO.1-sh-control (Erasmus MC) was used as a control. Thirty hours after transfection, total RNA and protein were extracted. For functional assays, pLKO.1-sh-Sox2 (sh-Sox2) or pLKO.1-sh-control (sh-control) viruses were produced by transient transfection of HEK293T cells, using a four-plasmid lentiviral expression system, as previously described. Selection was performed using puromycin (Sigma-Aldrich) at 0.5 μg/ml for 5 days.

Enhanced Green Fluorescence Protein Reporter Assay

The Sox2 3′UTR enhanced green fluorescence protein (EGFP) reporter construct was generated from genomic DNA by PCR amplification of the human Sox2 3′UTR sequence using the primers described in supplemental online Table 1. The mutated form of the Sox2 3′UTR sequence was generated by site-directed mutagenesis (Promega, Madison, WI, http://www.promega.com) according to the manufacturer’s instructions. The fragments were inserted into the pEGFP-C2 vector (Promega) downstream of the EGFP cassette. HEK293T cells were transduced with the miR-21 lentivirus, then transiently transfected with the pEGFP-C2 constructs (Sox2 3′UTR or mutant Sox2 3′UTR). Forty-eight hours later, EGFP-positive cells were identified in all tested samples by FACS analysis. Three independent experiments were performed, each including two replicates, and the mean of each experiment was calculated.

Fibroblast Colony-Forming Unit Assay

The clonogenic potential of SS-AF-MSCs expressing miR-21 or sh-Sox2 was evaluated by performing a fibroblast colony-forming unit (CFU-F) assay. Specifically, transduced SS-AF-MSCs expressing miR-21, EF1, sh-Sox2, or sh-control were plated at three clonal densities (70, 140, and 280 cells) in 60-mm plates for 14 days. The CFU-Fs were quantified by Giemsa staining. The number of CFU-Fs was estimated per 100 MSCs initially plated based on a linear regression analysis of the three different initial cell concentrations. The data are presented as the mean ± SEM of at least three independent experiments.

MTS Proliferation Assay

To determine the effect of miR-21 on proliferation, the transduced SS-AF-MSCs expressing miR-21, EF1, sh-Sox2, or sh-control were plated at a density of 10³ cells per well in 96-well plates and cultured for 3, 7, or 10 days. RS-AF-MSCs or BM-MSCs transfected with miR-21 antagonist or the scrambled sequence were plated at a density of 10³ cells per well in 96-well plates and cultured for 24 or 48 hours. At each time point, the appropriate amount of MTS reagent (Promega) was added to each well and the cells were incubated for 3 hours, according to the manufacturer’s instructions. The absorbance was recorded at 490 nm with an enzyme-linked immunosorbent assay plate reader (ELX 800; Biotek Instruments Inc., Winooski, VT, http://www.biotek.com/). The percentage increase in proliferation was calculated using the following formula: [(OD dayx − OD day0)/(OD day0) × 100]. Three independent experiments were performed, each including five replicates, and the mean ± SEM of each experiment was calculated.

Apoptosis Assay

One million transduced SS-AF-MSCs expressing miR-21, EF1, sh-Sox2, or sh-control were assessed by annexin V-fluorescein isothiocyanate staining (BD Biosciences), according to the manufacturer’s instructions. 7-Aminoactinomycin D (Sigma-Aldrich) was used for live-dead cell discrimination. Flow cytometry was performed using the Beckman Coulter Cytomics FC 500 flow cytometer (Beckman Coulter). Three independent experiments were performed, and the mean ± SEM of each experiment was calculated.

Cell Cycle Analysis

SS-AF-MSCs expressing miR-21 or EF1 were stained with propidium iodide. The cells were fixed in ice-cold 70% ethanol for at least 16 hours at 4°C. After fixation, the cells were incubated with 2 mg/ml RNase A (Sigma-Aldrich) at 37°C for 40 minutes. Then, 50 μg/ml propidium iodide (Sigma-Aldrich) was added, and the cells were analyzed by flow cytometry. The nuclear debris, and overlapping nuclei were gated out. The data are presented as the mean ± SEM of at least three independent experiments.

Senescence-associated β-galactosidase (SA β-gal) staining was performed using a cellular senescence assay kit (Merck KGaA). Briefly, 5 × 10⁴ transduced SS-AF-MSCs expressing miR-21 or EF1 were fixed in 1× fixing solution and stained for SA β-gal activity overnight. The positive (blue) cells were counted and were expressed relative to the number of positive cells in SS-AF-MSCs expressing EF1. The data are presented as the mean ± SEM of at least three independent experiments.

Adipogenic, Osteogenic, and Chondrogenic Differentiation

Detailed protocols are provided in the supplemental online data.

Statistics

The unpaired Student’s t test method was used to determine the statistical significance, and p values are indicated in the figures, where * represents p < .05, ** represents p < .01, and *** represents p < .001.

