Author contributions: J.H. and P.M.: collection and/or assembly of data, data analysis and interpretation, and manuscript writing; P.L.: data analysis and interpretation and manuscript writing; D.W. and S.L.: statistical analysis of microarray data; S.T.A.: conception and design, data analysis and interpretation, manuscript writing, and final approval of manuscript.
Disclosure of potential conflicts of interest is found at the end of this article.
First published online in STEM CELLSEXPRESS September 4, 2012.
Although the therapeutic potential of mesenchymal stem cells (MSCs) is widely accepted, loss of cell function due to donor aging or culture senescence are major limiting factors hampering their clinical application. Our laboratory recently showed that MSCs originating from older donors suffer from limited proliferative capacity and significantly reduced myogenic differentiation potential. This is a major concern, as the patients most likely to suffer from cardiovascular disease are elderly. Here we tested the hypothesis that a single pluripotency-associated transcription factor, namely Nanog, may reverse the proliferation and differentiation potential of bone marrow-derived MSC (BM-MSC) from adult donors. Microarray analysis showed that adult (a)BM-MSC expressing Nanog clustered close to Nanog-expressing neonatal cells. Nanog markedly upregulated genes involved in cell cycle, DNA replication, and DNA damage repair and enhanced the proliferation rate and clonogenic capacity of aBM-MSC. Notably, Nanog reversed the myogenic differentiation potential and restored the contractile function of aBM-MSC to a similar level as that of neonatal (n)BM-MSC. The effect of Nanog on contractility was mediated—at least in part—through activation of the TGF-β pathway by diffusible factors secreted in the conditioned medium of Nanog-expressing BM-MSC. Overall, our results suggest that Nanog may be used to overcome the effects of organismal aging on aBM-MSC, thereby increasing the potential of MSC from aged donors for cellular therapy and tissue regeneration. STEM CELLS 2012;30:2746–2759
Mesenchymal stem cells (MSCs) provide a promising cell source for tissue regeneration that is already under investigation in several clinical trials [1–3]. However, previous studies showed that MSC suffer from several drawbacks hampering clinical applications including: (a) decreased number and quality of cells with donor age [4–6] and (b) loss of proliferation and differentiation potential upon expansion in vitro [7–9]. Previously, we reported our findings on the effects of donor aging on bone marrow-derived smooth muscle cells (BM-SMCs) originating from ovine BM-MSCs . Using neonatal (n)BM-MSCs and adult (a)BM-MSCs, we found that aging affected not only proliferation and clonogenic potential but also the myogenic differentiation and contractile properties of aBM-MSC significantly. In addition, culture senescence limits the culture time of aBM-MSC to approximately 8–10 passages, thereby preventing their expansion to the numbers required for cellular therapies. These observations pose a significant challenge that must be overcome in order to enable cellular therapies for older patients, the population mostly in need for tissue replacement.
Nanog is a homeodomain transcription factor that inhibits differentiation and maintains pluripotency of embryonic stem cells (ESCs) along with Oct4 and Sox2 [11, 12]. Nanog is expressed in ESCs, embryonic germ, and pluripotent cells from preimplantation and early postimplantation embryos and disappears rapidly upon differentiation . Nanog overexpression enabled ESCs to maintain pluripotency independent of feeder cells  or leukemia inhibitory factor (LIF) , while downregulation of Nanog induced ESC differentiation toward extraembryonic lineages . While some studies reported expression of pluripotency-related factors in MSC, the literature lacks consensus on this issue and the function of these factors in the context of adult stem cells has not been determined [17–20].
Ectopic expression of Nanog was shown to accelerate the growth of somatic cells such as NIH3T3 [21, 22] and BM-MSCs [23, 24]. The effects, however, of Nanog on differentiation potential varied depending on the cell type and the differentiation lineage. Forced expression of Nanog had no effect on terminal differentiation of myogenic progenitors into muscle fibers, but significantly impaired their transdifferentiation into the osteogenic lineage . However, combined expression of Nanog with Oct4 inhibited terminal differentiation of myoblast progenitors . Conversely, ectopic expression of Nanog in human BM-MSC enhanced the differentiation potential along the chondrogenic and osteogenic lineages but abated adipogenic differentiation [23, 24].
In this study, we examined the hypothesis that ectopic expression of Nanog might restore the function of BM-MSC from older donors. Using BM-MSCs from neonatal and adult donors, we discovered that Nanog reversed the aging-mediated loss of proliferation and myogenic differentiation potential of aBM-MSC to a similar level as those of nBM-MSC. In addition, we identified a signaling pathway that mediated—at least in part—the effects of Nanog on the contractile function of BM-MSC that were coaxed to differentiate along the myogenic lineage.
