Gene expression modifications in Wharton’s Jelly mesenchymal stem cells promoted by prolonged in vitro culturing

It has been demonstrated that the umbilical cord matrix, represented by the Wharton’s Jelly (WJ), contains a great number of mesenchymal stem cells (MSCs), characterized by the expression of specific MSCs markers, shared by both human and animal models. The easy access to massive WJ amount makes it an attractive source of MSCs for cell-based therapies. However, as in other stem cell models, a deeper investigation of WJ-derived MSCs (WJ-MSCs) biological properties, probably modulated by their prolonged expansion and fast growth abilities, is required before their use in clinical settings. In this context, in order to analyze specific gene expression modifications occurring in WJ-MSCs, along with their culture prolongation, we investigated the transcriptomic profiles of WJ-MSCs after 4 and 12 passages of in vitro expansion by microarray analysis. Hierarchical clustering analysis of the data set originated from a total of 6 experiments revealed that in vitro expansion of WJ-MSCs up to 12 passages promote selective over-expression of 157 genes and down-regulation of 440 genes compared to the 4th passage. IPA software analysis of the biological functions related to the identified sets of genes disclosed several transcripts related to inflammatory and cell stress response, cell proliferation and maturation, and apoptosis. Taken together, these modifications may lead to an impairment of both cell expansion ability and resistance to apoptosis, two hallmarks of aging cells. In conclusion, results provided by the present study suggest the need to develop novel culture protocols able to preserve stem cell plasticity.


Background
Mesenchymal stem cells (MSCs) are generally considered the main toolbox for cell-based therapies. Compared to embryonic stem cells (ESCs), MSCs exhibit the following advantages: accessibility with fewer ethical controversies [1], no reports of teratoma formation after transplantation, and versatile therapeutic applications [2][3][4][5][6][7]. Based on literature data, the phenotype of MSCs obtained from different sources is typically characterized by the expression of CD44, CD73, CD90 and CD105, representing the best suited markers currently used to characterize these cells, together with the lack of the expression of endothelial/hematopoietic markers (CD144, CD34, CD45) [8,9]. The umbilical cord (UC) is an extraembryonic formation essential to provide feeding to the fetus during the intrauterine development. It has been shown that UC matrix, represented by the Wharton's Jelly (WJ), surrounding umbilical vessels, contains a great number of mesenchymal cells, which have been characterized as expressing aforementioned markers, shared by MSCs in both human and animal models [10][11][12]. The abundant amount of WJ makes it an attractive source of MSCs for cell-based therapies [13,14]. However, as in other stem cell models, a deep investigation of WJ-MSCs biological properties is required before their use in clinical settings. In this context, a critical point is represented by the analysis of functional modifications affecting WJ-MSCs along with their prolonged in vitro cultures. In fact, a recent study highlighted the changes in protein expression profiling, along with the in vitro expansion of WJ-MSCs, probably related to the gradual impairment of their stem cell plasticity and of the biological mechanisms occurring in cellular aging [15]. In order to provide a different investigation model of the biological modifications occurring during WJ-MSCs in vitro growth, we analyzed the transcriptomic profile of the aforementioned cells following prolonged culture times [12 th passage compared to an early (4 th ) passage] by microarray analysis. The aim of the present study was to identify possible novel markers related to their in vitro prolonged expansion and to their fast growth abilities.

Cell isolation and culture
Institutional review board approval was obtained for all cell culture procedures. Fresh human UC (N = 5) were obtained from full-term births, after written informed consent was obtained from parents. UC were aseptically stored in sterile saline solution and processed within 6 hours from the partum to obtain WJ-MSCs, as previously described [15]. Briefly, after the removal of blood vessels, the extracellular matrix of WJ was scraped off, treated with 2 mg/ml collagenase IV (Sigma) for 16 hours at 37°C and then with 2.5% trypsin for 30 minutes at 37°C, under agitation. Finally, the obtained cell suspension was seeded in complete Human mesenchymal stem cell growth medium (hMSCGM, Lonza) and cultured in 5% CO 2 in a 37°C incubator. When 80% of confluence was reached, the adherent fraction of cells was detached with 0.05% trypsin-EDTA, counted by Trypan Blue exclusion test, and reseeded at 3000 cells/cm 2 to reach the 90% of confluence after 3-4 population doublings.