Results

Analysis of miRNA Patterns in AF-MSCs, BM-MSCs, and UCB-MSCs

We previously performed a detailed molecular and proteomic analysis of SS-AF-MSCs and RS-AF-MSCs and BM-MSCs that revealed fundamental differences between their characteristics [4, 5]. As a next step, we have attempted to analyze the post-transcriptional differences between the MSCs derived from AF, BM, and UCB by evaluating and comparing their miRNA profiles. Initially, we performed an miRNA microarray (miRCURY LNA Array v.8.1) analysis of three AF-MSC (SS-AF-MSCs and RS-AF-MSCs), three BM-MSC, and two UCB-MSC samples, each hybridized to the captured probes against the pool of all the samples. The miRNA profile revealed 67 differentially expressed miRNAs among the three MSC sources, as shown in the heat map plot (Fig. 1A). The expression matrix contains the normalized Hy3/Hy5 ratios (log2 transformed) from all of the hybridizations (supplemental online Table 2). miRNAs with p value < .05 were considered to be significantly differentially expressed. For this expression analysis, the calculated p values were based on Student’s t test. In addition, the Benjamini and Hochberg multiple testing adjustment method was applied to the p values. A detailed analysis of the expression levels of the detected miRNA in AF-MSCs, BM-MSCs, and UCB-MSCs is presented in supplemental online Fig. 1. Specifically, 16 miRNAs in AF-MSCs, 9 in BM-MSCs, and 44 in UCB-MSCs were expressed at higher levels, respectively. Similarly, 51 miRNAs in AF-MSCs, 58 in BM-MSCs, and 23 in UCB-MSCs were expressed at lower levels, respectively. Furthermore, some miRNAs, such as miR-143, miR-487, miR-326, and miR-199*, were downregulated, whereas no miRNA was upregulated in all three MSC categories. The fold expression difference was calculated for each group versus the pool of samples. To validate the microarray results, we performed real-time PCR analysis of some of the differentially expressed miRNAs, such as hsa-miR-221, hsa-miR-222, hsa-miR-210, and hsa-let-7d, which confirmed the same trends in the respective miRNA expression levels (Fig. 1B).

Figure 1.

miRNA profiles of AF-MSCs, BM-MSCs, and UCB-MSCs. (A): Hierarchical clustered heat map illustrating the differential miRNA expression profiles of AF-MSCs, BM-MSCs, and UCB-MSCs compared with the pool of all samples (statistically significant differences, p < .05). (B): Validation of the microarray through real-time PCR expression analysis of random miRNAs. The data are presented as mean ± SEM. (C): Representative photos of the cultured UCB-MSC (Ci), BM-MSC (Cii), SS-AF-MSC (Ciii), and RS-AF-MSC (Civ) morphology. Magnification, ×10. Abbreviations: AF, amniotic fluid; BM, bone marrow; miRNA, microRNA; MSC, mesenchymal stem cell; PCR, polymerase chain reaction; RS, round-shaped; SS, spindle-shaped; UCB, umbilical cord blood.

Differentially Expressed miRNAs in AF-MSC Subpopulations

We recently isolated two morphologically different subpopulations of AF-MSCs, termed RS and SS AF-MSCs [4]. These cells have distinct morphologies, exhibit phenotypic and molecular differences, and differ in their ability to differentiate into multiple cell types. Notably, the proliferation rate of SS-AF-MSCs is higher than that of RS-AF-MSCs [4]. In this study, a detailed miRNA microarray analysis of the two AF-MSC subpopulations indicated a differential miRNA expression pattern. More specifically, 32 miRNAs were differentially expressed between the SS-AF-MSCs and RS-AF-MSCs (Fig. 2A). Interestingly, RS-AF-MSCs exhibited increased expression of miR-21 (6.6-fold expression) compared with SS-AF-MSCs (0.52-fold expression), as determined by both the miRNA microarray results and real-time PCR analysis (Fig. 2B). An extensive bioinformatics analysis of miR-21 using GOmir software [25] indicated a list of 55 predicted target genes supplemented with gene descriptions and functional analyses (supplemental online Table 3). The transcription factor Sox2 was identified as a target gene containing predicted miR-21 binding sites. Our group recently showed that SS-AF-MSCs exhibited higher expression levels of Sox2 compared with RS-AF-MSCs [4]. Similarly, in this study, we confirmed Sox2 expression at the RNA and protein level in both AF-MSC populations (Fig. 2Ci, 2Cii). Based on these findings, we focused on the role of miR-21 and its binding partner Sox2 in the proliferation rate and differentiation properties of SS-AF-MSCs and RS-AF-MSCs (Fig. 2D).

Figure 2.

miR-21 expression in SS-AF-MSCs and RS-AF-MSCs. (A): Differentially expressed miRNAs in SS-AF-MSCs versus RS-AF-MSCs, according to miRNA microarray data. (B): miR-21 expression levels in SS-AF-MSCs and RS-AF-MSCs were estimated by real-time PCR. The values shown are the means ± SEM of three samples. (Ci): Analysis of Sox2 expression in SS-AF-MSCs and RS-AF-MSCs by real-time PCR. The results were normalized to the human glyceraldehyde 3-phosphate dehydrogenase-positive control and then to the SS-AF-MSCs. (Cii): Western blot analysis of Sox2 in cell extracts from SS-AF-MSCs and RS-AF-MSCs. A protein band of 37 kDa corresponding to Sox2 was detected. β-Actin was used as a positive control for equal loading. Quantification was performed using Quantity One software, and the results were normalized to the β-actin-positive control and then to the SS-AF-MSCs. The values shown are the mean ± SEM of three independent experiments (*, p < .05; **, p < .01; Student’s t test). (D): Working hypothesis for the role of miR-21 in SS-AF-MSCs and RS-AF-MSCs. Abbreviations: AF, amniotic fluid; miRNA, microRNA; MSCs, mesenchymal stem cells; PCR, polymerase chain reaction; RS, round-shaped; SS, spindle-shaped.