MATERIALS AND METHODS
Cultivation of BM-MSCs and Lentivirus Transduction
Pools of three neonatal (nBM-MSC) and three adult (aBM-MSC) MSCs were prepared. Isolation of individual cell derived from BM of three neonatal lambs (<3 days old) and three adult sheep (4–4.5 years old) were described previously . Isolation of human hair follicle-derived-MSCs (HF-MSCs) was performed as described previously . Briefly, human scalp of a 73-year-old male containing hair follicles was obtained from the Cooperative Human Tissue Network (Philadelphia, PA, http://www.chtn.nci.nih.gov). The sample was washed extensively and subsequently cut into small pieces. To dissociate hair follicles from the surrounding matrix, the samples were treated with 0.5% collagenase type I (Invitrogen, Carlsbad, CA, http://www.invitrogen.com) for 4 hours at 37°C. Next, single hair follicles were unplugged from the dermis and plated each in a well of a 24-well plate. Hair follicles were cultured in Dulbecco's modified Eagle's medium (DMEM; Gibco, Grand Island, NY, http://www.invitrogen.com) supplemented with 10% fetal bovine serum (FBS; Gibco). HF-MSCs migrated out of the follicle and were further expanded. Human adult BM-MSCs from two donors (22 years old female and 29 years old male) were obtained from Stem Cell Technologies (Vancouver, Canada, http://www.stemcell.com). MSCs were cultured in DMEM containing 10% FBS supplemented with 2 ng/ml basic fibroblast growth factor (bFGF; BD Biosciences, San Jose, CA, http://www.bdbiosciences.com). Nanog was expressed from a lentiviral vector, pSIN-EF2-Nanog-Puro (Addgene, Cambridge, MA, http://www.addgene.org), under the control of EF1α promoter. The internal ribosome entry site was placed right after Nanog to allow expression of puromycin phosphotransferase under the same promoter. For lentivirus production, 293T/17 cells (ATCC, Manassas, VA, http://www.atcc.org) were cotransfected with the following three plasmids as described recently : Nanog lentiviral vector, psPAX2-gag/pol/tat/rev, and pMD2.G-VSVG. Neonatal and adult MSCs were transduced with viral supernatant and were selected with 0.5 μg/ml puromycin until all noninfected cells were killed.
Proliferation and Clonogenic Assays
Neonatal and adult BM-MSCs were seeded at 1.0 × 105 cells per well in six-well plates and cultured in DMEM containing 10% FBS plus bFGF (2 ng/ml). After reaching near confluence, the cells were trypsinized, counted, and the doubling time was calculated. Clonogenic assay was previously described in detail . In short, nBM-MSC and aBM-MSC were seeded (500 cells per dish) in a 100 mm culture dish and cultured for 10–15 days. The cells were then fixed and stained with trypan blue. Each dish was photographed at a constant distance using a gel documentation imaging system (UVP, Upland, CA, http://www.uvp.com). Images were analyzed using the NIH software ImageJ (version 1.43u; National Institutes of Health, Bethesda, MD, http://rsb.info.nih.gov/ij) to determine the area and effective diameter of each clone.
The senescence-associated-β-galactosidase (SA-β-gal) activity was detected using β-Gal Staining Kit (Invitrogen) according to the manufacturer's recommendations except that citric acid/sodium phosphate buffered-staining solution (pH 6.0) was used as described before . Cells were photographed using the EVOS phase-contrast microscope (Advanced Microscopy Group, Bothell, WA, http://www.amgmicro.com). SA-β-gal-positive cells were counted in five randomly selected fields of view to determine the percentage of β-gal+ cells (total of >400 cells were counted).
Immunostaining and Western Blots
Immunostaining for alpha smooth muscle actin (αSMA) and Calponin was performed as described previously .
Western blots (WBs) were performed as described previously [31, 32] using antibodies against Nanog (1:1,000 in 5% milk; BD Biosciences), αSMA (1:1,000 in 5% milk; Sigma-Aldrich, St. Louis, MO, http://www.sigmaaldrich.com), Calponin (1:1,000 in 5% milk; Santa Cruz Biotechnology, Santa Cruz, CA, http://www.scbt.com), p-Smad2 (1:1,000 dilution in 5% bovine serum albumin (BSA); Cell Signaling, Danvers, MA, http://www.cellsignal.com), and Smad2 (1:1,000 dilution in 5% BSA; Cell Signaling).
