TY - JOUR
AU - Soong, Yung-Kuei
AB - Abstract Using high-density oligonucleotide microarrays and functional network analyses, we examined whether MSCs derived from four different origins exhibited unique gene expression profiles individually and then compared the gene expression profiles of all MSCs with those of fetal organs. Our results indicated that within each group of MSCs from the same origin, the variability of the gene expression levels was smaller than that between groups of different origins. Functional genomic studies revealed the specific roles of MSCs from different origins. Our results suggest that amniotic fluid MSCs may initiate interactions with the uterus by upregulating oxytocin and thrombin receptors. Amniotic membrane MSCs may play a role in maintaining homeostasis of fluid and electrolytes by regulating the networks of endothelin, neprilysin, bradykinin receptors, and atrial natriuretic peptide. Cord blood MSCs may be involved in innate immune systems as the neonatal defense system against the earliest encountered pathogens. Adult bone marrow MSCs may be an important source not only of all blood lineages but also of bone formation. However, in spite of the different gene expression profiles seen in MSCs derived from different origins, a set of core gene expression profiles was preserved in these four kinds of MSCs. The core signature transcriptomes of all MSCs, when contrasted against those of fetal organs, included genes involved in the regulation of extracellular matrix and adhesion, transforming growth factor-β receptor signaling, and the Wnt signaling pathways. Disclosure of potential conflicts of interest is found at the end of this article. Mesenchymal stem cells, Microarray, Transcriptome, Functional network analysis Introduction Multipotent MSCs are a promising cell resource for tissue engineering and cell-based therapeutics because of their ability to self-renew and differentiate into specific functional cell types [1–3]. Murine MSCs have been identified in all organs and tissues except the peripheral blood of adult mice [4]. Human MSCs have been isolated from bone marrow [5], adipose tissues [6, 7], cord blood [8], amniotic fluid [9, 10], amniotic membrane [11, 12], placentae [13, 14], and umbilical cord tissues [15, 16]. MSCs may also be transdifferentiated into cells outside of the mesenchymal lineages [17, 18]. The heterogeneity of MSCs has been proposed as an explanation for their pluripotency and differentiation potentials [1]. It has been suggested that the MSC pool comprises both putative mesenchymal stem cells and subpopulations at different states of differentiation [19]. On the other hand, single-cell-derived, colony-expanded multipotent MSCs have been isolated from umbilical cord blood [20] and from amniotic fluid [10]. High-throughput, genome-wide analyses of transcriptome have recently been used to identify characteristics of human cord blood-derived MSCs [21], murine MSCs [22], mouse embryonic stem cells and adult neural stem cells [23], and human cord blood-derived CD133+ cells [24]. It is interesting to explore both the core feature of all MSCs and the individual signature gene expression profile of each group of MSCs derived from different origins. In this study, we examined whether gene expression levels of human MSCs remain stable during continuous culturing in vitro. We also hypothesized that MSCs derived from different origins will exhibit unique gene expression profiles, whereas all MSCs may express a common core profile that is distinct from differentiated organs. Using high-density oligonucleotide microarrays in this study, we compared the gene expression profiles of four groups of MSCs that were isolated from amniotic fluid, amniotic membrane, umbilical cord blood, and adult bone marrow. Furthermore, we contrasted the gene expression profiles of all these MSCs to those of six fetal organs (brain, heart, lung, liver, kidney, and muscle). Materials and Methods Culture of Mesenchymal Stem Cells MSCs derived from amniotic fluid (AF; n = 6), amniotic membrane (AM; n = 5), cord blood (CB; n = 5), and bone marrow (BM; n = 5) were obtained from the Bioresource Collection and Research Center (Hsinchu, Taiwan, http://www.bcrc.firdi.org.tw). Originally, AF MSCs were isolated from amniotic fluid that was collected during 16–20 weeks of pregnancy [9]. CB MSCs were isolated from umbilical cord blood collected after delivery of term pregnancies [25]. BM MSCs were isolated from three different batches of frozen bone marrow-derived mononuclear cells supplied by Cambrex (Walkersville, MD, http://www.cambrex.com) and two fresh bone marrow samples (from individuals ages 19 and 31) [25]. AM MSCs were isolated from collagenase type II (Sigma-Aldrich, St. Louis, http://www.sigmaaldrich.com)-treated amniotic membrane stromal matrix collected during the caesarean sections of term pregnancies. The Institutional Review Board of Cathay General Hospital (Taipei, Taiwan) approved this protocol, and each patient signed a written informed consent except those whose samples were commercially supplied by Cambrex. MSCs were cultured in α-modified minimum essential medium (α-MEM; HyClone, Logan, UT, http://www.hyclone.com) supplemented with 20% fetal bovine serum (HyClone) and 4 ng/ml basic fibroblast growth factor (R&D Systems Inc., Minneapolis, http://www.rndsystems.com) and incubated at 37°C in a humidified atmosphere with 5% CO2. All MSCs had normal karyotypes and were tested at passages of 3–6 in this study. Phase-contrast microscopic examinations, differentiation capabilities, and cell surface antigen profiles by flow cytometrical analyses were performed to validate MSC characteristics suggested by the International Society for Cellular Therapy [26]. Flow Cytometry MSCs of different origins at passages 5–6 were immunophenotyped by flow cytometric analysis. Cells were trypsinized, washed, and resuspended in phosphate-buffered saline (PBS; Sigma-Aldrich). After fixing and blocking, the cells were immunolabeled with the following mouse anti-human antibodies: CD34, CD45 (Miltenyi Biotec, Auburn, CA, http://www.miltenyibiotec.com); CD29, CD31, CD44, human leukocyte antigen (HLA)-A, -B, and -C and HLA-DR (BD Biosciences, San Jose, CA, http://www.bdbiosciences.com); CD105 (also known as endoglin or SH2); and CD73 (also known as SH3 and SH4; purified from culture supernatants of hybridoma cells; American Type Culture Collection, Manassas, VA, http://www.atcc.org). The nonspecific mouse IgG (Vector Laboratories, Burlingame, CA, http://www.vectorlabs.com) was substituted for the primary antibodies as isotype control. The secondary antibody, anti-mouse IgG-fluorescein isothiocyanate or IgG-phycoerythrin (Vector Laboratories), was incubated with cells and analyzed by flow cytometry (Becton, Dickinson and Company, San Jose, CA, http://www.bd.com). Multipotent Differentiation MSCs established from different origins were seeded into six-well plates and grown to confluence. To induce osteogenic differentiation, cells were incubated in α-MEM supplemented with 10% fetal bovine serum (FBS), 0.1 μM dexamethasone (Sigma-Aldrich), 10 mM β-glycerophosphate (Sigma-Aldrich), and 50 μM ascorbic acid (Sigma-Aldrich) for 3 weeks and analyzed by von Kossa staining as previously described [5]. To initiate adipogenic differentiation, cells were incubated in α-MEM supplemented with 10% FBS, 1 μM dexamethasone, 0.5 mM methylisobutylxanthine (Sigma-Aldrich), 10 μg/ml insulin (Gibco-BRL, Carlsbad, CA, http://www.gibcobrl.com), and 100 μM indomethacin (Sigma-Aldrich) for 3 weeks and assessed by oil red O staining [5]. To promote chondrogenic differentiation, cells were trypsinized, centrifuged, and cultured as pelleted aggregates in serum-free medium supplemented with 10 ng/ml transforming growth factor (TGF)-β3 (R&D Systems) in 15-ml centrifuge tubes. After incubation for 3 weeks, the cell pellets were embedded in paraffin, sectioned (5 μm thickness), histologically examined, and