DNA Methylation Plasticity of Human Adipose-Derived Stem Cells in Lineage Commitment

Adult stem cells have an enormous potential for clinical use in regenerative medicine that avoids many of the drawbacks characteristic of embryonic stem cells and induced pluripotent stem cells. In this context, easily obtainable human adipose-derived stem cells offer an interesting option for future strategies in regenerative medicine. However, little is known about their repertoire of differentiation capacities, how closely they resemble the target primary tissues, and the potential safety issues associated with their use. DNA methylation is one of the most widely recognized epigenetic factors involved in cellular identity, prompting us to consider how the analyses of 27,578 CpG sites in the genome of these cells under different conditions reflect their different natural history. We show that human adipose-derived stem cells generate myogenic and osteogenic lineages that share much of the DNA methylation landscape characteristic of primary myocytes and osteocytes. Most important, adult stem cells and in vitro–generated myocytes and osteocytes display a significantly different DNA methylome from that observed in transformed cells from these tissue types, such as rhabdomyosarcoma and osteosarcoma. These results suggest that the plasticity of the DNA methylation patterns plays an important role in lineage commitment of adult stem cells and that it could be used for clinical purposes as a biomarker of efficient and safely differentiated cells. Adult stem cells have an enormous potential for clinical use in regenerative medicine that avoids many of the drawbacks characteristic of embryonic stem cells and induced pluripotent stem cells. In this context, easily obtainable human adipose-derived stem cells offer an interesting option for future strategies in regenerative medicine. However, little is known about their repertoire of differentiation capacities, how closely they resemble the target primary tissues, and the potential safety issues associated with their use. DNA methylation is one of the most widely recognized epigenetic factors involved in cellular identity, prompting us to consider how the analyses of 27,578 CpG sites in the genome of these cells under different conditions reflect their different natural history. We show that human adipose-derived stem cells generate myogenic and osteogenic lineages that share much of the DNA methylation landscape characteristic of primary myocytes and osteocytes. Most important, adult stem cells and in vitro–generated myocytes and osteocytes display a significantly different DNA methylome from that observed in transformed cells from these tissue types, such as rhabdomyosarcoma and osteosarcoma. These results suggest that the plasticity of the DNA methylation patterns plays an important role in lineage commitment of adult stem cells and that it could be used for clinical purposes as a biomarker of efficient and safely differentiated cells. Human adipose-derived stem cells (hASCs) refer to the plastic-adherent and multipotent cell population isolated from collagenase digests of adipose tissue. Although the differentiation capacity of hASCs was initially thought to be limited to their tissue of origin, recent data have demonstrated that multipotent stem cells retain a broad differentiation potential. hASCs can be induced to differentiate along several mesenchymal tissue lineages, including adipocytes, osteoblasts, myocytes, and chondrocytes.1Zuk P.A. Zhu M. Ashjian P. De Ugarte D.A. Huang J.I. Mizuno H. Alfonso Z.C. Fraser J.K. Benhaim P. Hedrick M.H. Multilineage cells from human adipose tissue: implications for cell-based therapies.Tissue Eng. 2001; 7: 211-228Crossref PubMed Scopus (6322) Google Scholar, 2Halvorsen Y.D. Bond A. Sen A. Franklin D.M. Lea-Currie Y.R. Sujkowski D. Ellis P.N. Wilkison W.O. Gimble J.M. Thiazolidinediones and glucocorticoids synergistically induce differentiation of human adipose tissue stromal cells: biochemical, cellular, and molecular analysis.Metabolism. 2001; 50: 407-413Abstract Full Text PDF PubMed Scopus (182) Google Scholar, 3Erickson G.R. Gimble J.M. Franklin D.M. Rice H.E. Awad H. Guilak F. Chondrogenic potential of adipose tissue-derived stromal cells in vitro and in vivo.Biochem Biophys Res Commun. 