Doubling time and cell cycle analyses by bromodeoxyuridine incorporation assay
Exponentially growing WJ-MSCs were exposed to 10 μM bromodeoxyuridine (BrdU) (Sigma, St. Louis, MO, USA) for 1 h, then fixed in 70% ethanol and kept at 4°C before labeling as previously described [17]. To detect BrdU incorporation, cells were washed with PBS and treated with 1 ml of a solution containing 2 N HCl/0.5% Triton X-100 (Sigma) for 30 min at room temperature. 1 ml per sample of 0.1 M Na 2 B 4 O 7 (pH 8.57) was added to stop the HCl reaction. Cells were then washed with 1 ml of a solution containing 0.5% Triton X-100/1% BSA, followed by an incubation for 30 min at room temperature in the dark with fluorescein isothiocyanate (FITC)-conjugated anti-BrdU antibody (BD Biosciences, San Jose, CA; dilution: 1:5 in 0.5% v/v Triton X-100). Cells were washed and resuspended in a solution containing 5 μg/ml Propidium Iodide (PI, Sigma) and 200 μg/ml RNase (Sigma). After 30 min of incubation biparametric BrdU/DNA data were acquired on a FACSCalibur flow cytometer (two-lasers, four-color configuration) with CellQuest 3.2.1.f1 (BD) software; data were analysed using FlowJo™ software (TreeStar, Ashland, OR) or ModFit LT™ software (Verity Software House, Toshan, ME, USA). Debris was excluded from the analysis by gating a forward scatter versus side scatter plot. Cell aggregates were excluded by gating FL2 area versus FL2 width [17].

Telomere length assay
Genomic DNA was extracted from WJ-MSC at different passages using Wizard Genomic DNA Purification Kit (Promega) following the manufacturer's instructions. The length of telomere regions was assessed using the Telo TAGGG kit (Roche) according to the manufacturer's instructions. Appropriate controls, represented by DNA extracted from cells with long or short telomere regions, were also provided with the kit [15].

Determination of cell senescence
The amount of senescent cells was evaluated in the different reported conditions by using the Senescence β-Galactosidase Staining Kit (Abcam, Cambridge, UK) in accordance to the manufacturer's instructions, as previously described [15].

Expression profiling
Total RNA was extracted from about 10 6 cells/sample of two different WJ-MSCs cultures after 4 and 12 passages during their in vitro expansion, using the SVtotal RNA Izolation System kit (Promega, Madison, WI, USA). The purity and quantity of RNA was assessed using the Agilent 8453 Spectrophotometer (Agilent, Santa Clara, CA, USA). RNA quality was determined by both the evaluation of the rRNA band integrity, using agarose electrophoresis, and absorption readings at 260 and 280 nm. Extracted RNA was linearly amplified using the Amino AllylMessageAmp™ II aRNA Amplification Kit (Ambion, Austin, TX, USA). Five to ten μg of amplified aRNA were fluorescently labeled with Cy3-Cy5 cyanins and then hybridized on high-density array Human Whole Genome OneArray™ Microarray V5 (30,968 total probe; Biosense, Italy). Amplified aRNAs were used for microarray experiments carried out by hybridization of WJ-MSCs after 4 passages, compared to WJ-MSCs after 12 passages, for a total of 6 experiments ( Table 1). The same biological samples have been compared at 4th and 12th passage. After hybridization, Cy3-Cy5 fluorescent signals were captured by a Confocal Laser Scanner "ScanArray Express" (Packard BioScience) and analyzed using the software "ScanArray Express-MicroArray Analysis System" version 3.0 (Perkin Elmer). Raw data of the performed experiments were recorded in the GEO public database (accession number: GSE34929). The values of the median signal intensity from each spot were subtracted from the local median background intensity. For each slide, after local background subtraction, a LOWESS algorithm was used for row data normalization, to evaluate signal to noise ratio and generate log ratios of sample vs reference signal. A gene was considered to be differentially expressed when showing an absolute value of log-ratio higher or equal to ± 0.5, an index that translates to a fold-change of 1.4 in transcript quantity. Analysis of data obtained by microarray experiments was carried out by means of hierarchical gene clustering [18] using Cluster 3.0 (open source 2006) and TreeView (Stanford University Labs) software. In order to include in clustering analysis only well measured transcripts, we selected spots with a present call (identified transcripts with measurable expression) in at least 80% of experiments and being > 1.7 fold up-or down regulated in at least 5 experiments. Identified clusters were then analyzed by the Ingenuity Pathways Analysis (IPA) software (Ingenuity Systems, Redwood City, CA), in order to classify genes based on their biological functions and disclose functional networks connecting specific genes. IPA infers and ranks networks by a score, expressed as a numerical value, which is a probabilistic fit between the amount of focus genes that are potentially eligible for a network composition and present on a given gene list, the size of the network, as well as all the molecules present in the Ingenuity Knowledge Base that can be part of such a network. To validate the microarray results, qRT-PCR analysis was performed on three down-regulated (p53, HSPE1 and HIST1H3C) and three up-regulated genes (IL1B, CREBBP and LYN) as evidenced by microarray experiments, using the housekeeping gene GAPDH as internal control to normalize the relative expressions of target genes. The quantitative RT-PCR was carried out in a total volume of 50 μl containing 1× TaqMan Universal PCR Master mix, no AmpErase UNG and 2 μl of cDNA using the TaqMan assay on the Abi 7900HT Sequencing Detection System. Genes primers and probe sets used were NM_000546 (p53), NM_002157 (HSPE1), NM_003531 (HIST1H3C), NM_004380 (CREBBP), NM_000877 (IL1B), NM_002350 (LYN), NM_002046 (GAPDH) (Integrated DNA Technologies, Coralville, Iowa, USA). The real time amplifications included 10 minutes at 95°C, followed by 48 cycles of 15 seconds at 95°C and 1 minute at 60°C. Relative expression levels were calculated for each sample after normalization against the housekeeping gene GAPDH, using the ΔΔCt method for comparing relative fold expression differences [19].