stained with Safranin O [5]. Designs of DNA Microarray Experiments To evaluate the stability of gene expression profiles for MSCs, all the AF, AM, CB, and BM MSCs at the third passage were thawed and cultured in vitro at different time periods until the sixth passage. Three collections from each sample of cells were obtained for RNA extraction. To analyze the variations among MSC samples within the same group, RNA was extracted from additional MSCs (five AF, four AM, four CB, and four BM). For DNA microarray analysis, however, only two more MSC samples in each group were used (Fig. 1). Figure 1. Open in new tabDownload slide Flow chart of the study design. Three samples (passages 3–6) of one MSC from each group were analyzed after storage in liquid nitrogen for various period of time (labeled as 1a, 1b, and 1c) to study the stability of gene expression profile of individual MSCs. To identify the signature expression profile in each group of MSCs, three MSCs from each group were compared with the remaining nine MSCs that were selected from three other groups (three in each group). To identify the core expression profile in all MSCs, a total of 12 MSCs derived from four origins were compared with six organs from fetuses at the sixth gestational week. Abbreviations: AF, amniotic fluid; AM, amniotic membrane; BM, bone marrow; CB, cord blood. Figure 1. Open in new tabDownload slide Flow chart of the study design. Three samples (passages 3–6) of one MSC from each group were analyzed after storage in liquid nitrogen for various period of time (labeled as 1a, 1b, and 1c) to study the stability of gene expression profile of individual MSCs. To identify the signature expression profile in each group of MSCs, three MSCs from each group were compared with the remaining nine MSCs that were selected from three other groups (three in each group). To identify the core expression profile in all MSCs, a total of 12 MSCs derived from four origins were compared with six organs from fetuses at the sixth gestational week. Abbreviations: AF, amniotic fluid; AM, amniotic membrane; BM, bone marrow; CB, cord blood. RNA Preparation and Microarray Analysis At 90% confluence, the MSCs were briefly rinsed with ice-cold PBS and lysed in TRIZOL reagent (Invitrogen, Carlsbad, CA, http://www.invitrogen.com) for RNA extraction. The procedures of RNA extraction using TRIZOL and the RNeasy purification kit (Qiagen, Valencia, CA, http://www1.qiagen.com) and confirmation of RNA quality and quantity with the Bioanalyzer 2100 (Agilent Technologies, Palo Alto, CA, http://www.agilent.com) were similar to previous reports [27–30]. Total RNA specimens of human fetal brain, heart, lung, liver, kidney, and muscle at the sixth gestational week were obtained from ViroGen Inc. (Watertown, MA, http://www.virogen.com). Gene expression profiles in MSCs and early fetal tissues were analyzed with the human U133A GeneChip (Affymetrix, Santa Clara, CA, http://www.affymetrix.com), and the manufacturer's protocol was strictly followed. Real-Time Quantitative Polymerase Chain Reaction Analysis Real-time quantitative polymerase chain reaction (RT-Q-PCR) was carried out in the ABI 7900 HT RT-PCR system (Applied Biosystems, Foster City, CA, http://www.appliedbiosystems.com) using Assay-on-Demand TaqMan probe sets (Applied Biosystems) according to the manufacturer's protocol. Statistical Analysis The nonparametric Mann-Whitney U test was used with Statistica software, version 6.1 (Statsoft Inc., Tulsa, OK, http://www.statsoft.com). Depending on the desired stringency of the comparison, p values of less than 0.05–0.000005 were used for statistical significance and specified in the individual results. Pearson linear regression was used to analyze the similarity of gene expression levels between every two microarray profiles. We only selected genes that were designated “present” by Affymetrix data analysis suites for linear regression statistics. All the specimens in this study resulted in 42%–50% of genes designated present. Network Visualization and Analysis Network analyses of differentially expressed genes were performed using MetaCore Analytical Suite (GeneGo Inc., St Joseph, MI, http://www.genego.com) [31, 32]. MetaCore is a web-based computational platform designed for systems biology and drug discovery. It includes a curated database of human protein interactions and metabolism; thus, it is useful for analyzing a cluster of genes in the context of regulatory networks and signaling pathways. For the network analysis of a group of genes, MetaCore can be used to calculate the statistical significance (p value) based on the probability of assembly from a random set of nodes (genes) of the same size as the input list [32]. Details of the analytical algorithm and statistics of MetaCore are described in the supplemental online data. Results Characteristics of Mesenchymal Stem Cells of Different Origins Mesenchymal stem cells isolated from human amniotic fluid, amniotic membrane, cord blood, and bone marrow showed a uniform spindle-shaped morphology (Fig. 2A). In vitro differentiation analysis confirmed that all isolated MSCs from different origins exhibited the capacity to differentiate into various cell types, such as osteoblasts, chondrocytes, and adipocytes (Fig. 2A). The population doubling time of these four types of MSCs at passages 5–6 was 30–36 hours in this study. For further characterization of MSCs, a panel of surface markers was tested using flow cytometric analysis. MSCs from all four origins were negative for CD31 (endothelial cell marker), CD34, CD45 (both as hematopoietic markers), and HLA-DR (human leukocyte differentiation antigen class II), whereas they were positive for CD29, CD44 (both as adhesion markers), CD105, CD73 (both as mesenchymal markers), and HLA-A, -B, and -C (class I) (Fig. 2B). Figure 2. Open in new tabDownload slide Characteristics of four mesenchymal stem cells. (A): Morphology and multilineage differentiation capacity. All MSCs that were isolated from AF, AM, CB, and BM showed a homogeneous spindle-shaped morphology. Osteogenesis was analyzed by von Kossa staining for the mineral nodule deposition. Chondrogenesis was assessed by Safranin O staining for proteoglycan deposition. Adipogenesis was observed by the existence of lipid vesicles and confirmed by oil red O staining. (B): Immunophenotype of MSCs by flow cytometric analysis. Representative histograms are demonstrated, and their respective isotype controls are shown by the dotted lines. The staining patterns of these four types of MSCs were highly similar. The cells were positive for CD29, CD44, CD73, CD105, and HLA-A, -B, and -C and negative for CD31, CD34, CD45, and HLA-DR. Abbreviations: AF, amniotic fluid; AM, amniotic membrane; BM, bone marrow; CB, cord blood; HLA, human leukocyte antigen. Figure 2. Open in new tabDownload slide Characteristics of four mesenchymal stem cells. (A): Morphology and multilineage differentiation capacity. All MSCs that were isolated from AF, AM, CB, and BM showed a homogeneous spindle-shaped morphology. Osteogenesis was analyzed by von Kossa staining for the mineral nodule deposition. Chondrogenesis was assessed by Safranin O staining for proteoglycan deposition. Adipogenesis was observed by the existence of lipid vesicles and confirmed by oil red O staining. (B): Immunophenotype of MSCs by flow cytometric analysis. Representative histograms are demonstrated, and their respective isotype controls are shown by the dotted lines. The staining patterns of these four types of MSCs were highly similar. The cells were positive for CD29, CD44, CD73, CD105, and HLA-A, -B, and -C and negative for CD31, CD34, CD45, and HLA-DR. Abbreviations: AF, amniotic fluid; AM, amniotic membrane; BM, bone marrow; CB, cord blood; HLA, human leukocyte antigen. Stability of Gene Expression Profiles of MSCs To examine the stability of gene expression of MSCs, we calculated the coefficient of determination (R2) of each pair of comparison among