2002; 290: 763-769Crossref PubMed Scopus (580) Google Scholar hASCs also differentiate into neuron-like cells expressing neuronal markers.4Kang S.K. Putnam L.A. Ylostalo J. Popescu I.R. Dufour J. Belousov A. Bunnell B.A. Neurogenesis of Rhesus adipose stromal cells.J Cell Sci. 2004; 117: 4289-4299Crossref PubMed Scopus (152) Google Scholar, 5Safford K.M. Safford S.D. Gimble J.M. Shetty A.K. Rice H.E. Characterization of neuronal/glial differentiation of murine adipose-derived adult stromal cells.Exp Neurol. 2004; 187: 319-328Crossref PubMed Scopus (257) Google Scholar, 6Trottier V. Marceau-Fortier G. Germain L. Vincent C. Fradette J. IFATS collection: using human adipose-derived stem/stromal cells for the production of new skin substitutes.Stem Cells. 2008; 26: 2713-2723Crossref PubMed Scopus (178) Google Scholar The differentiation ability of hASCs has generated interest because of their potential clinical use in regenerative medicine.7Zuk P.A. The adipose-derived stem cell: looking back and looking ahead.Mol Biol Cell. 2010; 21: 1783-1787Crossref PubMed Scopus (278) Google Scholar They meet the criteria for application in regenerative medicine in the following ways: i) they are found in abundant quantities, ii) they can be collected and harvested by a minimal invasive procedure, iii) they can be differentiated into multiple cell lineage pathways in a reproducible manner, and iv) they can be safely and effectively transplanted into an autologous or allogenic host.8Gimble J.M. Katz A.J. Bunnell B.A. Adipose-derived stem cells for regenerative medicine.Circ Res. 2007; 100: 1249-1260Crossref PubMed Scopus (1828) Google ScholarGene expression potential in stem cell differentiation is regulated by epigenetic processes that confer a specific chromatin conformation on the genome, of which post-translational modifications of histone tails and CpG dinucleotide methylation are the best characterized. Much attention is being paid to the effects of CpG methylation on stemness and differentiation. The first evidence came from the observation that genes important for the maintenance of pluripotency in embryonic stem cells (ESCs), such as OCT4 and NANOG, are usually hypomethylated when activated, whereas they become hypermethylated during differentiation.9Lagarkova M.A. Volchkov P.Y. Lyakisheva A.V. Philonenko E.S. Kiselev S.L. Diverse epigenetic profile of novel human embryonic stem cell lines.Cell Cycle. 2006; 5: 416-420Crossref PubMed Scopus (95) Google Scholar, 10Fouse S.D. Shen Y. Pellegrini M. Cole S. Meissner A. Van Neste L. Jaenisch R. Fan G. Promoter CpG methylation contributes to ES cell gene regulation in parallel with Oct4/Nanog, PcG complex, and histone H3 K4/K27 trimethylation.Cell Stem Cell. 2008; 2: 160-169Abstract Full Text Full Text PDF PubMed Scopus (335) Google Scholar High-throughput strategies for genome-wide DNA profiling demonstrate that human ESCs have a unique CpG methylation signature, which, in combination with histone modifications, drives stem cell differentiation by restricting the developmental potential of progenitor cells.11Bibikova M. Chudin E. Wu B. Zhou L. Garcia E.W. Liu Y. Shin S. Plaia T.W. Auerbach J.M. Arking D.E. Gonzalez R. Crook J. Davidson B. Schulz T.C. Robins A. Khanna A. Sartipy P. Hyllner J. Vanguri P. Savant-Bhonsale S. Smith A.K. Chakravarti A. Maitra A. Rao M. Barker D.L. Loring J.F. Fan J.B. Human embryonic stem cells have a unique epigenetic signature.Genome Res. 2006; 16: 1075-1083Crossref PubMed Scopus (234) Google Scholar, 12Calvanese V. Horrillo A. Hmadcha A. Suarez-Alvarez B. Fernandez A.F. Lara E. Casado S. Menendez P. Bueno C. Garcia-Castro J. Rubio R. Lapunzina P. Alaminos M. Borghese L. Terstegge S. Harrison N.J. Moore H.D. Brüstle O. Lopez-Larrea C. Andrews P.W. Soria B. Esteller M. Fraga M.F. Cancer genes hypermethylated in human embryonic stem cells.PLoS One. 2008; 3: e3294Crossref PubMed Scopus (67) Google Scholar, 13Mohn F. Weber M. Rebhan M. Roloff T.C. Richter J. Stadler M.B. Bibel M. Schübeler D. Lineage-specific polycomb targets and de novo DNA methylation define restriction and potential of neuronal progenitors.Mol Cell. 2008; 30: 755-766Abstract Full Text Full Text PDF PubMed Scopus (678) Google Scholar The epigenetic control of pluripotency is not restricted to the classic pluripotency-related genes, because it has been recently described that hypermethylation of tissue-specific genes also controls the reprogramming ability of somatic cells into pluripotent cells.14Barrero M.J. Berdasco M. Paramonov I. Bilic J. Vitaloni M. Esteller M. Izpisua Belmonte J.C. DNA hypermethylation in somatic cells correlates with higher reprogramming efficiency.Stem Cells. 