Vascular Contractility of MSCs on Tissue Equivalents
Cylindrical vascular tissue equivalents containing MSC in fibrin hydrogels (106 cells per milliliter) were cultured around a 6.0 mm mandrel of poly(dimethyl siloxane) as described previously [33, 34]. The medium containing DMEM and 10% FBS supplemented with insulin (2 μg/ml; Sigma-Aldrich), TGF-β1 (2 ng/ml, BioLegend, San Diego, CA, http://www.biolegend.com), ascorbic acid (300 μM; Sigma-Aldrich), and ε-amino-n-caproic acid (2 mg/ml; Sigma-Aldrich) was changed every 2 days. After 2 weeks in culture, the tissue constructs were photographed and the wall thickness and length were measured using ImageJ. Then the tissues were released from the mandrel and mounted on two stainless hooks through the lumen while one side was fixed, and the other side was connected to a force transducer. Each constructs resided in an isolated tissue bath filled with Krebs–Ringer solution. The tissues were continuously bubbled with 94% O2, 6% CO2 to obtain a pH of 7.4, a P of 38 mmHg, and a P >500 mmHg at 37°C. Tissues were equilibrated at a basal tension of around 1.0 g and constant length for 30–60 minutes. After equilibration, endothelin-1 (20 nM; Sigma-Aldrich), the thromboxane A2 mimetic U46619 (1 μM; Sigma-Aldrich), or potassium chloride (KCl: 118 mM; Sigma-Aldrich) was added to the tissue bath and isometric contraction was recorded using a PowerLab data acquisition unit and analyzed with Chart5 software (ADInstruments, CO Springs, CO, http://www.adinstruments.com). Contractility was normalized by the tissue area applying force and expressed in kPa.
Compaction of Fibrin Hydrogels
Fibrin gel compaction assay was previously described in detail . Briefly, BM-MSCs were embedded in fibrin gels (106 cells per milliliter) that contained fibrinogen (2.5 mg/ml) and thrombin (2.5 U/ml). One hour after polymerization, gels were released from the plate walls and cells were allowed to compact the gels over time. To quantify gel compaction activity, gels were photographed at the indicated times and their area was measured using ImageJ software.
Total RNA Preparation and GeneChip Arrays
Total RNA was isolated from three replicates per group. Cells were seeded in 100 mm culture dishes under growth condition (DMEM + 10% FBS + 2 ng/ml bFGF). When they reached confluence, total RNA was isolated using the RNeasy kit (Qiagen, Chatsworth, CA, http://www.qiagen.com). The quality of purified RNA was analyzed on an Agilent Bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA, http://www.agilent.com). Intact whole RNA was labeled to obtain biotinylated cRNA for hybridization to Affymetrix GeneChip Bovine Genome Arrays according to the manufacturer's protocol. This array contains 24,027 probe sets representing approximately 23,000 bovine transcripts and the ovine and bovine genomes have a higher than 90% genetic similarity. Scanned microarray images were imported into GeneChip Command Console Software (AGCC, Affymetrix, Santa Clara, CA, http://www.affymetrix.com) to generate raw signal values for each probe.
Affymetrix Bovine GeneChip data files were processed using the MAS5.0 algorithm (implemented in the affy R library of Bioconductor package) to generate expression summary values for each probe set. MAS5.0-based “present calls” was used to keep the probe sets with “present” status across all three samples in at least one of the four groups for downstream analysis. We then performed three separate comparisons based on the following samples characteristics: Nanog+ aBM-MSC (aBM.N) versus control aBM-MSC (aBM.C); Nanog+ nBM-MSC (nBM.N) versus control nBM-MSC (nBM.C); and nBM.C versus aBM.C. The Limma program in the Bioconductor package was used to calculate the level of differential gene expression. Briefly, a linear model was fit to the data with cell means corresponding to the different conditions and a random effect for array. For each comparison, we obtained the list of differentially expressed genes (DEGs) constrained by p-value <.01 and at least twofold change. Following single gene-based significance testing, we used the expression value of DEGs to cluster the samples for each comparison. Hierarchical clustering based on the average linkage of Pearson correlation was used. The list of DEGs was further analyzed for enriched Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway using the NCBI DAVID server with default setting. The statistical significance was calculated using the Fisher's exact test in which the null hypothesis is that no difference exists between the number of genes falling into a given pathway in the target DEG list and the genome as a whole.
Quantitative Real-Time Polymerase Chain Reaction
Total RNA was reverse transcribed using a cDNA synthesis kit (Qiagen) according to the manufacturer's instructions. Quantitative polymerase chain reaction (PCR) was performed using the iCycler (Bio-Rad Laboratories, Hercules, CA, www.bio-rad.com). The reaction was carried out in a volume of 25 μl containing 1 μl of cDNA, 0.4 μM of each primer (Sigma Genosys, Woodlands, TX, https://www.sigma-genosys.com), and 12.5 μl of 2× IQ TM SYBR Green Supermix (Bio-Rad Laboratories). The primer sequences for the genes used in this study were listed in Supporting Information Table 5. Each reaction comprises 40 cycles each with melting at 95°C for 10 seconds, annealing and extension at 55°C for 30 seconds. The fluorescence intensity was recorded during the extension step of each cycle. The specificity of the PCR products was verified using the melting curve generated by MyiQ software and by electrophoresis on 1% agarose gels. The PCR data analysis was performed as described before . Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) served as a loading control.
Pairwise statistical analysis of the data was performed using a two-tailed Student's t test using Microsoft Excel software. The data were considered statistically different when p < .05. Each experiment was repeated at least three times with triplicate samples each time unless indicated otherwise.