three RNA preparations from the same MSCs at the sixth passage, some of which had been cryopreserved in liquid nitrogen for 7–14 days from passages 3–6 (Fig. 1). The R2 values are presented as means ± SD (Table 1). Likewise, we calculated the R2 of each pair of comparison among 5 DNA microarrays within each MSC group and derived all the coefficients of determination (Table 1). Specimens within the same MSCs but from different time points had significantly greater values of R2 than those within the same MSC group but from different samples; these data indicate that the gene expression levels of each MSCs were stable at the same passage of in vitro culture, even when some of them had been cryopreserved in liquid nitrogen before they were thawed and cultured (Table 1). A similar trend was observed in the CB group, although the p value did not reach statistical significance. This may be due to the small sample size. Dot plot analyses for comparison of those differences are shown in supplemental online Figure 1. Table 1. Analysis of correlations between MSCs within the same group using the coefficient of determination (r2) Open in new tabDownload slide Open in new tabDownload slide Open in new tabDownload slide Open in new tabDownload slide Open in new tab Table 1. Analysis of correlations between MSCs within the same group using the coefficient of determination (r2) Open in new tabDownload slide Open in new tabDownload slide Open in new tabDownload slide Open in new tabDownload slide Open in new tab Gene Expression Profiles of MSCs from Different Origins Are Closer to Each Other Than Those in Different Fetal Organs Using the same statistical approach, we identified the average R2 of comparisons between MSC groups to range from 0.874 to 0.902 (Table 2), which was a little lower than the range of 0.884–0.930 for the average R2 of comparisons within each group (Table 1). Nevertheless, the average R2 of the comparisons between MSC groups was significantly greater than the average R2 of comparisons between the six fetal organs, 0.779 ± 0.097 (p < .00001). These results indicated that gene expression profiles of all MSCs were more similar than those of different fetal organs, suggesting the existence of a core gene expression profile for all MSCs, independent of tissue origins. Table 2. Analysis of correlations between MSCs from different origins using the coefficient of determination (r2) Open in new tabDownload slide Open in new tabDownload slide Open in new tabDownload slide Open in new tabDownload slide Open in new tab Table 2. Analysis of correlations between MSCs from different origins using the coefficient of determination (r2) Open in new tabDownload slide Open in new tabDownload slide Open in new tabDownload slide Open in new tabDownload slide Open in new tab Validation of Microarray Results with Real-Time Quantitative PCR The results of our cDNA microarrays have been shown to be consistent with those of RT-Q-PCR [27–30]. Nevertheless, we have selected two genes from each group of MSCs and obtained gene expression levels for all specimens (six AF, five AM, five CB, and five BM specimens) using RT-Q-PCR (supplemental online Fig. 2). Results were consistent with DNA microarray data. Based on the findings of recent studies [33, 34], we did not further reanalyze all genes of interest exhaustively with RT-Q-PCR. The MicroArray Quality Control study, a collaborative effort led by the U.S. Food and Drug Administration that included 137 scientists from private and public sector laboratories, found that microarray gene expression analysis can be used as a stand-alone quantitative comparison [33]. The correlation between Affymetrix gene expression results and TaqMan RT-Q-PCR results was further shown in a good linearity (R2 = 0.95) [34]. As long as the manufacturer's protocols are strictly followed, Affymetrix chips have been shown to have good interlaboratory reproducibility [33]. Functional Analysis of Signature Gene Expression in AF MSCs To identify the individual gene expression signature of each group of MSCs, we first averaged the three gene expression levels (AF1a, AF1b, and AF1c) of one AF MSC (Fig. 1). We then combined this result with two additional MSCs (AF2 and AF3), and compared them with three other MSCs (AM, CB, and BM) using Mann-Whitney U tests. To generate the list of differentially expressed genes, a fold-change ranking method was recently advocated to be more reproducible than t test p value or significance analysis of microarrays [35]. After filtering with p < .05, we ranked genes by fold change and chose the top 25 genes that were upregulated in AF MSCs (Table 3). The upregulated genes in AF MSCs included genes involved in uterine maturation and contraction, such as OXTR (oxytocin receptor) and PLA2G10 (regulation of prostaglandin synthesis). Other upregulated genes in this group are involved in signal transduction pathways: thrombin-triggered responses (F2R and F2RL), hedgehog (HHAT), and G-protein related (RHOF, RGS5, PLCB4, and RGS7). Several AF MSC-downregulated genes are listed in supplemental online Table 1. Table 3. Signature gene expressions Open in new tabDownload slide Open in new tabDownload slide Open in new tabDownload slide Open in new tabDownload slide Open in new tabDownload slide Open in new tabDownload slide Open in new tabDownload slide Open in new tabDownload slide Open in new tab Table 3. Signature gene expressions Open in new tabDownload slide Open in new tabDownload slide Open in new tabDownload slide Open in new tabDownload slide Open in new tabDownload slide Open in new tabDownload slide Open in new tabDownload slide Open in new tabDownload slide Open in new tab To annotate gene functions at the network level, signature genes that were upregulated in AF MSCs were extrapolated using the intersection algorithm of MetaCore, as shown in Figure 3A. The p value for the resulting network was 2.02 × 10−11, indicating that the probability of assembly from a random set of nodes (genes) was very low [32]. The top 10 biological processes associated with genes in this network are summarized in supplemental online Table 2. Figure 3. Open in new tabDownload slide Functional networks of individual mesenchymal stem cells. Some genes are labeled by aliases. (A): The genes (blue circles) that were upregulated in amniotic fluid MSCs (also listed in Table 3) were extrapolated into the networks using the MetaCore program. The related biological processes in this network are listed in supplemental online Table 2. (B): The genes (blue circles) that were upregulated in amniotic membrane MSCs (also listed in Table 3) were extrapolated into the networks using the MetaCore program. The related biological processes in these networks are listed in supplemental online Table 3. (C): The genes (blue circles) that were upregulated in cord blood MSCs (also listed in Table 3) were extrapolated into the networks using MetaCore program. The related biological processes in this network are listed in supplemental online Table 4. (D): The genes (blue circles) that were upregulated in bone marrow MSCs (also listed in Table 3) were extrapolated into networks using the MetaCore program. The related biological processes in these networks are listed in supplemental online Table 5. Alias names: 70Z-PEP, TPTN22; B1R, BDKRB1; DAL1, EPB41L3; IBP1, IGFBP1; IL-33, C9orf26; Mindin, SPON2; Neprilysin, MME; OTR, OXTR; PAR1, F2R; PAR3, F2RL2. Figure 3. Open in new tabDownload slide Functional networks of individual mesenchymal stem cells. Some genes are labeled by aliases. (A): The genes (blue circles) that were upregulated in amniotic fluid MSCs (also listed in Table 3) were extrapolated into the networks using the MetaCore program. The related biological processes in this network are listed in supplemental online Table 2. (B): The genes (blue circles) that were upregulated in amniotic membrane MSCs (also listed in Table 3) were extrapolated into the networks using the MetaCore program. The related