2012; 30: 1696-1702Crossref PubMed Scopus (15) Google Scholar In contrast with the wide-ranging information obtained from ESCs, the role of CpG methylation in regulating differentiation of adult multipotent stem cells has been less extensively examined.15Berdasco M. Esteller M. DNA methylation in stem cell renewal and multipotency.Stem Cell Res Ther. 2011; 2: 42Crossref PubMed Scopus (79) Google Scholar Adult stem cells of different origin (eg, adipose tissue, bone marrow, or hematopoietic progenitor cells) display a range of differentiation potentials in mesodermal, endodermal, and ectodermal tissues,16Kern S. Eichler H. Stoeve J. Klüter H. Bieback K. Comparative analysis of mesenchymal stem cells from bone marrow, umbilical cord blood, or adipose tissue.Stem Cells. 2006; 24: 1294-1301Crossref PubMed Scopus (2496) Google Scholar and strong methylation of lineage-specification promoters restricts the ability of adult stem cells of different origin to differentiate.17Sørensen A.L. Timoskainen S. West F.D. Vekterud K. Boquest A.C. Ahrlund-Richter L. Stice S.L. Collas P. Lineage-specific promoter DNA methylation patterns segregate adult progenitor cell types.Stem Cells Dev. 2010; 19: 1257-1266Crossref PubMed Scopus (52) Google Scholar For example, methylation of endothelial cell–specific genes (CD31 and CD144) has been described in freshly isolated hASCs, but not in differentiated cells, after endothelial stimulation.18Boquest A.C. Noer A. Sørensen A.L. Vekterud K. Collas P. CpG methylation profiles of endothelial cell-specific gene promoter regions in adipose tissue stem cells suggest limited differentiation potential toward the endothelial cell lineage.Stem Cells. 2007; 25: 852-861Crossref PubMed Scopus (59) Google Scholar Preliminary genome-wide approaches, including gene expression, CpG methylation, histone marks, and microRNA (miRNA) analysis,19Aranda P. Agirre X. Ballestar E. Andreu E.J. Román-Gómez J. Prieto I. Martín-Subero J.I. Cigudosa J.C. Siebert R. Esteller M. Prosper F. Epigenetic signatures associated with different levels of differentiation potential in human stem cells.PLoS One. 2009; 4: e7809Crossref PubMed Scopus (90) Google Scholar allow a connection to be established between changes in the epigenetic signature and progression from pluripotent to multipotent cells, highlighting the existence of specific epigenetic profiles associated with each degree of differentiation potential. The epigenetic control of stem cell differentiation is also reinforced by several in vitro experimental studies with chromatin-modifying drugs. Specific epigenetic treatments can alter the potential of pluripotent and multipotent stem cells to differentiate into several lineages.15Berdasco M. Esteller M. DNA methylation in stem cell renewal and multipotency.Stem Cell Res Ther. 2011; 2: 42Crossref PubMed Scopus (79) Google Scholar For example, the use of DNA demethylation treatment (5-aza-2`-deoxycytidine) promotes differentiation of multipotent cells into cardiac myogenic cells20Choi S.C. Yoon J. Shim W.J. 5-Azacytidine induces cardiac differentiation of P19 embryonic stem cells.Exp Mol Med. 2004; 36: 515-523Crossref PubMed Scopus (100) Google Scholar and drives the osteogenic differentiation of mesenchymal stem cells.21Zhou G.S. Zhang X.L. Wu J.P. Zhang R.P. Xiang L.X. Dai L.C. Shao J.Z. 5-Azacytidine facilitates osteogenic gene expression and differentiation of mesenchymal stem cells by alteration in DNA methylation.Cytotechnology. 