miR-21 Binds Directly to the Sox2 3′UTR

Based on the hypothesis that Sox2 might be a target of miR-21, a GFP reporter assay was performed to demonstrate the potential binding. The Sox2 3′UTR was fused to an enhanced green fluorescence protein (EGFP) reporter gene (supplemental online Fig. 2A) and EGFP expression was determined by FACS analysis in HEK293T cells transduced with miR-21 lentivirus. As shown in supplemental online Fig. 2A, miR-21 significantly repressed EGFP expression (EGFP expression: 15.1% ± 2.04%). Consistently, neither the Sox2 3′UTR in the absence of miR-21 (EGFP expression: 41.125% ± 4.06%) nor the mutant Sox2 3′UTR (containing a T→C mutation at the miR-21 binding site) in the presence of miR-21 (EGFP expression: 36.7% ± 2.81%) had any effect, as determined by FACS analysis and immunofluorescence microscopy (supplemental online Fig. 2B). Notably, the HEK293T cells did not exhibit endogenous miR-21 expression (supplemental online Fig. 3A). These experiments confirmed the direct interaction of miR-21 with the Sox2 3′UTR.

We then sought to evaluate the expression levels of Sox2 in SS-AF-MSCs expressing miR-21. We transduced SS-AF-MSCs with miR-21 lentivirus at an MOI of 4 and sorted the cells for NGFR (98% efficiency) (Fig. 3Ai). The miR-21 expression levels in the transduced SS-AF-MSCs were 2.5- to 9.8-fold higher than in the control cells (Fig. 3Aii). Interestingly, a statistically significant decrease in Sox2 at the RNA and protein level was observed in SS-AF-MSCs expressing miR-21 compared with EF1-transduced cells (Fig. 3B). This finding clearly indicates the direct interaction between miR-21 and the Sox2 3′UTR, which results in translational repression and target degradation.

Figure 3.

miR-21 induction in SS-AF-MSCs decreases the expression levels of Sox2, Oct4, and Nanog. (Ai): NGFR expression in SS-AF-MSCs transduced with miR-21 or EF1 lentivirus at MOI of 4 before and after cell sorting. (Aii): Real-time PCR for miR-21 detection in SS-AF-MSCs transduced with miR-21 lentivirus at MOI of 4 after sorting for NGFR. The results were normalized to the RNU44 internal control and then to the EF1-transduced cells. (B): Real-time PCR expression analysis of Sox2 (Bi) and Western blot analysis of Sox2 (Bii) in SS-AF-MSCs expressing miR-21 compared with EF1-transduced cells. (C): Real-time PCR expression analysis of Oct4 (Ci) and Western blot analysis of Oct4 (Cii) in SS-AF-MSCs expressing miR-21 compared with EF1-transduced cells. (D): Real-time PCR expression analysis of Nanog (Di) and Western blot analysis of Nanog (Dii) in SS-AF-MSCs expressing miR-21 compared with EF1-transduced cells. The real-time PCR results were normalized to the human glyceraldehyde 3-phosphate dehydrogenase-positive control and then to the EF1-transduced SS-AF-MSCs. Western blot quantification was performed using Quantity One software, and the results were normalized to the β-actin-positive control and then to EF1-transduced SS-AF-MSCs. The values presented are the mean ± SEM of three independent experiments (*, p < .05; **, p < .01; Student’s t test). Abbreviations: AF, amniotic fluid; EF1, elongation factor 1; MSCs, mesenchymal stem cells; NGFR, nerve growth factor receptor; PCR, polymerase chain reaction; PE, phycoerythrin; SS, spindle-shaped.

miR-21 Induction Decreases Oct4 and Nanog Expression in SS-AF-MSCs

Accumulating evidence suggests that Sox2 promotes pluripotency and represses differentiation in concert with Oct4 and Nanog as a core transcription factor network [29, 30]. To determine whether Sox2 inhibition by miR-21 also affects these two transcription factors, we performed real-time PCR and Western blot analysis of Oct4 and Nanog expression in transduced SS-AF-MSCs expressing miR-21 or EF1. Statistically significant decreases in Oct4 and Nanog were observed at both the mRNA and protein levels in SS-AF-MSCs expressing miR-21 compared with EF1-transduced cells (Fig. 3C, 3D). Oct4 and Nanog were not predicted as potential targets of miR-21 by in silico target prediction analysis, suggesting that the effect observed was the result of an indirect mechanism.