biological processes in these networks are listed in supplemental online Table 3. (C): The genes (blue circles) that were upregulated in cord blood MSCs (also listed in Table 3) were extrapolated into the networks using MetaCore program. The related biological processes in this network are listed in supplemental online Table 4. (D): The genes (blue circles) that were upregulated in bone marrow MSCs (also listed in Table 3) were extrapolated into networks using the MetaCore program. The related biological processes in these networks are listed in supplemental online Table 5. Alias names: 70Z-PEP, TPTN22; B1R, BDKRB1; DAL1, EPB41L3; IBP1, IGFBP1; IL-33, C9orf26; Mindin, SPON2; Neprilysin, MME; OTR, OXTR; PAR1, F2R; PAR3, F2RL2. Figure 3. Open in new tabDownload slide (Continued) Figure 3. Open in new tabDownload slide (Continued) Functional Analysis of Signature Gene Expression in AM MSCs Using a method similar to the aforementioned analysis of AF MSCs, we compared the gene expression profiles of AM MSCs with those of the other three MSC groups. After filtering with p < .05, we ranked genes by fold change and chose the top 25 genes that were upregulated in AM MSCs (Table 3). Several upregulated genes in AM MSCs are involved in the regulation of the immune adaptation between the materno-placental interface (SPON2, IFI27, BDKRB1, SCYB5, SCYB6, and LYN). Other upregulated genes in AM MSCs include transcription factors (FOXF1, HAND2, and TCF21) and metabolic enzymes (DPP6, TDO2, and ST6GALNAC5). Several AM MSC-downregulated genes are listed in supplemental online Table 1. To annotate gene functions at the network level, signature genes that were upregulated in AM MSCs were extrapolated using the intersection algorithm of MetaCore as shown in Figure 3B. The p values for the resulting two networks were 3.17 × 10−10 and 4.68 × 10−11, indicating that the probabilities of assembly from random sets of nodes (genes) were very low [32]. The top 10 biological processes associated with genes in these networks are summarized in supplemental online Table 2. Significantly, the two networks that consisted of different sets of genes (Table 3) (SPON2 [Mindin], DPP4, LYN, TCF21, and BDKRB1 [B1R], shown in the upper panel of Figure 3B; and C9orf26 [IL-33], membrane metalloendopeptidase [MME; Neprilysin], and HAND2, shown in the lower panel of Figure 3B) were directed at the same functions: maintenance of the homeostasis of fluid and electrolytes and regulation of blood pressure (supplemental online Table 3A, 3B). Functional Analysis of Signature Gene Expression in CB MSCs After filtering with p < .05, we ranked genes by fold change and chose the top 25 genes that were upregulated in CB MSCs (Table 3). Upregulated genes in CB MSCs are those involved in the regulation of insulin growth factors (IGFBP1); in the regulation of tumorigenesis (ZIC1 and EPB41L3); in blood function (thrombin-related [THBD]) and maintenance of blood cell shape (ANK1); in the development of neural tissues (MAB21L2), craniofacial structure (DLX2), and tendons and ligaments (SIX2); in the peroxisome proliferator-activated receptor pathway (PPARG and PTGIS); and in the metabolic catalysis of neuropeptide (CPE), prostacyclin (PTGIS), and arginine (ASS); there were also two relatively novel genes (APOLD1 and KBTBD11). Interestingly, downregulated genes in CB MSCs include a series of homeodomain proteins that control basic developmental programs (Hox-A9, -B6, -B7, -C6, and -C10 and PITX2) and the gene CDKN2A, which codes for inhibitors of the CDK4 (p14ARF and p16INK4) (supplemental online Table 1). To annotate gene functions at the network level, signature genes that were upregulated in CB MSCs were extrapolated using the intersection algorithm of MetaCore as shown in Figure 3C. The p value for the resulting network was 2.87 × 10−10, indicating that the probability of assembly from a random set of nodes (genes) was very low [32]. The top 10 biological processes associated with genes in this network are summarized in supplemental online Table 4. Functional Analysis of Signature Gene Expression in BM MSCs After filtering with p < .05, we ranked genes by fold change and chose the top 25 genes that were upregulated in BM MSCs (Table 3). Upregulated genes in BM MSCs are those involved in major histocompatibility complex (HLA-DPA1, -DRA, -DPB1, -DRB1, and MR1), in interferon-derived functions (IFI30 and IFI44L), in regulation of myeloid differentiation and growth arrest (SCAP2), in regulation of the complement cascade (SERPING1), in development programs regulated by homeodomain proteins (HOX-A1, -A5, -C10, EN1, DLX5, and CART1), in regulation of transport (HEPH, SLC14A1, SLC1A3, and KCNN4), in regulation of mRNA polyadenylation and translational activation (CPEB1), in regulation of bone and cartilage formation (BMP2), and in the coding of various enzymes (ENPP1, PTPN22, and HAS2). On the other hand, some genes that are involved in the differentiation and functions of blood cells were downregulated in BM MSCs, including those that encode coagulation factor III (F3), those that facilitate pre-B-cell growth (BST1), and those required for cytokines (SEPT6) (supplemental online Table 1). To annotate gene functions at the network level, signature genes that were upregulated in BM MSC were extrapolated using an intersection algorithm of MetaCore as shown in Figure 3D. The p values for the resulting two networks were 3.27 × 10−12 and 2.16 × 10−06, indicating that the probabilities of assembly from random sets of nodes (genes) were very low [32]. The top 10 biological processes associated with genes in these networks are summarized in supplemental online Table 5. The Core Gene Expression Profile Shared by Four Kinds of MSCs We then compared 12 MSCs as one group with the six fetal organs as another group. By contrasting the transcriptomes of MSCs from various origins with differentiated human tissues at the earliest ages, we identified a core gene expression profile that is common to these four kinds of MSCs. Using the highly stringent criteria (p < .000005 and >4 fold change), we identified 48 genes that were differentially upregulated in MSCs (Table 4) and 11 genes that were downregulated in MSCs (Table 4). Table 4. Core expression profiles Open in new tabDownload slide Open in new tabDownload slide Open in new tabDownload slide Open in new tabDownload slide Open in new tab Table 4. Core expression profiles Open in new tabDownload slide Open in new tabDownload slide Open in new tabDownload slide Open in new tabDownload slide Open in new tab Genes in the core expression profile of all MSCs are involved in plasmin-related extracellular matrix remodeling (CD44, collagen II, IGF-2, PLAU, PLAUR, SERPINE 1, and TIMP1), cytoskeletal regulation (ACTN1, ARPC1B, Cav1, Cav2, CDKN1A, PLAU, PLAUR, and SERPINE 1), chemokine regulation and adhesion (ACTN1, ARPC1B, Cav1, Cav2, CDKN1A, PLAU, PLAUR, SERPINE 1, and thrombospondin 1), plasmin activation (PLAU and TFPI-2), and TGF-β receptor signaling (Cav1, Cav2, CDKN1A, SMURF, and SERPINE-1; Fig. 4). Figure 4. Open in new tabDownload slide The core expression profile in all mesenchymal stem cells derived from four origins. The core expression profiles of MSCs were identified by contrasting the gene expression profiles of all types of MSCs from four origins (amniotic fluid, amniotic membrane, fetal cord blood, and adult bone marrow) with those of six fetal organs (brain, heart, lung, liver, kidney, and muscle). The genes of core expression profile in all MSCs (also listed in Table 4) are tightly associated with plasmin-related extracellular matrix remodeling, chemokines and adhesion, regulation of cytoskeleton, and TGF-β receptor signaling. The genes that were significantly upregulated in MSCs are marked by red boxes, and those that were downregulated in MSCs are marked by green boxes. Abbreviations: IGF, insulin growth factor; MMP, matrix metalloproteinase; TGF, transforming growth factor. Figure 4. Open in new