2009; 60: 11-22Crossref PubMed Scopus (51) Google ScholarDespite the knowledge about how to govern stem cell differentiation, we still lack information about the degree of similarity between stem cell derivatives and their normal primary counterparts. The success of stem cell differentiation is usually addressed by expression of a set of differentiation markers and loss of pluripotency-related genes. In addition, to achieve appropriate quality and control of these cells, we must also ensure the integrity of the epigenome and avoid inappropriate gene expression in transplanting cells or tumorigenesis. Herein, we have performed a high-throughput analysis using methylation arrays of well-characterized and defined populations of hASCs before and after in vitro induction of osteogenic and myogenic differentiation. Most important, the CpG methylation profile of these cells has been compared with those obtained from normal primary cells and tumor samples. Overall, our results demonstrate that hASCs generate osteogenic and myogenic lineages that resemble the DNA methylome of primary tissues, but do not present the epigenetic hallmarks of cancer cells. These findings suggest that the profile of CpG methylation could be used as a biomarker of efficient and safe cell identity after stem cell reprogramming.Materials and MethodsHuman Adipose-Derived Stem Cell PopulationshASCs were established from the adipose tissue of patients aged 30 to 55 years. Samples were obtained by minimally invasive liposuction procedures. Donors previously gave their written informed consent, in accordance with the guidelines of the Ethics Committee of the Nuestra Señora de la Salud Hospital (Granada, Spain). hASC lines were established as previously described.22Bunnell B.A. Flaat M. Gagliardi C. Patel B. Ripoll C. Adipose-derived stem cells: isolation, expansion and differentiation.Methods. 2008; 45: 115-120Crossref PubMed Scopus (768) Google Scholar Briefly, lipoaspirates were digested with 0.075% collagenase type I (Invitrogen SA, Barcelona, Spain) prepared in PBS containing 1% bovine serum albumin for 1 hour at 37°C with constant shaking, followed by filtration through a 100-μm filter. After washing with PBS, cells were treated with erythrocyte lysis buffer. Resultant cells were cultivated at 1000 cells/cm2 in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS). Cultures were washed with buffer for 24 to 48 hours after plating to remove unattached cells, and then refed with fresh medium. Only cells in passage 5 or 6 were used in these experiments. Positive and negative surface markers for hASCs, defined by the International Society for Cellular Therapy, were studied by fluorescence-activated cell sorting (FACSCanto II Cytometer; BD Biosciences, San Jose, CA). Cells were removed from culture using a nonenzymatic cell dissociation solution (Sigma-Aldrich, Madrid, Spain) and washed with PBS. The following antibodies were used: CD73, CD90, CD105, CD45, CD34, and CD133 (BD Biosciences). Approximately 2 × 105 cells were incubated with primary antibody directly coupled to fluorescein isothiocyanate, allophycocyanin, or phycoerythrin for 15 minutes in the dark at room temperature.In Vitro DifferentiationhASCs were differentiated to osteogenic and myogenic lineages by in vitro induction using specific culture medium.23Zuk P.A. Zhu M. Mizuno H. Huang J. Futrell J.W. Katz A.J. Benhaim P. Lorenz H.P. Hedrick M.H. Human adipose tissue is a source of multipotent stem cells.Mol Biol Cell. 2002; 13: 4279-4295Crossref PubMed Scopus (5410) Google Scholar Osteogenesis was induced in the presence of DMEM with 10% FBS, 0.1 mol/L dexamethasone, 10 mmol/L β-glycerophosphate, and 50 g/mL ascorbic acid-2-phosphate. After 28 days, the culture medium was removed and cells were washed with PBS and processed. Calcium deposition was visualized by alizarin red staining and confirmed by X-ray microanalysis of cultured cells under the scanning electron microscope (SEM), following the method of Kim et al.24Kim H.K. Kim J.H. Abbas A.A. Kim D.O. Park S.J. Chung J.Y. Song E.K. Yoon T.R. Red light of 647 nm enhances osteogenic differentiation in mesenchymal stem cells.Lasers Med Sci. 2009; 24: 214-222Crossref PubMed Scopus (68) Google Scholar In addition, RT-PCR analysis using mRNA obtained from total cells incubated in osteogenic medium was performed to determine gene expression of osteonectin, osteopontin, and osteocalcin. Myogenic induction of hASCs was performed in the presence of DMEM with 10% FBS, 50 mol/L hydrocortisone, and 5% horse serum. After 42 