miR-21 Induction Alters the Cell Growth Rate of SS-AF-MSCs

To analyze the role of miR-21 in SS-AF-MSCs, a CFU-F assay was performed to determine the clonogenic potential of these cells (Fig. 4Ai, 4Aii). SS-AF-MSCs expressing miR-21 showed reduced clonogenic potential (17.1 CFU-F ± 1.59) compared with EF1-transduced cells (20.9 CFU-F ± 0.53). To further explore the role of miR-21 in the proliferation rate of SS-AF-MSCs, MTS assays were performed at 3, 7, and 10 days. miR-21 induction in SS-AF-MSCs caused a significant decrease in cell proliferation compared with EF1-transduced cells (Fig. 4Aiii). In addition, no apoptotic effect was observed in SS-AF-MSCs expressing miR-21, as determined by annexin V and 7ADD staining (Fig. 4Aiv, 4Av). Similar results were observed after transduction of SS-AF-MSCs with sh-Sox2 virus (Fig. 4Bi). In particular, Sox2 downregulation in SS-AF-MSCs reduced their clonogenic potential (3.1 CFU-F ± 0.1) compared with SS-AF-MSCs transduced with sh-control virus (14.3 CFU-F ± 1.1) (Fig. 4Bii, 4Biii) and also decreased their proliferation rate during ten days in culture (Fig. 4Biv). Finally, only 3.8% of the sh-Sox2-transduced SS-AF-MSCs were observed to be positive for annexin V, as determined by FACS (Fig. 4Bv, 4Bvi).

Figure 4.

miR-21 induction or Sox2 inhibition alters the growth of SS-AF-MSCs. (Ai): The clonogenic potential of SS-AF-MSCs expressing miR-21 was determined by CFU-F assay. The mean numbers ± SEM of CFU-F per 100 cells in a 14-day clonogenic assay are presented. The values presented are the mean ± SEM of three independent experiments (*, p < .05; Student’s t test). (Aii): Representative images of the colonies formed. (Aiii): Comparative analysis of the growth rates of miR-21 versus EF1-transduced SS-AF-MSCs during 10 days in culture. The values presented are the mean ± SEM of three independent experiments (*, p < .05; **, p < .01; Student’s t test). (Aiv): Apoptosis was examined by FACS analysis of annexin V staining in miR-21 and EF1-transduced SS-AF-MSCs. (Av): 7AAD was used for live-dead cell discrimination. (Bi): Western blot analysis of Sox2 in SS-AF-MSCs transduced with sh-Sox2 or sh-control viruses. Quantification was performed using Quantity One software, and the results were normalized to the β-actin positive control and then to the SS-AF-MSCs. (Bii): The clonogenic potential of SS-AF-MSCs transduced with sh-Sox2 or sh-control virus was determined by CFU-F assay. The mean numbers ± SEM of CFU-F per 100 cells in a 14-day clonogenic assay are presented. The values presented are the mean ± SEM of three independent experiments (**, p < .01; ***, p < .001; Student’s t test). (Biii): Representative images of the colonies formed. (Biv): Comparative analysis of the growth rates of sh-Sox2 versus sh-control transduced SS-AF-MSCs during 10 days in culture. The values presented are the mean ± SEM of three independent experiments (*, p < .05; **, p < .01; ***, p < .001; Student’s t test). (Bv): Apoptosis was examined by FACS analysis of annexin V staining in sh-Sox2 and sh-control transduced SS-AF-MSCs. (Bvi): 7AAD was used for live-dead cell discrimination. Abbreviations: 7AAD, 7-aminoactinomycin D; AF, amniotic fluid; CFU-F, fibroblast colony-forming unit; EF1, elongation factor 1; FACS, fluorescence-activated cell sorting; MSCs, mesenchymal stem cells; NGFR, nerve growth factor receptor; PCR, polymerase chain reaction; PE, phycoerythrin; SS, spindle-shaped.

To gain further insight into the mechanisms underlying the altered proliferation of SS-AF-MSCs on miR-21 induction, cell cycle analysis was performed by propidium iodide staining (Fig. 5Ai). After miR-21 induction in SS-AF-MSCs, the proportion of cells in G₀/G₁ phase increased (69.45% ± 0.85%), whereas the portion in S phase decreased (6.25% ± 1.25%), compared with EF1-transduced cells (G₀/G₁: 53.5% ± 5 and S: 8.3% ± 2.7%) (Fig. 5Aii). Moreover, G₁ cell cycle arrest was demonstrated mainly through a decrease in G₁-specific cyclin D1 at both the RNA (Fig. 5Bi) and protein levels (Fig. 5Ci). Consistently, miR-21 induction was sufficient to reduce levels of cyclin E1 and cyclin A at the protein level (Fig. 5Cii, 5Ciii). Accordingly, an increase in the expression of cyclin-dependent kinase inhibitors (CDKIs) such as p27 (CDKN1B) and p18 (CDKN2C) was also evident after miR-21 expression in SS-AF-MSCs (Fig. 5Bii). In addition, the expression level of Cdc25A (cell division cycle 25A) was decreased in SS-AF-MSCs expressing miR-21 compared with the expression level of EF1-transduced cells (Fig. 5Biii). Finally, SS-AF-MSCs expressing miR-21 induced SA β-gal activity (Fig. 5D). These observations suggest that miR-21 may regulate the proliferation of SS-AF-MSCs through a combined effect on different components of the cell cycle machinery.

Figure 5.