Cells isolation, culturing and characterization
Cells isolated from the human WJ displayed a consistent spindle-shaped elongated fibroblast-like morphology at the 4 th passage, a feature retained up to the 12 th passage ( Figure 1A, B). A representative immunophenotype of cells used in our experiments is reported in Additional file 1: Figure S1. A positive response pattern expression of CD13, CD29, CD44, CD73, CD90, CD105, CD146, CD166, HLA-ABC markers and a negative reactivity for CD14, CD34, CD45, CD117, CD133, CD144, CD326, HLA-DR were detected at both 4 th and 12 th passages (Table 1). Such a high homogeneous marker expression suggests that nonstem cell populations did not significantly contaminate samples used in the present study. On the other hand, the expression of CD13, CD44 and CD73 underwent a progressive and statistically significant reduction at the 12 th passage (Table 1) while the variation of the adhesion molecule expression (CD14, CD44 and CD73) is in accordance with a stem cell aging process, during in vitro expansion. The steady ability of WJ-MSCs to differentiate into both adipogenic and osteogenic lineages at the 4 th and the 12 th passage was evidenced ( Figure 1). After their expansion, cells showed a homogeneous diploid content during the G1 cell-cycle phase (Additional file 2: Figure S2). Furthermore, G1 and G2 cell-cycle checkpoints appeared intact. This finding is consistent with actively cycling cells. The long telomeric end of DNA extracted from WJ-MSCs at all examined passages also confirms that these cells preserve their capability to undergo a high number of cellular divisions up to the 12 th passage (data not shown) [15]. Accordingly, a low frequency of cells staining positive for βgalactosidase was found at both studied passages (4 th and 12 th ). On the other hand, long-term in vitro culture passages led to an impairment of cell expansion ability, as demonstrated by different exponential curves of growth at the 4 th and the 12 th passage, as demonstrated by cell counts ( Figure 1C). These data were confirmed by the analysis of the doubling time obtained through the BrdU incorporation assay, associated to the DNA staining ( Figure 1D): results evidenced that the doubling time progressively increased from 32 h at the 4 th passage to 74 h at the 12 th passage (p < 0.001).