tabDownload slide The core expression profile in all mesenchymal stem cells derived from four origins. The core expression profiles of MSCs were identified by contrasting the gene expression profiles of all types of MSCs from four origins (amniotic fluid, amniotic membrane, fetal cord blood, and adult bone marrow) with those of six fetal organs (brain, heart, lung, liver, kidney, and muscle). The genes of core expression profile in all MSCs (also listed in Table 4) are tightly associated with plasmin-related extracellular matrix remodeling, chemokines and adhesion, regulation of cytoskeleton, and TGF-β receptor signaling. The genes that were significantly upregulated in MSCs are marked by red boxes, and those that were downregulated in MSCs are marked by green boxes. Abbreviations: IGF, insulin growth factor; MMP, matrix metalloproteinase; TGF, transforming growth factor. Other genes that were upregulated in MSCs but not in early fetal organs include genes involved in metabolism regulation: NQO1 (NADH dehydrogenase), LOXL2 (metabolism of arginine), PLOD2 (metabolism of lysine), SMURF2 (metabolism of tryptophan), NNMT (NAD metabolism), and NT5E (nucleotide and NAD metabolisms). To regulate intracellular signal transduction, MSC-upregulated genes included KDELR3, which is involved in both cAMP signaling and protein kinase A (cAMP-dependent protein kinase) signaling, and RGS4, which drives G proteins into their inactive GDP-bound forms (Table 4). Discussion The advantage of clinical applications of human mesenchymal stem cells (MSCs) is their high potential for self-renewal and the capacity to differentiate toward multiple cell types. MSCs retained normal karyotypes, morphology, and cell markers, although a recent study clearly revealed that some stromal cell-associated markers (CD13, CD29, CD44, CD63, CD73, CD90, and CD166) increased significantly at the transition from freshly isolated human adipose tissue-derived stromal vascular fraction cells to serial-passaged adipose-derived stem cells at passage 0 [36]. Results of this study, nevertheless, demonstrate that (a) gene expression profiles in individual human MSCs remain stable between passage 3 and passage 6 during in vitro culture, enduring cryopreservation and thawing well, (b) each group of MSCs derived from four different origins exhibits its signature expression profile, and (c) all human MSCs share a core MSC profile distinct from that of various fetal organs. We have previously reported that CB MSCs have a stronger osteogenic potential but a lower capacity for adipogenic differentiation than BM MSCs, suggesting that the disparate differentiation tendencies of MSCs from different sources should be considered in further applications [25]. In this study, we further analyzed the differences of transcriptomes in MSCs derived from four sources. Within each group of MSCs from the same origin, the variability of the gene expression levels is less than the variability between groups from different origins (Table 1). The findings agree with the notion that MSCs are multipotent precursors with low variability regardless of their donors [5]. In spite of the different profiles of individual groups of MSCs derived from different origins, the four MSC groups shared a set of core gene expression profiles (Tables 2, 4). It is intriguing that the most prominent functions of AF MSCs may be involved in uterine contraction and its related signaling transduction pathways, as suggested by the upregulation of oxytocin receptor (OXTR) and thrombin receptors (F2R [PAR1] and F2RL2 [PAR3]) (Table 3; Fig. 3A; supplemental online Table 2). Thrombin has been shown to be a potent uterotonic agonist, and its effects in the myometrium are mediated by intracellular signaling events that are comparable to those activated by oxytocin [37]. Thrombin, therefore, is suggested to be one of the initiators for uterine contraction in the presence of intrauterine hemorrhages. Given that the origin of AF MSCs is the fetus, our results, for the first time in the literature, suggest a mechanistic role of fetal MSCs in regulating uterine contraction. The results of functional networks (Fig. 3B; supplemental online Table 3) on the signature genes of AM MSCs (Table 3) allow us to appreciate the unique advantage that such analysis may have in the identification of additional gene products and their roles. In addition to pointing to the same functions (supplemental online Table 3), both networks (Fig. 3B) pointed to the involvement of endothelin, even with its absence in the list of Table 3. Since 1991, endothelins have been shown to be produced by the amniotic membrane [38–40]. Endothelin is a potent endothelial-derived vasoconstrictor and a sodium-regulating peptide [41]. Among the genes revealed in Figure 3B, endothelin, MME (also known as neutral endopeptidase or neprilysin), bradykinin receptor B1 (BDKRB1, also known as B1R), and atrial natriuretic peptide are involved in the regulation of vascular tone and electrolyte homeostasis [42]. Given that the amniotic membrane is the compartment lining of amniotic fluid, it is expected to play a role in the homeostasis of fluid and electrolytes, as the network analyses indicated (supplemental online Table 3). Endothelins have recently been linked to the pathogenesis of pre-eclampsia [43, 44]. However, speculation that the AM MSC endothelin may play a role in pre-eclampsia remains to be studied. The functional network analysis of signature genes of CB MSCs (Table 3; Fig. 3C) revealed that many of them might be part of the innate immune system against viruses and bacteria (supplemental online Table 4), representing the neonatal defense system against the earliest encountered pathogens. Of equal interest are the genes that were downregulated in CB MSCs but not in the other three MSCs (supplemental online Table 1). They included six homeobox genes (Hox-A9, -B6, -B7, -C6, and -C10 and PITX2). The products of the mammalian Hox genes control the anteroposterior pattern of parts in the hindbrain, neck, and trunk. Eliminating the function of a Hox gene in mice leads to a defect in the body region that corresponds to the domain of expression of that gene. Because all the CB MSCs in this study were derived from term babies, one would expect that Hox gene expression in CB MSCs was lower than that of other MSCs, including AF MSCs. Hox genes were recently found to be continuously expressed in adults to meet specific regional needs, especially in tissues that undergo frequent renewal [45]. Frequent renewal appears to be an important feature of fetal amniotic membrane and adult bone marrow, a view that is supported by our findings of a high expression of Hox-B6 in AM MSCs (Table 3) and upregulated Hox-A1, -A5, -C10, EN1, DLX5, and CART1 in BM MSCs (Table 3). Given that bone marrow is the ultimate source of myelo-lympho-hemopoietic systems, it is not a surprise that the biological processes of BM MSCs are involved in antigen processing via MHC class II, B-cell, and T-cell receptor signaling pathways (Fig. 3D; supplemental online Table 5A). In addition, echoing the recent concept of osteoimmunology [46], our results indicate that BM MSCs participate in skeletal development, the BMP signaling pathway, ossification, and bone mineralization (Fig. 3D; supplemental online Table 5B). Our network analyses also suggest that BM MSCs may be involved in dopamine biosynthesis, synaptic transmission, and regulation of neurotransmitter secretion (Fig. 3D; supplemental online Table 5B). If this is proved to be the case, our results would support the argument that bone marrow is a source of neural stem cells [47], although there is an alternative possibility that some bone marrow stroma are innervated and contain nerve terminals, which may be responsible for the expression of synapse-related proteins. As expected, many genes in the list of MSC core transcriptomes (Table 4) have previously been identified in the stroma cells isolated from adipose tissues