Cell cycle analysis in SS-AF-MSCs expressing miR-21. (Ai, Aii): Propidium iodide staining of SS-AF-MSCs expressing miR-21 and EF1-transduced cells, as determined by FACS. The values shown are the means ± SEM of three independent experiments. (Bi): Real-time PCR expression analysis of cyclin D1, cyclin E1, cyclin A2, and cyclin B2 in SS-AF-MSCs expressing miR-21 compared with EF1-transduced cells. (Bii): Real-time PCR expression analysis of p18 and p27 in SS-AF-MSCs expressing miR-21 compared with EF1-transduced cells. (Biii): Real-time PCR expression analysis of Cdc25A in SS-AF-MSCs expressing miR-21 compared with EF1-transduced cells. The real-time PCR results were normalized to the human β-actin positive control and then to the EF1-transduced SS-AF-MSCs. The values presented are the mean ± SEM of three independent experiments. (C): Western blot analysis of cyclin D1, cyclin E1, and cyclin A in SS-AF-MSCs expressing miR-21 or EF1. Quantification was performed using Quantity One software, and the results were normalized to the β-actin positive control and then to the EF1-transduced SS-AF-MSCs. The values presented are the mean ± SEM of three independent experiments (*, p < .05; Student’s t test). (D): Representative images of SS-AF-MSCs expressing EF1 (Di) or miR-21 (Dii) assayed for SA β-gal activity. Magnification, ×20. Graph shows relative levels of SA β-gal-positive cells of three independent experiments (*, p < .05; **, p < .01; Student’s t test). Abbreviations: AF, amniotic fluid; CFU-F, fibroblast colony-forming unit; EF1, elongation factor 1; FACS, fluorescence-activated cell sorting; MSCs, mesenchymal stem cells; NGFR, nerve growth factor receptor; PCR, polymerase chain reaction; PE, phycoerythrin; SA β-gal, senescence-associated β-galactosidase; SS, spindle-shaped.

Inhibition of miR-21 Increases the Proliferation Rate of RS-AF-MSCs and BM-MSCs Through Upregulation of Sox2

As described above, RS-AF-MSCs express miR-21 at high levels. Thus, we transfected these cells with a miR-21 antagonist to explore a possible transition to the SS-AF-MSC phenotype. After transfection with the antagonist, miR-21 expression was decreased in RS-AF-MSCs compared with cells transfected with a scrambled sequence of the miR-21 antagonist (Fig. 6Ai). The proliferation rate of the RS-AF-MSCs was increased 24 hours after miR-21 inhibition, as determined by the MTS assay (Fig. 6Aii). In addition, reduction of miR-21 expression led to a significant increase in the expression levels of the pluripotency genes Sox2, Oct4, and Nanog (Fig. 6Aiii). However, no morphological alterations were observed in RS-AF-MSCs. To test whether Sox2 expression difference was mediated through miR-21 directly, we performed rescue experiments. RS-AF-MSCs were treated with miR-21 antagonist followed by transfection with Sox2 shRNA. As expected, transfection with sh-Sox2 reduced the increase of Sox2 resulted from miR-21 antagonist treatment (Fig. 6Aiv).

Figure 6.

Inhibition of miR-21 in RS-AF-MSCs and BM-MSCs. (Ai): Real-time PCR analysis of miR-21 in RS-AF-MSCs transfected with miR-21 antagonist or the scrambled sequence of the miR-21 antagonist. The results were normalized to the RNU44 internal control and then to the RS-AF-MSC expression levels. (Aii): Comparative analysis of the growth rates of RS-AF-MSCs transfected with miR-21 antagonist or the scrambled sequence 24 and 48 hours after transfection. (Aiii): Western blot analysis of Sox2, Oct4, and Nanog in RS-AF-MSCs transfected with miR-21 antagonist or the scrambled sequence. (Aiv): Western blot analysis of Sox2 in RS-AF-MSCs that were transfected with scramble or miR-21 antagonist and subsequently treated with pLKO.1-sh-Sox2 or pLKO.1-sh-control for 32 hours. (Bi): Real-time PCR analysis of miR-21 in BM-MSCs transfected with the miR-21 antagonist or the scrambled sequence. The results were normalized to the RNU44 internal control and then to the BM-MSC expression level. (Bii): Comparative analysis of the growth rates of BM-MSCs transfected with the miR-21 antagonist or the scrambled sequence 24 and 48 hours after transfection. (Biii): Western blot analysis of Sox2, Oct4, and Nanog in BM-MSCs transfected with the miR-21 antagonist or the scrambled sequence. (Biv): Western blot analysis of Sox2 in BM-MSCs that were transfected with scramble or miR-21 antagonist and subsequently treated with pLKO.1-sh-Sox2 or pLKO.1-sh-control for 32 hours. Western blot quantification was performed using Quantity One software, and the results were normalized to the β-actin positive control and then to the RS-AF-MSCs or BM-MSCs, as appropriate. The values presented are the mean ± SEM of three independent experiments (*, p < .05; **, p < .01; ***, p < .001; Student’s t test). Abbreviations: AF, amniotic fluid; BM, bone marrow; MSCs, mesenchymal stem cells; PCR, polymerase chain reaction; RS, round-shaped.