Generation of Nanog-Expressing nBM-MSC and aBM-MSC
Previously, we demonstrated that BM-MSC-derived SMC from aged donors showed dramatic loss of proliferative and differentiation potential when compared with their neonatal counterparts . Here, we attempted to overcome the aging-induced loss of cellular function by introducing the ESC pluripotency-related transcription factor, Nanog. We hypothesized that the negative effects of donor aging on BM-MSC proliferation and myogenic differentiation may be reversed by ectopic expression of a single pluripotency factor, Nanog. To test this hypothesis, we generated BM-MSCs overexpressing Nanog protein from neonatal (<3 days old) or adult (4–4.5 years old) ovine cells using lentiviral vectors encoding for human Nanog and puromycin phosphotransferase. Control cells were modified with the same construct lacking the Nanog coding sequence. After puromycin selection expression of Nanog protein in nBM-MSC and aBM-MSC was confirmed by Western blot analysis (Fig. 1A). In addition, immunostaining showed that Nanog was absent from control cells but was highly expressed and localized exclusively in the nucleus of transduced aBM-MSC and nBM-MSC (Fig. 1B, Supporting Information Fig. S1). However, Nanog expression did not induce expression of other pluripotency factors neither at the mRNA (Fig. 1C) or the protein level (data not shown).
Nanog-Induced Changes in the Global Gene Expression Profile aBM-MSC and nBM-MSC
Next we used Affymetrix GeneChip Bovine Genome microarrays to investigate the global molecular effects of Nanog expression on nBM-MSC and aBM-MSC. A total of 12 microarrays representing triplicates of nBM.C, nBM.N, aBM.C, or aBM.N were performed. The data were analyzed using the following criteria: (a) a gene was included in the analysis when it was expressed in at least one of the four groups; (b) a gene was differentially expressed if its expression changed by at least twofold (fold change (FC) ≥2) and the change was statistically significant with p-value less than .01. Using the first criterion, we obtained 11,818 transcripts for downstream analysis (Supporting Information Table S1). Among them, 1,671 transcripts were differentially expressed (p < .01, FC ≥2) in at least one of the following three pairs of conditions: (a) nBM.C versus aBM.C; (b) aBM.N versus aBM.C; and (c) nBM.N versus nBM.C (Supporting Information Table S2). DEGs in the first group are likely to be affected by the donor age, while genes in the second and third groups are likely affected by Nanog expression in adult and neonatal cells, respectively.
Comparison between aBM-MSC and nBM-MSC showed that 5.3% of genes might be affected by organismal aging. Specifically, 623 transcripts were differentially expressed (p < .01, FC ≥2) between neonatal and adult BM-MSC; 243 of them were upregulated and 380 genes were downregulated in nBM-MSC. Nanog expression affected gene expression of aBM-MSC to a larger extent than that of nBM-MSC: 967 genes (8.2%) and 678 genes (5.7%) were affected by Nanog in aBM-MSC and nBM-MSC, respectively. Compared to their corresponding control, 387 genes were upregulated and 580 genes were downregulated in aBM.N, while 283 genes were upregulated and 395 genes were downregulated in nBM.N.
Hierarchical clustering showed that the Nanog-expressing cells, nBM.N and aBM.N, were clustered with each other instead of their control counterparts (Fig. 2A). Interestingly, as shown in the Venn diagrams (Fig. 2B) among 132 common DEGs between the first group (Donor age; nBM.C vs. aBM.C) and the second group (Nanog in adult; aBM.N vs. aBM.C), the vast majority (128 out of 132) changed in the same direction for both groups (35 were upregulated and 93 downregulated). Conversely, among 107 DEGs shared between the first and the third group (Nanog in neonate; nBM.N vs. nBM.C), only a small number of genes changed in the same direction (two up and 11 downregulated), and the remaining genes were regulated in opposite directions (Supporting Information Table S2). These results suggest that Nanog-induced gene expression changes might have driven aBM-MSC closer to nBM-MSC, thereby supporting the hypothesis that Nanog might have reversed the effects of aging in aBM-MSC.
To identify direct targets of Nanog in BM-MSC, we compared the DEGs from our analysis with the list of 1,687 genes whose promoters were found to contain Nanog binding sites in hESC . Our analysis identified 75 genes in aBM-MSC and 56 genes in nBM-MSC that might be directly affected by Nanog (Supporting Information Table S3). The small number of genes that might be direct targets of Nanog may be caused by the limited accessibility of promoters in somatic versus ESCs or potential differences in the regulatory targets of Nanog in ovine versus human cells.