days, the degree of myogenic differentiation was analyzed by immunofluorescence using monoclonal antibodies against α-sarcomeric actin (clone 5C5; Sigma-Aldrich), α-sarcomeric actinin (clone EA-53; Sigma-Aldrich), and troponin T-C (clone C-19; Santa Cruz Biotechnology, Inc., Santa Cruz, CA). In addition, RT-PCR analysis using mRNA of these cells was used to determine gene expression of desmin, myosin-1, and myogenic differentiation 1. A complementary study to determine the regression of osteogenic and myogenic differentiation was performed by analyzing the induced cells after removing the culture medium supplemented with the differentiation agents over 14 and 20 days for osteogenic and myogenic induction, respectively. The RT-PCR primers and annealing temperatures (Tm) used were as follows: desmin (Tm = 51°C), 5′-GGTGGAGGTGCTCACTAACC-3′ (antisense) and 5′-TGTTGTCCTGGTAGCCACTG-3′ (antisense); myosin-1 (Tm = 51°C), 5′-TGTGAATGCCAAATGTGCTT-3′ (sense) and 5′-GTGGAGCTGGGTATCCTTGA-3′ (antisense); myogenic differentiation 1 (Tm = 52°C), 5′-AAGCGCCATCTCTTGAGGTA-3′ (sense) and 5′-GCGCCTTTATTTTGATCACC-3′ (antisense); osteocalcin (Tm = 56°C), 5′-GCTCTAGAATGGCCCTCACACTC-3′ (sense) and 5′-GCGATATCCTAGACCGGGCCGTAG-3′ (antisense); osteonectin (Tm = 53°C), 5′-TGTGGGAGCTAATCCTGTCC-3′ (sense) and 5′-TCAGGACGTTCTTGAGCCAGT-3′ (antisense); and osteopontin (Tm = 50°C), 5′-GCTCTAGAATGAGAATTGCACTG-3′ (sense) and 5′-GTCAATGGAGTCCTGGCTGT-3′ (antisense).Cancer Cell Lines and Primary TissuesThe human rhabdomyosarcoma (RD and TE.32.7) and osteosarcoma (MG-63) cell lines were obtained from ATCC (Rockland, MD). Cell lines were maintained in monolayer cultures at 37°C in an atmosphere containing 5% CO2, with DMEM supplemented with 10% FBS. Normal primary cells were obtained from biopsy specimens of the rectus abdominis muscle (myocytes) and ribs (osteocytes) under histological validation, and the samples were stored at −80°. All samples were obtained in accordance with the guidelines of the Ethics Committee of the Bio-Health Research Foundation of Eastern Andalusia (Granada, Spain).DNA Methylation Profiling Using Universal Bead ArraysDNA from adipose-derived stem cells, in vitro–differentiated cells, cancer cell lines, and primary tissues was isolated by applying the QIAamp DNA Mini Kit (Qiagen, Iberia, Spain). Microarray-based DNA methylation profiling was performed with the HumanMethylation2 BeadChip Infinium Methylation Arrays (Illumina, Inc.) on a total of five adipose-derived stem cells, six in vitro–differentiated cells, three normal primary tissues, and three cancer cell lines. The panel was designed to compare the DNA methylation status of each group of samples, which allow 27,578 CpG loci covering 14,495 genes at single-nucleotide resolution to be interrogated by typing bisulfite-converted DNA. The sequences included in the panel were derived from the well-annotated National Cancer for Biotechnology Information consensus coding sequence (CCDS) database (Genome Build 36) and were supplemented by >1000 cancer-related genes described in the literature. The probe content was enriched to include >150 well-established cancer genes known to show differential methylation patterns. The methylation array content also targeted the promoter regions of 110 miRNA genes.Methylation arrays were then performed. First, 1 μg of genomic DNA was bisulfite converted using the CpGenomicTM DNA Modification Kit (Intergen Company, Purchase, NY). After sodium bisulfite treatment, the remaining assay steps used Infinium technology (Illumina Inc, San Diego, CA), previously described for single-nucleotide polymorphism genotyping25Steemers F.J. Chang W. Lee G. Barker D.L. Shen R. Gunderson K.L. Whole-genome genotyping with the single-base extension assay.Nat Methods. 