BM-MSCs exhibit reduced proliferation during multiple passages [31], similar to RS-AF-MSCs, and both cell types express miR-21 at high levels. BM-MSCs (passage 6) expressing miR-21 at high levels (supplemental online Fig. 3B) were also transfected with the miR-21 antagonist or the scrambled sequence. Downregulation of miR-21 expression was observed (Fig. 6Bi), and the proliferation rate of the BM-MSCs was increased 24 hours after miR-21 inhibition, as shown by MTS assay (Fig. 6Bii). Increased expression levels of the pluripotency genes Sox2, Oct4, and Nanog were also detected (Fig. 6Biii). The rescue experiment was also performed for BM-MSCs. Likewise, transfection with sh-Sox2 reduced the increase of Sox2 that resulted from miR-21 antagonist treatment (Fig. 6Biv).

miR-21 Induction Results in Altered Differentiation Properties of SS-AF-MSCs

To evaluate the role of miR-21 in the differentiation potential of SS-AF-MSCs, adipogenic, osteogenic, and chondrogenic induction studies were performed. Transduced SS-AF-MSCs expressing miR-21 or EF1 were cultured under adipogenic conditions for 3, 5, 7, 9, 14, and 21 days. Fewer oil droplets were observed in SS-AF-MSCs expressing miR-21 compared with EF1-transduced cells (Fig. 7Ai). Consistently, the relative expression of adipogenic differentiation markers such as peroxisome proliferator activated receptor-γ (PPARγ) was decreased in SS-AF-MSCs expressing miR-21 compared with EF1-transduced cells during differentiation, as determined by real-time PCR analysis (Fig. 7Aiii).

Figure 7.

miR-21 induction alters the differentiation potential of SS-AF-MSCs. (Ai): Oil Red O staining for adipogenic differentiation in transduced SS-AF-MSCs expressing miR-21 or EF1. Magnification, ×20. (Aii): Quantitative analysis of Oil Red O staining in SS-AF-MSCs expressing miR-21 and EF1-transduced cells at different time points. (Aiii): Real-time PCR expression analysis of PPARγ in SS-AF-MSCs expressing miR-21 compared with EF1-transduced cells. (Bi): Alizarin red S staining for osteogenic differentiation of SS-AF-MSCs expressing miR-21 and of EF1-transduced cells. Magnification, ×20. (Bii): Quantitative analysis of Alizarin red S staining in SS-AF-MSCs expressing miR-21 and EF1-transduced cells. (Biii): Real-time PCR expression analysis of Runx2 and osteocalcin in SS-AF-MSCs expressing miR-21 compared with EF1-transduced cells. (Ci): Pellets formed by SS-AF-MSCs expressing miR-21 and EF1-transduced cells. (Cii): Alcian Blue and collagen IV staining for chondrogenic differentiation in SS-AF-MSCs expressing miR-21 and EF1-transduced cells. Magnification, ×10. (Ciii): Real-time PCR expression analysis of aggrecan and Sox9 in SS-AF-MSCs expressing miR-21 compared with EF1-transduced cells. The results of the real-time PCR analysis were normalized to the human β-actin-positive control and then to the EF1-transduced SS-AF-MSCs. The values presented are the mean ± SEM of three independent experiments (*, p < .05; **, p < .01; Student’s t test). Abbreviations: AF, amniotic fluid; EF1, elongation factor 1; MSCs, mesenchymal stem cells; PCR, polymerase chain reaction; PPARγ, peroxisome proliferator-activated receptor-γ; SS, spindle-shaped.

After osteogenic induction, transduced SS-AF-MSCs expressing miR-21 or EF1 were stained for calcium deposits using Alizarin red S reagent (Fig. 7Bi). In contrast to adipogenesis, osteogenic differentiation was increased by miR-21 expression compared with EF1-transduced cells (Fig. 7Bii). The expression of osteogenic molecular markers such as osteocalcin and Runx2 was increased in SS-AF-MSCs expressing miR-21 compared with EF1-transduced cells (Fig. 7Biii).

After chondrogenic induction, the SS-AF-MSCs aggregated and formed pellets (Fig. 7Ci). For histological evaluation, the pellets were cut into serial 3-μm sections and stained with Alcian Blue (Fig. 7C). The pellets formed by the SS-AF-MSCs expressing miR-21 were smaller than those formed by EF1-transduced cells. Chondrogenic differentiation was also detected by immunofluorescence staining for collagen IV to confirm the production of collagen, which plays a role in the formation of a dense mass through strong interactions (Fig. 7Cii) [32]. In addition, the expression levels of the chondrogenic markers Sox9 and aggrecan were decreased after miR-21 induction in SS-AF-MSCs (Fig. 7Ciii). These results demonstrate that miR-21 induction in SS-AF-MSCs not only reduces the cell proliferation rate but also accelerates osteogenesis and decelerates adipogenesis and chondrogenesis.