We also performed pathway analysis (Fig. 2C) to identify pathways that were significantly enriched with DEGs in each of the three sets that were compared (nBM.C vs. aBM.C, aBM.N vs. aBM.C, and nBM.N vs. nBM.C). As shown in Supporting Information Table S4, a few pathways were found to be common between these groups. For example, the chemokine-signaling pathway was significantly changed in nBM-MSC as well as in aBM-MSC upon Nanog expression. Similarly, the PPAR signaling pathway was significantly changed in both nBM-MSC and aBM-MSC upon Nanog expression, in agreement with a previous study, which reported that Nanog inhibited the adipogenic potential of hBM-MSC .
Conversely, Nanog affected some pathways differentially in aBM-MSC compared to nBM-MSC. In particular, Nanog induced changes in cell cycle, DNA synthesis/replication, and nucleotide mismatch/repair pathways in aBM-MSC and to a lesser extent in nBM-MSC. Only a small subset (4 out of 26) of cell cycle-related genes were upregulated by Nanog in neonatal cells (nBM.N vs. nBM.C), possibly because nBM-MSCs were proliferating fast even in the absence of Nanog. Instead, in nBM-MSC, Nanog upregulated several genes in the TGF-β and cancer signaling pathways including TGFB1, TGFB2, TGFB3, FGF2, FGF10, VEGFC, HGF, and BMP2.
Validation of Microarray Results with Quantitative Reverse Transcription-PCR
Next we performed quantitative reverse transcription-PCR (qRT-PCR) to validate the microarray data for a selected set of genes. To this end, we selected DEGs from cell cycle pathways including CCNA2, CDC2, CCNE2, CCNB1, CCNB2, CDKN1C, PCNA, and TP53. The expression level of each gene was normalized to GAPDH and the ratio of aBM.N to aBM.C was compared to the fold change obtained from microarray data. In all cases, qRT-PCR analysis showed similar trend as well as quantitatively similar results with microarray analysis (Fig. 2D).
Nanog Increased Proliferation in Neonatal and Adult BM-MSC
Upregulation of cell cycle-related genes suggested that Nanog might enhance the proliferation potential of aBM-MSC. Indeed, experiments measuring the clonogenic potential and cell proliferation rate confirmed this hypothesis. To measure clonogenicity, cells were seeded at clonal density (500 cells per 100 mm dish) and cultured for 9 days (nBM-MSC) or 18 days (aBM-MSC). Only colonies with diameter larger than 2 mm were counted, as these were more likely to have originated from highly proliferative stem cells. Although control neonatal cells showed high colony forming potential, the number of colonies larger than 2 mm in diameter was significantly higher in Nanog-expressing cells (Fig. 3A, 3B). In addition, proliferation assays showed that the doubling time of Nanog-expressing cells was approximately 12-hour shorter than that of control cells (Fig. 3C). Interestingly, Nanog had a more profound effect on aBM-MSC. Specifically, the Nanog-expressing aBM-MSCs formed four times more colonies than their control counterparts and the cell density was significantly higher in Nanog+ colonies (Fig. 3A, 3B). Nanog expression also decreased the doubling time of aBM-MSCs from 70 hours to approximately 50 hours (Fig. 3C).
Notably, serial passaging showed that expression of Nanog delayed culture senescence significantly. Specifically, aBM-MSC stopped dividing after 14 population doublings (estimated after passage 2) while Nanog-expressing aBM-MSC maintained their proliferation potential up to 39 population doublings. Indeed, while more than 70% of passage 8 control aBM-MSC stained positive for SA-β-gal, only 12% of Nanog-expressing cells was SA-β-gal positive even at passage 20 (Fig. 3D, 3E). In addition, a gene that is strongly associated with cellular senescence, the cyclin-dependent kinase inhibitor 2A or p16INK4a, was highly upregulated (> sevenfold) in aBM-MSC but decreased significantly in Nanog-expressing cells (> 30-fold) (Fig. 3F).
Consistently, the p16INK4a activator mesenchyme homeobox 2 (MEOX2) was significantly decreased in Nanog-expressing aBM-MSC and nBM-MSC (Supporting Information Fig. S2A), while the p16INK4a inhibitor EZH2 was significantly upregulated by Nanog in aBM-MSC (Supporting Information Fig. S2B). Finally, increased proliferation was not accompanied by chromosomal abnormalities as demonstrated by karyotype analysis showing that both Nanog-expressing nBM-MSC and aBM-MSC possessed 54 diploid chromosomes as expected for normal ovine cells with no signs of transformation (Supporting Information Fig. S3).
Nanog Improved Myogenic Differentiation of aBM-MSC and nBM-MSC
When expressed in MSC, Nanog was found to promote differentiation toward osteogenic and chondrogenic lineages but prevented adipogenesis [23–26]. However, the effect of Nanog on myogenic differentiation and especially in the context of organismal aging has not been investigated. To address this question, we first examined the expression of SMC-specific markers such as αSMA and calponin after induction of myogenic differentiation for 5 days. Immunostaining showed that both nBM-MSC and aBM-MSC increased expression of both proteins, which also assumed fibrillar organization indicative of contractile SMC (Fig. 4A, 4B).