2006; 3: 31-33Crossref PubMed Scopus (290) Google Scholar with manufacturers' supplied reagents and conditions. A thermocycling program with a short denaturation step included for bisulfite conversion (16 cycles at 95°C for 30 seconds, followed by 50°C for 1 hour) was performed to improve the bisulfite conversion efficiency. After bisulfite conversion, each sample was whole-genome amplified and enzymatically fragmented. The bisulfite-converted whole-genome amplified DNA samples were purified and applied to the BeadChips (Illumina Inc, San Diego, CA). During hybridization, the whole-genome amplified DNA molecules annealed to locus-specific DNA oligomers linked to individual bead types. The two bead types corresponded to each CpG locus: one to the methylated and the other to the unmethylated state. Allele-specific primer annealing was followed by single-base extension using dinitrophenyl- and biotin-labeled dNTPs. Both bead types for the same CpG locus incorporated the same type of labeled nucleotide, determined by the base preceding the interrogated cytosine in the CpG locus, and could, therefore, be detected in the same color channel. After extension, the array was fluorescently stained and scanned, and the intensities of the unmethylated and methylated bead types were measured.DNA methylation values, described as β values, were recorded for each locus in each sample via BeadStudio software (Illumina Inc). The DNA methylation β value is a continuous variable between 0 (completely unmethylated) and 1 (completely methylated), and represents the ratio of the intensity of the methylated bead type/combined locus intensity. The DNA methylation microarray data are freely available for download from the National Cancer for Biotechnology Information Gene Expression Omnibus (http://www.ncbi.nlm.nih.gov/geo; accession number GSE33896).Hierarchical Cluster Analysis and Definition of CpG Methylation DifferencesHierarchical clustering was performed on all of the studied samples using the Cluster Analysis tool of the BeadStudio software version 3.2 (Illumina Inc). CpGs included in the analysis had to meet two criteria: a false-discovery rate of <0.01 and not being located on the X chromosome (to avoid a sex-specific bias). To investigate the methylation profiles, we first excluded the CpGs with a coefficient of variation of >0.25 for samples to discount intrasample variation among those of the same category. Averages were calculated from the resulting sequences. Differentially methylated CpG sites were determined calculating the differences in average β values between groups. A threshold of >0.20 change in average β values and a false-discovery rate of <0.01 in an analysis of variance test, adjusted for multiple testing, was applied to assign significant differentially methylated sites.26Benjamini Y. Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing.J R Stat Soc. 1995; B57: 289-300Google ScholarBisulfite Genomic Sequencing of Multiple ClonesWe determined the CpG island methylation status of the selected genes by PCR analysis of bisulfite-modified genomic DNA, which induces the chemical conversion of unmethylated, but not methylated, cytosine to uracil. After PCR and cloning, 10 clones of each sequence and sample were automatically sequenced to determine their degree of methylation. Primer sequences and annealing temperatures used were as follows: CD79B (Tm = 59°C), 5′-GGGGTAAGTATAGATAGAGGGGA-3′ (sense) and 5′-AATAAAAACAAACCCCACAAAC-3′ (antisense); CDKL1 (Tm = 59°C), 5′-CAAAATCATCTTCTAATTCCAAAA-3′ (sense) and 5′-TTAGTTTTTTGAGTTGTTGGGA-3′ (antisense); PIWI-like 2 (PIWIL2; Tm = 58°C), 5′-GTAGGTTGGGTTTTTGTTTAAGT-3′ (sense) and 5′-TAACCAAAAACAAAACCCC-3′ (antisense); receptor-type tyrosine-protein phosphatase S (PTPRS; Tm = 58°C), 5′-TGTGGGGGAGATATTTTAATTT-3′ (sense) and 5′-CCTTCTACTACACAACCTCCAA-3′ (antisense); serum amyloid A1 (SAA1; Tm = 58°C), 5′-GTTTTGAGTGAGGTTTTGTTGT-3′ (sense) and 5′-ATTTTCCCCAATTTATCAAATC-3′ (antisense); solute carrier family 44 member 2 (SLC44A2; Tm = 58°C), 5′-GGGTTTTTTAGTTTGGGTAGTT-3′ (sense) and 5′-TATTCTCTCAAAACCCCCTC-3′ (antisense); and transmembrane channel-like 8 (TMC8; Tm = 62°C), 5′-ACTCCRACACCCCAAAAA-3′ (sense) and 5′-GTGTGTTTTTTTGAGAGTTGGAG-3′ (antisense).Expression MicroarraysFor the expression microarray experiment in hASCs and in vitro differentiated cells, total RNA was prepared using TRIzol (Invitrogen, Carlsbad, CA) and further purified using RNeasy columns (Qiagen, GmbH, Germany), according to the manufacturers' instructions. The integrity of RNA was monitored by both Byoanalyzer (Agilent, Santa Clara, CA) and spectrometric measurement. Biotinylated target RNA was prepared from 200 ng of total RNA using the Affymetrix protocol (Affymetrix, Santa