and/or bone marrow. These genes included SERPINE1 (plasminogen activator inhibitor 1 [PAI-1]) [48], CD44, CD73 [36, 49], CD59 [50], annexin, and transgelin [51]. Our functional network analysis suggests that the regulation of extracellular matrix and adhesion is a prominent feature of all MSCs (Fig. 4), in agreement with the notion that the adhesion to the extracellular matrix is the main determinant of differentiation of stromal stem cells to somatic mesenchymal cells [52]. SERPINE1 can bind to the urokinase-type plasminogen activator (PLAU) and its receptor (PLAUR), and the complex can further interact with vitronectin and different integrins [53, 54]. Plasmin can cleave vitronectin, lamin, and fibronectin, which will degrade extracellular matrices (Fig. 4). It has been proposed that the endocytic clearance of the complexes of integrins, PLAU, PLAUR, and SERPINE1 can lead to the disengagement of integrins from extracellular matrix and cell detachment [55]. Through interactions with SERPINE1, plasmin, Cav2, Smurf, and Smad, the functional network analysis reveals TGF-β receptor signaling to be a common feature of all MSCs (Fig. 4). Depending on the differentiation stage of the target cell, the local environment, and the concentration and isoform of TGF-β, in vivo or in vitro, TGF-β can exert opposite functions in cell proliferation, apoptosis, and differentiation and can inhibit or increase terminally differentiated cell functions [56]. In vitro, TGF-β is considered to be a major regulator of stem cells quiescence [56]. Our finding that TGF-β is more highly expressed in all MSCs than in fetal organs implies an important role of TGF-β in MSCs, although its functions may differ drastically depending on the context, that is, in vitro culture versus in vivo residency in the niche for MSCs. CD44, a cell-surface glycoprotein that serves as a receptor for hyaluronic acid, is involved in cell-cell interactions, cell adhesion and migration, and interaction with other ligands, such as osteopontin, collagens, and matrix metalloproteinases. CD44, which is considered a marker of MSCs, was highly expressed in all MSCs used in this study (Fig. 2B; Table 4). CD44 participates in a wide variety of cellular functions, including lymphocyte activation, recirculation and homing, hematopoiesis, and tumor metastasis. CD44 has been shown to be critical for the recruitment of MSCs into wound sites for tissue regeneration, as well as for migration of fibroblast progenitors to allografts in the development of graft fibrosis [57]. Targeting CD44 as one way to eradicate cancer stem cells has been recently shown in two studies. BCR-ABL-expressing chronic myeloid leukemia (CML) stem cells depend much more on CD44 for homing and engraftment than normal hematopoietic stem cells do, and the CD44 blockade may be beneficial in autologous transplantation in CML [58]. Acute myeloid leukemia (AML) leukemic stem cells (LSCs) require CD44 to interact with a niche to maintain their stem cell properties, and this requirement provides a therapeutic strategy to eliminate quiescent AML LSCs and may be applicable to other types of cancer stem cells [59]. When DNA microarrays are used to analyze the similar tissues, gene expression profiles obtained from different studies have been notoriously varied, even sometimes conflicting. Possible causes for these discrepancies include different assay platforms, nonuniform coverage of gene sets, distinct data filtering strategies, various degrees of statistical stringency, and data complexity and variability [60–62]. Nevertheless, using different microarray platforms, this study with the Affymetrix U133A system identified some upregulated genes in MSCs that were confirmed by another study using the Amersham CodeLink system [21] (Amersham Biosciences, Piscataway, NJ, http://www.amersham.com). Jeong et al. compared CB MSCs with cord blood-derived mononuclear cells and identified 47 genes that were differentially up-expressed in CB MSCs [21]. Among their 47 genes, seven were also noted in core expression profiles that were more upregulated in MSCs than in early fetal organs: SERPINE1, CTGF, TAGLN, NNMT, CYR61, LOX, and KDELR3 (Table 4). Both SERPINE1 (PAI-1) and CTGF (connective tissue growth factor) interact with low-density lipoprotein related protein-1 (LRP1) [63, 64]. Notably, the internalization of the PLAU/ PLAU/integrin complex by SERPINE1 (Fig. 4) is an LRP1-dependent process [55], whereas LRP1 interacts with human Frizzled-1 (the Wnt receptor) and downregulates the canonical Wnt signaling pathway that triggers cell proliferation, oncogenic transformation, and inhibition of apoptosis [65]. The Wnt signaling pathway can further regulate CYR61 (cysteine-rich angiogenic inducer 61) and CTGF [66]. By interrogating these genes through network analysis, we recognize the roles of LRP1 and the Wnt signaling pathway in MSCs. The Wnt signaling pathway plays a central role in development and homeostasis [65]. Lysyl oxidase (LOX) catalyzes cross-linking of elastin and collagens and accounts for their integrity [67]. Transgelin (TAGLN) is frequently used as a smooth muscle marker [68], and the loss of transgelin gene expression may be an important early event in tumor progression and can be used as a diagnostic marker for breast and colon cancer development [69]. KDELR3 is responsible for the retention of resident soluble proteins in the lumen of the endoplasmic reticulum [70]. However, the exact roles of LOX, TAGLN, and KDELR3 in MSCs remain unclear. In conclusion, functional network analyses in this study were proven useful for providing insights into the transcriptomes of MSCs derived from different origins and the core signature profiles of all MSCs. AF MSCs may initiate the interaction with the uterus by upregulating the oxytocin and thrombin receptors. AM MSCs may play a role in the homeostasis of fluid and electrolytes by regulating networks of endothelin, neprilysin, bradykinin receptor, and atrial natriuretic peptide. As the neonatal defense system against the earliest encountered pathogens, CB MSCs may be involved in the innate immune system. BM MSCs may be important sources not only for all blood lineages but also for bone formation. When compared with fetal organs, the core signature transcriptomes of all MSCs may involve in the regulation of extracellular matrix and adhesion, TGF-β receptor signaling, and the Wnt signaling pathway. Disclosure of Potential Conflicts of Interest The authors indicate no potential conflicts of interest. Acknowledgements We thank En-Shih Chen, Hsiao-Wen Lu, Yu-Chih Lin, Pei-Cih Wei, Yi-Jun Lin, and Ching-Ling Wang (Chang Gung Memorial Hospital) for technical assistance and Shihtien T. Wang (Children's Hospital of Wisconsin, Milwaukee, WI) for English editing. This study was supported by grants 94-EC-17-A-17-R7-0525 (S.-M.H.) from the Ministry of Economic Affairs, NSC 94-2314-B-281-004 (M.-S.T.), NSC 94-2314-B-182A-137 (S.-D.C.), and NSC 94-2314-B-182-050 (H.-S.W.) from National Science Council, Taiwan, and grants CMRPG1013-III (S.-D.C.), CMRPG340461 (T.-H.W.), and CMRPG340291 (H.-S.W.) from Chang Gung Memorial Hospital. M.-S.T. and S.