Next, we compared the expression level of αSMA and calponin between control and Nanog+ BM-MSCs under myogenic differentiation conditions using WB. As shown in Figure 4C, Nanog did not alter the protein level of αSMA or calponin in neonatal or adult BM-MSCs. We also performed real-time RT-PCR for other smooth muscle contractility-related genes including SM22, smoothelin, and caldesmon for which ovine-specific antibodies were not available (Fig. 4D). Although the expression levels of SM22 and caldesmon were similar between Nanog-expressing and control nBM-MSCs, SM22 increased slightly but significantly and caldesmon increased by 7.5-fold in Nanog-expressing aBM-MSCs as compared to their control counterparts. Conversely, smoothelin expression was significantly increased by Nanog expression in both neonatal and adult BM-MSCs. These results prompted us to examine the ability of Nanog-expressing cells to generate force by measuring the compaction of three-dimensional fibrin hydrogels.
To this end, nBM-MSCs and aBM-MSCs were embedded in fibrin gels (106 cells per milliliter), and 1 hour after polymerization, the gels were released from the plate wall and allowed to compact in the presence of TGF-β1 (2 ng/ml). At the indicated times, the area of each gel was measured using ImageJ and normalized to its initial area. As shown in Figure 5, control nBM-MSC showed significant contractility but the rate as well as the final extent of gel compaction was higher in Nanog-expressing nBM-MSC (Fig. 5A, 5B). Conversely, aBM-MSC showed limited gel compaction activity but similar to nBM-MSC, Nanog enhanced the force generation ability of aBM-MSC significantly. Specifically, Nanog-expressing aBM-MSC compacted gels to half of their original gel size within 50 hours when compared with 70 hours for control cells (Fig. 5A, 5B). These data suggested that Nanog enhanced the contractility of nBM-MSC and aBM-MSC that were coaxed to differentiate to SMC.
Nanog Restored Age-Related Loss of Vascular Contractility of aBM-MSC
Next, we evaluated the effect of Nanog on vasoreactivity of vascular constructs prepared from nBM-MSC or aBM-MSC. To this end, we prepared cylindrical tissue equivalents by embedding cells in fibrin hydrogels that were polymerized around cylindrical mandrels. After 2 weeks in culture in myogenic differentiation medium, the cells compacted the hydrogels down to ∼ 5% of their original volume yielding cylindrical constructs with wall thickness of less than 500 μm. At that time, vascular rings were placed in isolated tissue baths to measure isometric tension generated in response to receptor or non-receptor-mediated vasoagonists.
Although nBM-MSC-based constructs showed high vasoreactivity in response to receptor (Endothelin-1, U46619) or non-receptor (KCl)-mediated pathways, Nanog expression increased contractility significantly (Fig. 6A, 6B). Conversely, constructs generated with aBM-MSC showed very low level of vascular reactivity. Notably, Nanog enhanced vascular contractility of aBM-MSC by more than 10-fold restoring it to similar level as that of Nanog+ nBM-MSC and higher than control nBM-MSC. There was no statistically significant difference in the response of Nanog+ aBM-MSC (n = 10) versus Nanog+ nBM-MSC (n = 11) to any of the three agonists (p = .8, .65, and .87 for Endothelin-1, U46619, and KCl, respectively). Collectively, our data show that forced expression of Nanog restored the aging-related loss of vascular function of aBM-MSC to a similar level as that of nBM-MSC.
Nanog Increased Vascular Contractility of Human aMSCs
These results prompted us to hypothesize that Nanog may exert a similar effect on the contractile function of human aMSC. To this end, we used MSCs from three donors (two males and one female) and two different anatomic locations, that is, BM and hair follicle (HF-MSCs). Cylindrical tissue equivalents were prepared with Nanog-expressing or control MSCs and the vascular reactivity was measured in response to the same vasoactive agonists. Our results clearly demonstrate that Nanog enhanced the contractility of all MSCs irrespective of gender, anatomic location, or species (Fig. 6C–6E).
Nanog-Induced Increase in Fibrin Compaction Was Mediated in Part Through the TGF-β/Smad Signaling Pathway
Microarray analysis showed that Nanog upregulated expression of TGF-β1 in nBM-MSC and to a lesser extent in aBM-MSC. Real-time qRT-PCR verified this result and showed that TGF-β1 expression increased by approximately fivefold in nBM-MSC and approximately threefold in aBM-MSC upon Nanog expression (Supporting Information Fig. S4). Since TGF-β1 is known to increase SMC contractility, we examined whether Nanog enhanced contractility through the TGF-β1 pathway. Indeed, Nanog induced Smad2 phosphorylation in both nBM-MSC and aBM-MSC in the absence of exogenous TGF-β1 (Fig. 7A, 7B), without affecting the expression level of total Smad2 (Supporting Information Fig. S5). In addition, Smad2 phosphorylation was diminished by treatment with the TGF-β type I receptor kinase inhibitor, SB431542, implicating the TGF-β pathway in this process.