Clara, CA). Briefly, double-stranded cDNA was prepared from the RNA template using the Ambion WT Expression kit, following manufacturer's instructions. The resulting complementary RNA target was randomly fragmented using uracil DNA glycosylase and apurinic/apyrimidic endonuclease 1, labeled with the terminal transferase (WT Terminal Labeling kit; Affymetrix) and hybridized on the Human Gene 1.1 ST 16-array (Affymetrix). The hybridization reactions were processed and scanned according to the standard GeneTitan (Affymetrix) protocols. All arrays were globally scaled to a target intensity value of 100 and then the scaling factor, background, noise, and percentage present were calculated according to the Command Console (AGCC 2.0) and Expression Console (EE 1.1) software (Affymetrix, High Wycombe, UK). All resulting data sets were filtered using the absolute call metric (present or absent) using Microsoft Access (Microsoft Corporation, Redmond, WA). The resulting values represent the transcript abundance (absolute value) following the Affymetrix statistical procedure. Three biological replicates were obtained for each ASC sample, and in vitro–differentiated cells were analyzed in duplicate. The DNA expression microarray data are freely available for download from NCBI Gene Expression Omnibus (http://www.ncbi.nlm.nih.gov/geo; accession number GSE37329).Transcriptome Data AnalysisTo identify all genes or probe sets whose expression was significantly different between control adipose-derived stem cells and each one of the two differentiated cell types (osteogenis or myogenic differentiation), we used the following approach under R statistical language: the raw data coming from the Affymetrix Human Gene 1.1 ST platform were preprocessed and normalized using the Affymetrix package (background subtraction and quantile normalization), then the differential expression was assessed using a linear model procedure (limma package).27Smyth G.K. Limma: linear models for microarray data.in: Gentleman R. Carey V. Dudoit S. Irizarry R. Huber W. Bioinformatics and Computational Biology Solutions Using R and Bioconductor. Springer, New York2005: 397-420Crossref Google Scholar The differentially expressed genes with at least 20% expression increase (fold-change >1.2) or 20% expression decrease (fold-change <0.8) after induction were selected, and the resultant P values were adjusted for multiple testing using the Benjamini and Hochberg correction procedure.Gene Ontology AnalysisGene ontology (GO) analysis of the selected genes was performed using the Bioconductor package GOstats (Bioconductor, Boston, MA).28Falcon S. Gentleman R. Using GOstats to test gene lists for GO term association.Bioinformatics. 2007; 23: 257-258Crossref PubMed Scopus (1375) Google Scholar The set of selected genes was tested for enrichment of any GO category, and the P values for multiple testing were adjusted using the Benjamini and Hochberg correction procedure.Comparisons between Differentially Methylated and Differentially Expressed GenesBecause of the fact that the genes differentially expressed and methylated are not the same for myogenic and osteogenic lineages, we conducted a difference between two lists using their functional profiles. We used squared euclidean distance to quantify the difference between profiles26Benjamini Y. Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing.J R Stat Soc. 1995; B57: 289-300Google Scholar and calculate P values.RT-qPCR Expression AnalysisTotal RNA was prepared from all samples using TRIzol (Invitrogen, Carlsbad) and further purified using RNeasy columns (Qiagen, GmbH), according to the manufacturers' instructions. For quantitative RT-PCR assays, 2 μg of total RNA was converted to cDNA with the ThermoScriptTM RT-PCR System (Invitrogen) using oligo-dT as primer. PCR amplifications were performed as follows: 0.20 μg of cDNA and 5 pmol of each primer and SYBR Green PCR Master Mix (Applied Biosystems, Foster City, CA). Three measurements were analyzed using a Prism 7700 Sequence Detection (Applied Biosystems) instrument. Quantitative RT-PCR (RT-qPCR) primer sequences were as follows: PIWIL2, 5′-GTGGGTTGAGCTCGGTCTT-3′ (sense) and 5′-GGACGGGCTGTAGAGAACAC-3′ (antisense); PTPRS, 5′-ACTCGGCCAACTACACCTG-3′ (sense) and 5