To analyze the role of Sox2 inhibition in SS-AF-MSC differentiation properties, adipogenic, osteogenic, and chondrogenic induction studies were also performed. More oil droplets were observed in SS-AF-MSCs transduced with sh-Sox2 virus compared with SS-AF-MSCs transduced with sh-control virus, resulting in an increase in adipogenic differentiation determined by Oil Red O staining (supplemental online Fig. 4Ai, 4Aii). Consistently, the relative expression of adipogenic differentiation marker PPARγ was increased in SS-AF-MSCs transduced with sh-Sox2 virus compared with SS-AF-MSCs transduced with sh-control virus during differentiation, as determined by real-time PCR analysis (supplemental online Fig. 4Aiii). Osteogenic differentiation was decreased when Sox2 was inhibited in SS-AF-MSCs (supplemental online Fig. 4Bi, 4Bii). In agreement, the expression of osteogenic molecular markers such as osteocalcin was decreased in SS-AF-MSCs transduced with sh-Sox2 virus compared with SS-AF-MSCs transduced with sh-control virus (supplemental online Fig. 4Biii). Additionally, after chondrogenic induction, the pellets formed after Sox2 inhibition in SS-AF-MSCs were larger than those formed by SS-AF-MSCs transduced with sh-control virus (supplemental online Fig. 4Ci, 4Cii). Expression levels of the chondrogenic markers Sox9 and aggrecan were also increased after Sox2 inhibition in SS-AF-MSCs (supplemental online Fig. 4Ciii). Thus, it is evident that Sox2 and miR-21 act through independent mechanisms and alter the differentiation properties of SS-AF-MSCs.

Discussion

MicroRNAs and their role in the determination of stem cell fate have recently attracted intense interest. Many studies have already been conducted in MSCs in an attempt to define the miRNAs that regulate the unique properties of MSCs [33–35]. Recent interest has been focused on different MSC populations from fetal and adult sources [10, 36, 37]. The best characterized cells to date are adult BM-MSCs, but recently, fetal MSCs such as AF-MSCs and UCB-MSCs have also attracted increased attention [4, 5, 38–40]. Previously, we successfully isolated and characterized karyotypically normal subpopulations of AF-MSCs and performed a systematic phenotypic, molecular, and proteomic analysis [5, 6]. Interestingly, our studies have shown that SS-AF-MSCs may represent a valuable tool in cell therapy because they are able to induce liver repair [6] or serve as delivery vehicles for anticancer agents in vivo [2].

In this study, we defined the miRNA profiles of MSCs derived from fetal sources such as amniotic fluid and umbilical cord blood and directly compared them with the profile of BM-MSCs. Specifically, 67 miRNAs were differentially expressed in the three MSC sources. On the basis of the detailed analysis, no common miRNAs were detected as upregulated in all the populations, but miR-143, miR-487, miR-326, and miR-199* were downregulated in all the sample categories tested. These miRNAs have been shown to target genes that regulate various cellular processes that are critical for stem cells, such as proliferation, differentiation, and cell survival. For example, miR-143 has been found to target genes such as ERK5 (extracellular-signal-regulated kinase 5) [41], DNMT3A (DNA methyltransferase 3A) [42], FNDC3b (fibronectin type III domain containing 3B) [43], and Bcl-2 (B-cell lymphoma 2) [44], as indicated by the miRecords database [26]. To further study the source of AF-MSCs, a more detailed analysis was performed for the two subpopulations of AF-MSCs, which resulted in the identification of 32 differentially expressed miRNAs between SS-AF-MSCs and RS-AF-MSCs. These two populations exhibited differences in morphology, proliferation rate and differentiation properties, suggesting that the differential expression of miRNAs may regulate these functions. One of the most interesting upregulated miRNAs in RS-AF-MSCs compared with SS-AF-MSCs was miR-21. The role of miR-21 has been studied in various fields, including development, oncology, stem cell biology, and aging [45]. Interestingly, miR-21 has an oncogenic role in several cancers and is thought to be a potential disease biomarker [46, 47]. Furthermore, miR-21 has been implicated in the promotion of tumor growth by targeting genes such as the tumor suppressor Pdcd4 (programmed cell death 4), PTEN (phosphatase and tensin homolog), or TPM1 (tropomyosin 1) in several tumor types [46, 48, 49]. In addition to its oncogenic effect, miR-21 exhibits an antiangiogenic function by targeting the RhoB transcript in endothelial cells [50]. In mouse embryonic stem cells, RE1 silencing transcription factor is responsible for the transcriptional inhibition of miR-21, resulting in the suppression of self-renewal and pluripotency as a result of the loss of expression of the critical regulators Oct4, Sox2, Nanog, and c-Myc [16].