Furthermore, we examined whether increased Smad2 phosphorylation and contractility might be induced by soluble factors secreted by Nanog-expressing BM-MSC. To this end, we measured Smad2 phosphorylation in control aBM-MSC or nBM-MSC in the presence of medium that was conditioned (conditioned medium, CM) by Nanog+ aBM-MSC (aBM.N-CM) or Nanog+ nBM-MSC (aBM.N-CM). As control, we used CM from control nBM-MSC (nBM.C-CM) or aBM-MSC (aBM.C-CM), respectively. As shown in Figure 7C, both aBM.N-CM and nBM.N-CM induced Smad2 phosphorylation, which was completely abolished by SB431542 treatment. Interestingly, nBM-MSCs exhibited considerable compaction in the presence of nBM.N-CM and to a lesser extent aBM.N-CM but failed to contract fibrin gels in the presence of nBM.C-CM. Similar to Smad2 phosphorylation, the enhanced contractility was diminished by treatment with SB431542. In addition, a TGF-β1 neutralizing antibody inhibited nBM.N-CM-induced pSmad2 and hydrogel compaction partially (Supporting Information Fig. S6). Taken together, these results suggest that Nanog enhanced the myogenic differentiation potential of aBM-MSC and nBM-MSC by inducing expression of TGF-β1 and perhaps also other as yet unidentified factor(s) that activate the TGF-β signaling pathway.
Although the therapeutic potential of MSCs is widely accepted, loss of cell function due to donor aging or culture senescence are major limiting factors hampering their clinical application. In a recent study, we characterized the effects of donor aging on MSCs derived from neonatal or adult BM  and found that the BM-MSC proliferation and myogenic differentiation potential declined significantly with donor aging. The effects of organismal aging are also compounded by culture senescence limiting the culture time of MSC to approximately 8–10 passages, thereby preventing their expansion to the large-cell numbers required for cellular therapies. This is a major concern, as the patients mostly in need for cellular therapies in general and for vascular grafts in particular are elderly. Here, we attempted to address this challenge by ectopic expression of pluripotency-associated factors, Nanog or Oct4, into BM-MSC from neonatal and adult animals.
Some studies showed that MSCs express pluripotency-related factors such as Oct4 and Sox2  or Oct4 and Nanog , but their expression decreased over time in culture. Nanog and Oct4 expression correlated inversely with organismal aging as human first-trimester fetal MSCs, which expressed Nanog and Oct4, had longer telomeres and grew faster as compared to adult BM-MSCs lacking both factors . Other groups did not observe endogenous expression of pluripotency factors in MSC but showed that ectopic expression of Nanog, Oct4, or Sox2 increased the proliferation rate and improved osteogenesis and chondrogenesis. Interestingly, while Oct4 promoted Nanog inhibited adipogenesis [23, 24]. Similarly, our results show that ovine aBM-MSC or nBM-MSC did not express Nanog endogenously but ectopic expression had profound effects on proliferation and myogenic differentiation, especially for aBM-MSC.
Specifically, Nanog expression shortened the doubling time and delayed the senescence of adult MSCs significantly. Previous studies demonstrated that cell cycle and DNA replication/mitosis genes were significantly reduced in senescent or late passage adult cells [37–42]. Interestingly, microarray analysis revealed that cell cycle, DNA replication, and DNA repair-related genes were highly upregulated by Nanog in aBM-MSC but not in nBM-MSC, which were already highly proliferative. Conversely, p16INK4a and the p16INK4a activator MEOX2 (a suppressor of cell proliferation that is involved in senescence ) were significantly decreased in Nanog-expressing aBM-MSC and nBM-MSC. Conversely, the p16INK4a inhibitor EZH2, which was shown to restore proliferation of aged β-cells  and hematopoietic stem cells  was significantly upregulated by Nanog in aBM-MSC.
KEGG pathway analysis revealed that Nanog downregulated adipogenic genes such as CEBPα and PPARγ, in agreement with previous work showing that ectopic expression of Nanog suppressed adipogenic differentiation . Remarkably, several studies reported that aging correlated with decreased MSC capacity for osteogenesis, chondrogenesis, and myogenesis and increased propensity for adipogenesis [10, 46, 47]. Interestingly, the p53 level increased moderately but significantly in aBM-MSC upon Nanog expression. It is generally accepted that increased p53 results in cell cycle arrest through p21 but p53 was also shown to maintain genome stability and prevent oncogenic transformation . Interestingly, the level of p53/p21 was significantly increased during reprogramming leading to reduced efficiency of induced pluripotent stem cell (iPSC) generation [49, 50]. Others reported that p53/p21 expression was considerably decreased in MSCs from old donors  or late passage cells (>30 passages) , possibly indicating an increased propensity for generating cancer stem cells. Therefore, the increase in p53 by Nanog might help to sustain genomic stability and prevent oncogenic transformation of BM-MSC.