-M.H. contributed equally to this work. References 1 Mannello F , Tonti GAM, Bagnara GP et al. Role and function of matrix metalloproteinases in the differentiation and biological characterization of mesenchymal stem cells . Stem Cells 2006 ; 24 : 475 – 481 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Laflamme MA , Murry CE. Regenerating the heart . Nat Biotechnol 2005 ; 23 : 845 – 856 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Prockop DJ , Gregory CA, Spees JL. One strategy for cell and gene therapy: Harnessing the power of adult stem cells to repair tissues . Proc Natl Acad Sci U S A 2003 ; 100 : 11917 – 11923 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Meirelles L , Chagastelles PC, Nardi NB. Mesenchymal stem cells reside in virtually all post-natal organs and tissues . J Cell Sci 2006 ; 119 : 2204 – 2213 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Pittenger MF , Mackay AM, Beck SC et al. Multilineage potential of adult human mesenchymal stem cells . Science 1999 ; 284 : 143 – 147 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Zuk PA , Zhu M, Ashjian P et al. Human adipose tissue is a source of multipotent stem cells . Mol Biol Cell 2002 ; 13 : 4279 – 4295 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Xu Y , Malladi P, Wagner DR et al. Adipose-derived mesenchymal cells as a potential cell source for skeletal regeneration . Curr Opin Mol Ther 2005 ; 7 : 300 – 305 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 8 Bieback K , Kern S, Kluter H et al. Critical parameters for the isolation of mesenchymal stem cells from umbilical cord blood . Stem Cells 2004 ; 22 : 625 – 634 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Tsai MS , Lee JL, Chang YJ et al. Isolation of human multipotent mesenchymal stem cells from second-trimester amniotic fluid using a novel two-stage culture protocol . Hum Reprod 2004 ; 19 : 1450 – 1456 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Tsai MS , Hwang SM, Tsai YL et al. Clonal amniotic fluid-derived stem cells express characteristics of both mesenchymal and neural stem cells . Biol Reprod 2006 ; 74 : 545 – 551 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Zhao P , Ise H, Hongo M et al. Human amniotic mesenchymal cells have some characteristics of cardiomyocytes . Transplantation 2005 ; 79 : 528 – 535 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Ochsenbein-Kolble N , Bilic G, Hall H et al. Inducing proliferation of human amnion epithelial and mesenchymal cells for prospective engineering of membrane repair . J Perinat Med 2003 ; 31 : 287 – 294 . Google Scholar Crossref Search ADS PubMed WorldCat 13 In 't Anker PS , Scherjon SA, Kleijburg-van der Keur C et al. Isolation of mesenchymal stem cells of fetal or maternal origin from human placenta . Stem Cells 2004 ; 22 : 1338 – 1345 . Google Scholar Crossref Search ADS PubMed WorldCat 14 Portmann-Lanz CB , Schoeberlein A, Huber A et al. Placental mesenchymal stem cells as potential autologous graft for pre- and perinatal neuroregeneration . Am J Obstet Gynecol 2006 ; 194 : 664 – 673 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Kim JW , Kim SY, Park SY et al. Mesenchymal progenitor cells in the human umbilical cord . Ann Hematol 2004 ; 83 : 733 – 738 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Fu YS , Cheng YC, Lin MY et al. Conversion of human umbilical cord mesenchymal stem cells in Wharton's jelly to dopaminergic neurons in vitro: Potential therapeutic application for Parkinsonism . Stem Cells 2006 ; 24 : 115 – 124 . Google Scholar Crossref Search ADS PubMed WorldCat 17 Bianco P , Riminucci M, Gronthos S et al. Bone marrow stromal stem cells: Nature, biology, and potential applications . Stem Cells 2001 ; 19 : 180 – 192 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Grove JE , Bruscia E, Krause DS. Plasticity of bone marrow-derived stem cells . Stem Cells 2004 ; 22 : 487 – 500 . Google Scholar Crossref Search ADS PubMed WorldCat 19 Baksh D , Song L, Tuan RS. Adult mesenchymal stem cells: Characterization, differentiation, and application in cell and gene therapy . J Cell Mol Med 2004 ; 8 : 301 – 316 . Google Scholar Crossref Search ADS PubMed WorldCat 20 Lee OK , Kuo TK, Chen WM et al. Isolation of multipotent mesenchymal stem cells from umbilical cord blood . Blood 2004 ; 103 : 1669 – 1675 . Google Scholar Crossref Search ADS PubMed WorldCat 21 Jeong JA , Hong SH, Gang EJ et al. Differential gene expression profiling of human umbilical cord blood-derived mesenchymal stem cells by DNA microarray . Stem Cells 2005 ; 23 : 584 – 593 . Google Scholar Crossref Search ADS PubMed WorldCat 22 Phinney DG , Hill K, Michelson C et al. Biological activities encoded by the murine mesenchymal stem cell transcriptome provide a basis for their developmental potential and broad therapeutic efficacy . Stem Cells 2006 ; 24 : 186 – 198 . Google Scholar Crossref Search ADS PubMed WorldCat 23 Aiba K , Sharov AA, Carter MG et al. Defining a developmental path to neural fate by global expression profiling of mouse embryonic stem cells and adult neural stem/progenitor cells . Stem Cells 2006 ; 24 : 889 – 895 . Google Scholar Crossref Search ADS PubMed WorldCat 24 Jaatinen T , Hemmoranta H, Hautaniemi S et al. Global gene expression profile of human cord blood-derived CD133+ cells . Stem Cells 2006 ; 24 : 631 – 641 . Google Scholar Crossref Search ADS PubMed WorldCat 25 Chang YJ , Shih DT, Tseng CP et al. Disparate mesenchyme-lineage tendencies in mesenchymal stem cells from human bone marrow and umbilical cord blood . Stem Cells 2006 ; 24 : 679 – 685 . Google Scholar Crossref Search ADS PubMed WorldCat 26 Dominici M , Le Blanc K, Mueller I et al. Minimal criteria for defining multipotent mesenchymal stromal cells. The International Society for Cellular Therapy position statement . Cytotherapy 2006 ; 8 : 315 – 317 . Google Scholar Crossref Search ADS PubMed WorldCat 27 Wang TH , Lee YS, Chen ES et al. Establishment of cDNA microarray analysis at the Genomic Medicine Research Core Laboratory (GMRCL) of Chang Gung Memorial Hospital . Chang Gung Med J 2004 ; 27 : 243 – 260 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 28 Lee YS , Chen CH, Chao A et al. Molecular signature of clinical severity in recovering patients with severe acute respiratory syndrome coronavirus (SARS-CoV) . BMC Genomics 2005 ; 6 : 132 . Google Scholar Crossref Search ADS PubMed WorldCat 29 Chao A , Wang TH, Lee YS et al. Molecular characterization of adenocarcinoma and squamous carcinoma of the uterine cervix using microarray analysis of gene expression . Int J Cancer 2006 ; 119 : 91 – 98 . Google Scholar Crossref Search ADS PubMed WorldCat 30 Wang TH , Chan YH, Chen CW et al. Paclitaxel (Taxol) upregulates expression of functional interleukin-6 in human ovarian cancer cells through multiple signaling pathways . Oncogene 2006 ; 25 : 4857 – 4866 . Google Scholar Crossref Search ADS PubMed WorldCat 31 Nikolsky Y , Ekins S, Nikolskaya T et al. A novel method for generation of signature networks as biomarkers from complex high throughput data . Toxicol Lett 2005 ; 158 : 20 – 29 . Google Scholar Crossref Search ADS PubMed WorldCat 32 Mason CW , Swaan PW, Weiner CP. Identification of interactive gene networks: A novel approach in gene array profiling of myometrial events during guinea pig pregnancy . Am J Obstet Gynecol 2006 ; 194 : 1513 – 1523 . Google Scholar Crossref Search ADS PubMed WorldCat 33 Shi L , Reid LH, Jones WD et al. The MicroArray Quality Control (MAQC) project shows inter- and intraplatform reproducibility of gene expression measurements . Nat Biotechnol 2006 ; 24 : 1151 – 1161 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 34 Canales RD , Luo Y, Willey JC et al. Evaluation of DNA microarray results with quantitative gene expression platforms . Nat Biotechnol 2006 ; 24 : 1115 – 1122 . Google Scholar Crossref Search ADS PubMed WorldCat 35 Guo L , Lobenhofer EK, Wang C et al. Rat toxicogenomic study reveals analytical consistency across microarray platforms . Nat Biotechnol 2006 ; 24 : 1162 – 1169 . Google Scholar Crossref Search ADS PubMed WorldCat 36 Mitchell JB , McIntosh K, Zvonic S et al. Immunophenotype of human adipose-derived cells: Temporal changes in stromal-associated and stem cell-associated markers . Stem Cells 2006 ; 24 : 376 – 385 . Google Scholar Crossref Search ADS PubMed WorldCat 37 Elovitz MA , Ascher-Landsberg J, Saunders T et al. The mechanisms underlying the stimulatory effects of thrombin on myometrial smooth muscle . Am J Obstet Gynecol 2000 ; 183 : 674 – 681 . Google Scholar Crossref Search ADS PubMed WorldCat 38 Casey ML , Word RA, MacDonald PC. Endothelin-1 gene expression and regulation of endothelin mRNA and protein biosynthesis in avascular human amnion. Potential source of amniotic fluid endothelin . J Biol Chem 1991 ; 