After extensive comparative bioinformatics analysis using GOmir software, Sox2 was identified as a predicted target of miR-21. In this study, we proved for the first time that Sox2 is negatively regulated by miR-21 at the post-transcriptional level by means of a specific target site within the 3′UTR. miR-21 induction in SS-AF-MSCs led to a significant reduction in Sox2 expression at the RNA and protein levels, further supporting a direct interaction. Ours is also the first study to demonstrate that miR-21 may reduce Oct4 and Nanog expression, likely through an indirect mechanism, because miR-21 is not believed to target any of these transcription factors, according to in silico analysis. These three transcription factors are known to act in concert as the core components of a network that promotes pluripotency in stem cells [29]. In addition, Sox2 have been found to directly regulate Nanog expression [51]; thus, it is expected that lower levels of Sox2 would result in lower levels of Nanog expression [51]. The decrease in the expression levels of Sox2, Oct4, and Nanog after miR-21 induction in SS-AF-MSCs was followed by a decrease in the clonogenic potential and proliferation rate of these cells, but there was no evidence for an apoptotic effect. Similarly, recent studies demonstrated that miR-21 regulates the proliferation rate of adipose tissue-derived MSCs (AT-MSCs) and also that miR-21 overexpression decreased clonogenic potential cell proliferation in white adipose tissue in a mouse model with induced obesity from a high-fat diet [52]. More importantly, Sox2 downregulation resulted in a statistically significant reduction of the clonogenic potential and the proliferation rate of SS-AF-MSCs, similar to the phenotype observed after miR-21 induction. These results indicated that miR-21 acts as a regulator of clonogenic potential and proliferation of SS-AF-MSCs by targeting Sox2. In addition, G₀/G₁ cell cycle arrest was observed after miR-21 induction and sequential Sox2 suppression in SS-AF-MSCs. This effect was mainly demonstrated although there were decreased levels of cyclin D1, cyclin E1, and cyclin A. In agreement with our data, Sox2 has previously been reported to promote proliferation by facilitating the G₁/S transition and by transcriptional regulation of the CCND1 gene [53]. More importantly, in human MSCs, sh-Sox2 treatment resulted in a decrease in the proportion of cells in S phase and thus in a reduced proliferation rate [54]. We next analyzed the expression levels of the cell cycle inhibitors p27 and p18, which bind to cyclin-CDK complexes and inhibit their catalytic activity, resulting in suppression of G₁ phase [55, 56]. As expected, p27 and p18 were upregulated after miR-21 induction in SS-AF-MSCs. Additionally, recent studies have shown that miR-21 negatively regulates the cell division cycle 25A protein (Cdc25A), resulting in cell cycle arrest in colon cancer cells [57]. Similarly, in this study, SS-AF-MSCs expressing miR-21 exhibited decreased expression of Cdc25A compared with EF1-transduced cells. This may explain the cell cycle arrest observed after miR-21 induction, as Cdc25A positively regulates G₁/S transition [57]. In addition, we demonstrated relatively high SA β-gal activity of SS-AF-MSCs expressing miR-21 compared with EF1-transduced cells, which may be the result of cell cycle arrest [58, 59].

To further study the functional role of miR-21 and explain the robust expression of this miRNA in RS-AF-MSCs, we performed miR-21 inhibition experiments using a specific antagonist. The suppression of miR-21 in RS-AF-MSCs rescued the low proliferation rate and increased the expression levels of Sox2, Oct4, and Nanog. Although miR-21 appears to regulate the proliferation rate of the two AF-MSC subpopulations, no alterations in RS-AF-MSC morphology were observed after miR-21 inhibition. Interestingly, the inhibition of miR-21 in BM-MSCs had similar effects. Notably, the proliferation rate of BM-MSCs was lower than that of SS-AF-MSCs, as described in our previous studies [5]. In agreement, rescue experiments in RS-AF-MSCs and BM-MSCs with sh-Sox2 reduced the increased expression of Sox2, which initially resulted from the miR-21 antagonist.

In this study, the induction miR-21 resulted in enhanced osteogenesis and inhibited adipogenesis and chondrogenesis in SS-AF-MSCs. However, Sox2 inhibition resulted in opposite effects, inducing adipogenesis and chondrogenesis and reducing osteogenesis. These observations suggest that a different mechanism has been activated during the differentiation process in sh-Sox2-transduced SS-AF-MSCs that is likely independent of the miR-21 pathway.

Conclusion

Although many studies have indicated that miR-21 is an oncomir that regulates proliferation and apoptosis, we observed a contrasting role in AF-MSCs. Our data suggest that miR-21 may act as a regulator of the clonogenic potential, proliferation rate, and differentiation properties of AF-MSCs, most likely by means of the direct suppression of the transcription factor Sox2. The effect of miR-21 may also be responsible for the impaired proliferation rate of MSCs derived from other sources.

Acknowledgments

This research has been cofinanced by the European Union (European Social Fund) and Greek national funds through the Operational Program “Education and Lifelong Learning” of the National Strategic Reference Framework Research Funding Program: Heracleitus II, Investing in Knowledge Society through the European Social Fund. We are grateful to Dr. Helen A. Papadaki (Medical School, University of Crete) and Dr. A. Stravoropoulos (Cord Blood Bank, BRFAA) for providing BM-MSCs and umbilical cord blood samples, respectively. We also thank Prof. L. Naldini, Dr. Panagiotis Politis, and Dr. Elena Siapati for providing plasmids and reagents, and for discussions, and we thank Dr. Ariana Gavriil for technical assistance with FACS.

Author Contributions

O.T.: experimental design, experimental procedures, data analysis, manuscript writing; D.Z.: experimental procedures, data analysis; V.B.: experimental procedures, data analysis, review of manuscript; K.I.P. and A.A.: provision of amniotic fluid samples; N.P.A.: conception, financial support, review of manuscript; M.G.R.: conception and design, financial support, experimental procedures, data analysis, manuscript writing.

Disclosure of Potential Conflicts of Interest

The authors indicate no potential conflicts of interest.