Cellular senescence has been attributed to—among other factors—deficiencies in DNA damage repair or genome maintenance [51–53]. DNA damage has been associated with oxidative stress due to endogenous mitochondrial respiration or exogenous stressors such as UV irradiation. Replication-associated DNA damage also leads to replicative exhaustion and decreased proliferative capacity of aged cells . During DNA synthesis, deficiencies in DNA replication machinery, introduction of errors, or slow base excision repair could increase susceptibility of DNA to recombination, eventually resulting in the onset of replication-arrest . Besides cell cycle genes, our microarray analysis identified DNA replication, nucleotide (purine and pyrimidine) metabolism, base excision repair, and mismatch repair-related genes that were upregulated by Nanog in aBM-MSC. These microarray findings suggest that Nanog expression might delay adult stem cell senescence and sustain the proliferation potential of aBM-MSC by facilitating DNA repair and stabilizing genome integrity.
We also observed that Nanog enhanced differentiation of ovine BM-MSC into SMC as demonstrated by the dramatically increased contractility, especially of aBM-MSC. In addition, we observed that Nanog significantly increased the contractile function of human adult MSCs, independent of anatomic location or gender. Increased myogenic function correlated with higher expression of SMC markers such as SM22, smoothelin, and caldesmon. Interestingly, Nanog-expressing cells exhibited sustained phosphorylation of Smad2 in the absence of exogenous TGF-β1, suggesting that the TGF-β signaling might be sustained through autocrine signals. Indeed, CM from Nanog-expressing cells induced Smad2 phosphorylation, which was inhibited by the TGFβ type I receptor inhibitor, SB431542, supporting this hypothesis. However, ELISA measurements showed only a small increase in the level of TGF-β1 protein in CM of Nanog-expressing cells (data not shown). In addition, TGF-β neutralizing antibody inhibited Smad2 phosphorylation and contractility induced by CM from Nanog-expressing cells only partially (Supporting Information Fig. S6), suggesting that in addition to TGF-β1 other TGF-β family ligands might also be responsible for increased myogenic function. Alternatively, part of the TGF-β1 activity might not be in soluble form but might be bound to extracellular matrix, which was recently shown to enhance its bioactivity . Immobilized TGF-β1 might not be detected by ELISA and might not be neutralized by the anti-TGF-β1 antibody, although the TGF-β1 pathway would still be blocked by the TGFβ receptor inhibitor, SB431542. In other experiments, TGF-β1 was required for hydrogel compaction by control but not by Nanog-expressing BM-MSC. Although addition of exogenous TGF-β1 increased Nanog-induced compaction even further, SB431542 inhibited compaction only partially (Supporting Information Fig. S7), suggesting that Nanog-induced contraction might be mediated by TGF-β1-independent pathways as well. More studies are required to uncover these Nanog-activated pathway(s).
Interestingly, another pluripotency-associated transcription factor, Oct4, had different effects on BM-MSC. Although Oct4 also enhanced proliferation and clonogenic capacity of both neonatal and adult BM-MSCs significantly (Supporting Information Fig. S8), it had differential effects on the myogenic differentiation of adult and neonatal cells. Specifically, upon myogenic differentiation, Oct4-expressing aBM-MSC exhibited increased vascular contractility albeit to a much lower extent than Nanog-expressing cells. Conversely, Oct4 significantly decreased contractility of nBM-MSC, suggesting the Oct4 might inhibit differentiation of nBM-MSC into functional SMC. Previous studies showed that Oct4 promoted adipogenesis of human adult BM-MSC  but failed to enhance osteogenesis . Taken together these results may indicate that Oct4 and Nanog affect the transcription program of BM-MSC in different ways, despite recent findings showing that in ESC they co-occupy a substantial portion of their target genes and collaborate in regulating pluripotency and differentiation . Finally, our data may also suggest that the effects of Nanog and Oct4 on BM-MSC depend strongly on donor age, possibly due to differences in target gene accessibility in neonatal versus adult stem cells.
In summary, this study demonstrated that Nanog expression enhanced the proliferation and myogenic differentiation potential of nBM-MSC and to a significantly greater extent aBM-MSC, suggesting that Nanog may reverse or prevent aging-induced loss of aBM-MSC function, with no need for reprogramming into a pluripotent state. This in turn may allow for expansion of aBM-MSC to the large-cell numbers required for regenerative medicine, thereby facilitating development of cellular therapies for the elderly—the patients mostly in need for tissue replacement.
This work was supported by grants from the National Institutes of Health (R01 HL086582) and the New York State Stem Cell Science (NYSTEM Contract #C024316) to S.T.A.
DISCLOSURE OF POTENTIAL C ONFLICTS OF INTEREST
The authors indicate no potential conflicts of interest.