266 : 5762 – 5768 . Google Scholar Crossref Search ADS PubMed WorldCat 39 Marvin KW , Keelan JA, Eykholt RL et al. Expression of angiogenic and neurotrophic factors in the human amnion and choriodecidua . Am J Obstet Gynecol 2002 ; 187 : 728 – 734 . Google Scholar Crossref Search ADS PubMed WorldCat 40 Fried G , Sand A, Ostlund E et al. Endothelin-1 and macrophage colony-stimulating factor are co-localized in human amnion membrane cells and secreted into amniotic fluid . Mol Hum Reprod 2003 ; 9 : 719 – 724 . Google Scholar Crossref Search ADS PubMed WorldCat 41 Prabhakar SS . Regulatory and functional interaction of vasoactive factors in the kidney and extracellular pH . Kidney Int 2004 ; 66 : 1742 – 1754 . Google Scholar Crossref Search ADS PubMed WorldCat 42 Quaschning T , Galle J, Wanner C. Vasopeptidase inhibition: A new treatment approach for endothelial dysfunction. Kidney Int Suppl 2003 ; S54 – 57 . Google Scholar OpenURL Placeholder Text WorldCat 43 Fiore G , Florio P, Micheli L et al. Endothelin-1 triggers placental oxidative stress pathways: Putative role in preeclampsia . J Clin Endocrinol Metab 2005 ; 90 : 4205 – 4210 . Google Scholar Crossref Search ADS PubMed WorldCat 44 LaMarca BB , Cockrell K, Sullivan E et al. Role of endothelin in mediating tumor necrosis factor-induced hypertension in pregnant rats . Hypertension 2005 ; 46 : 82 – 86 . Google Scholar Crossref Search ADS PubMed WorldCat 45 Morgan R . Hox genes: A continuation of embryonic patterning? Trends Genet 2006 ; 22 : 67 – 69 Epub 2005 Dec 2001. Google Scholar Crossref Search ADS PubMed WorldCat 46 Walsh MC , Kim N, Kadono Y et al. Osteoimmunology: Interplay between the immune system and bone metabolism . Annu Rev Immunol 2006 ; 24 : 33 – 63 . Google Scholar Crossref Search ADS PubMed WorldCat 47 Rice CM , Scolding NJ. Adult stem cells—Reprogramming neurological repair? Lancet 2004 ; 364 : 193 – 199 . Google Scholar Crossref Search ADS PubMed WorldCat 48 Bastelica D , Morange P, Berthet B et al. Stromal cells are the main plasminogen activator inhibitor-1-producing cells in human fat: Evidence of differences between visceral and subcutaneous deposits . Arterioscler Thromb Vasc Biol 2002 ; 22 : 173 – 178 . Google Scholar Crossref Search ADS PubMed WorldCat 49 Alviano F , Fossati V, Marchionni C et al. Term amniotic membrane is a high throughput source for multipotent mesenchymal stem cells with the ability to differentiate into endothelial cells in vitro . BMC Dev Biol 2007 ; 7 : 11 . Google Scholar Crossref Search ADS PubMed WorldCat 50 Gronthos S , Franklin DM, Leddy HA et al. Surface protein characterization of human adipose tissue-derived stromal cells . J Cell Physiol 2001 ; 189 : 54 – 63 . Google Scholar Crossref Search ADS PubMed WorldCat 51 Panepucci RA , Siufi JL, Silva WA Jr et al. Comparison of gene expression of umbilical cord vein and bone marrow-derived mesenchymal stem cells . Stem Cells 2004 ; 22 : 1263 – 1278 . Google Scholar Crossref Search ADS PubMed WorldCat 52 Gartner S , Kaplan HS. Long-term culture of human bone marrow cells . Proc Natl Acad Sci U S A 1980 ; 77 : 4756 – 4759 . Google Scholar Crossref Search ADS PubMed WorldCat 53 Stefansson S , Lawrence DA. The serpin PAI-1 inhibits cell migration by blocking integrin alpha V beta 3 binding to vitronectin . Nature 1996 ; 383 : 441 – 443 . Google Scholar Crossref Search ADS PubMed WorldCat 54 Okumura Y , Kamikubo Y, Curriden SA et al. Kinetic analysis of the interaction between vitronectin and the urokinase receptor . J Biol Chem 2002 ; 277 : 9395 – 9404 . Google Scholar Crossref Search ADS PubMed WorldCat 55 Czekay RP , Aertgeerts K, Curriden SA et al. Plasminogen activator inhibitor-1 detaches cells from extracellular matrices by inactivating integrins . J Cell Biol 2003 ; 160 : 781 – 791 . Google Scholar Crossref Search ADS PubMed WorldCat 56 Ruscetti FW , Akel S, Bartelmez SH. Autocrine transforming growth factor-beta regulation of hematopoiesis: Many outcomes that depend on the context . Oncogene 2005 ; 24 : 5751 – 5763 . Google Scholar Crossref Search ADS PubMed WorldCat 57 Zhu H , Mitsuhashi N, Klein A et al. The role of the hyaluronan receptor CD44 in mesenchymal stem cell migration in the extracellular matrix . Stem Cells 2006 ; 24 : 928 – 935 . Google Scholar Crossref Search ADS PubMed WorldCat 58 Krause DS , Lazarides K, von Andrian UH et al. Requirement for CD44 in homing and engraftment of BCR-ABL-expressing leukemic stem cells . Nat Med 2006 ; 24 : 24 . Google Scholar OpenURL Placeholder Text WorldCat 59 Jin L , Hope KJ, Zhai Q et al. Targeting of CD44 eradicates human acute myeloid leukemic stem cells . Nat Med 2006 ; 24 : 24 . Google Scholar OpenURL Placeholder Text WorldCat 60 Yauk CL , Berndt ML, Williams A et al. Comprehensive comparison of six microarray technologies . Nucleic Acids Res 2004 ; 32 : e124 . Google Scholar Crossref Search ADS PubMed WorldCat 61 Carter SL , Eklund AC, Mecham BH et al. Redefinition of Affymetrix probe sets by sequence overlap with cDNA microarray probes reduces cross-platform inconsistencies in cancer-associated gene expression measurements . BMC Bioinformatics 2005 ; 6 : 107 . Google Scholar Crossref Search ADS PubMed WorldCat 62 Wang H , He X, Band M et al. A study of inter-lab and inter-platform agreement of DNA microarray data . BMC Genomics 2005 ; 6 : 71 . Google Scholar Crossref Search ADS PubMed WorldCat 63 Degryse B , Neels JG, Czekay RP et al. The low density lipoprotein receptor-related protein is a motogenic receptor for plasminogen activator inhibitor-1 . J Biol Chem 2004 ; 279 : 22595 – 22604 . Google Scholar Crossref Search ADS PubMed WorldCat 64 Gao R , Brigstock DR. Low density lipoprotein receptor-related protein (LRP) is a heparin-dependent adhesion receptor for connective tissue growth factor (CTGF) in rat activated hepatic stellate cells . Hepatol Res 2003 ; 27 : 214 – 220 . Google Scholar Crossref Search ADS PubMed WorldCat 65 Zilberberg A , Yaniv A, Gazit A. The low density lipoprotein receptor-1, LRP1, interacts with the human frizzled-1 (HFz1) and down-regulates the canonical Wnt signaling pathway . J Biol Chem 2004 ; 279 : 17535 – 17542 . Google Scholar Crossref Search ADS PubMed WorldCat 66 Si W , Kang Q, Luu HH et al. CCN1/Cyr61 is regulated by the canonical Wnt signal and plays an important role in Wnt3A-induced osteoblast differentiation of mesenchymal stem cells . Mol Cell Biol 2006 ; 26 : 2955 – 2964 . Google Scholar Crossref Search ADS PubMed WorldCat 67 Maki JM , Sormunen R, Lippo S et al. Lysyl oxidase is essential for normal development and function of the respiratory system and for the integrity of elastic and collagen fibers in various tissues . Am J Pathol 2005 ; 167 : 927 – 936 . Google Scholar Crossref Search ADS PubMed WorldCat 68 Miyata T , Iizasa H, Sai Y et al. Platelet-derived growth factor-BB (PDGF-BB) induces differentiation of bone marrow endothelial progenitor cell-derived cell line TR-BME2 into mural cells, and changes the phenotype . J Cell Physiol 2005 ; 204 : 948 – 955 . Google Scholar Crossref Search ADS PubMed WorldCat 69 Shields JM , Rogers-Graham K, Der CJ. Loss of transgelin in breast and colon tumors and in RIE-1 cells by Ras deregulation of gene expression through Raf-independent pathways . J Biol Chem 2002 ; 277 : 9790 – 9799 . Google Scholar Crossref Search ADS PubMed WorldCat 70 Pelham HR . The dynamic organisation of the secretory pathway . Cell Struct Funct 1996 ; 21 : 413 – 419 . Google Scholar Crossref Search ADS PubMed WorldCat Copyright © 2007 AlphaMed Press This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Functional Network Analysis of the Transcriptomes of Mesenchymal Stem Cells Derived from Amniotic Fluid, Amniotic Membrane, Cord Blood, and Bone Marrow
JO - Stem Cells
DO - 10.1634/stemcells.2007-0023
DA - 2007-10-01
UR - https://www.deepdyve.com/lp/oxford-university-press/functional-network-analysis-of-the-transcriptomes-of-mesenchymal-stem-gSaioTmWZy
SP - 2511
EP - 2523
VL - 25
IS - 10
DP - DeepDyve
ER -