Gene expression profile of WJ-MSCs at the 4 th passage vs the 12 th passage
To gain insights on the global changes in gene expression of human WJ-MSCs, produced by 12 passages in vitro expansion as compared to 4 passages, we performed a hierarchical clustering analysis of the data set originated from a total of 6 experiments (3 biological and 3 technical replicates). On a total of 30,968 transcripts investigated by the array, the analysis revealed that 12 passages in vitro expansion of WJ-MSCs promote the selective over-expression of 157 genes (cluster 1), while 440 genes were downregulated (cluster 2) as compared to 4 passages expansion cells ( Figure 2; Additional file 3: Table S1; Additional file 4: Table S2). Ingenuity Pathway Analysis (IPA) was carried out to investigate the main functions played by the selected clusters of genes. The up-regulated gene dataset (n = 157) was mainly composed by genes involved in functions regarding cellular development, cellular growth and proliferation, cellular movement, cell death, cellular assembly and organization, gene expression, cancer, cellular compromise, nervous system development and function, cell cycle, cell morphology, post-translational modifications and RNA post-transcriptional modification ( Figure 3A). This analysis also demonstrated that down-regulated genes (n = 440) were involved in functions regarding cell cycle, cancer, nervous system development and function, cell morphology, protein synthesis, cell death, cell signaling, cellular growth and proliferation, cellular movement, post-translational modification and free radical scavenging ( Figure 4A).
Networks associated with transcripts selectively modified at the 12 th passage in vitro expansion IPA software predicts functional networks based on known protein-protein and functional interactions. We therefore employed IPA to study how the genes selectively changing their expression at the 12 th passage as compared to the 4 th passage were interacting in specific networks. IPA analysis of the up-regulated genes indicated 8 networks with a score ranging from 46 to 16. The first top network generated by IPA (score = 46) ( Figure 3B) is composed by: i) genes around the gene node IL1, namely IL1B, IL1R1, S100113 and TWIST2, that participate in the inflammatory response and apoptosis; ii) genes around AKT gene node, namely RAC2 and PTPRF, involved in protein phosphorylation, cell differentiation, proliferation and maturation; iii) other genes not linked to any specific node, whose function is related to the transcription regulation (PARP1, HNRNPAB, NONO CREBBP), cytoskeletal organization (ACTR3, NISCH), MSC differentiation (EPHA4, MT2A, CDH2, RAC), cell stress response (ERRFI1), inhibition of apoptosis (RNF7, TWIST2, IL1B), cell proliferation and maturation (LYN) ( Table 2). IPA analysis of the down-regulated genes showed 14 networks with a score ranging from 45 to 11. The first top network generated by IPA (score = 45) ( Figure 4B) is composed by: i) genes around p53 gene node, namely UBE2D3, MAP2K3, ZHX1, XPO1, PIN1, CDC27, ANAPC2; ii) other genes not linked to any specific node, whose function was related to cell proliferation and pluripotency (TYMS, FGFR3, MITF and NCAPD2), oxidative stress (RPS3, NFE2L2, HSP1, HAT1), NF-κB signaling (IKBKG, MYL5), apoptosis (SLC6A4, TPH1) and histone modification (HAT1, HIST1H3C) ( Table 3).

TaqMan Real Time quantitative PCR and western blot: validation of the microarray data
qRT-PCR analysis, performed in order to validate microarray data by investigating three down-regulated (p53, HSPE1 and HIST1H3C) and three up-regulated genes (IL1B, CREBBP and LYN) present in the first top up and down gene networks, respectively, ( Figure 5) and western blot analysis of p53, HIST1H3C and IL1B protein (Additional file 5: Figure S3; Additional file 6: Western Blot Method), confirmed the results obtained by the microarray analysis.

Discussion
It is well known that during cell culture different kind of stem cells can undergo functional modifications related to the number of culture passages [70,71]. The identification of these modifications is of crucial relevance in order to better understand the biology of stem cells and the differences in their ability to proliferate and differentiate, along with cell expansion. The study of the whole stem cell transcriptome, carried out by microarray technology, allows to identify the global changes occurring in the expression profiles of these cells, providing useful information about their functional changes along with culture progression. In the present study, we carried out an analysis of gene expression profiles of WJ-MSCs at 4 th and 12 th passages, in order to evidence the modifications in the trascriptome induced by the culture time prolongation. IPA gene analysis revealed that the top upregulated network was characterized by the presence of 2 independent functional nodes (IL1, AKT). IL1 node is connected with IL1B, IL1R1, S100113 and TWIST2 genes, related to inflammation. The up regulation of these genes in older cells could reflect the presence of an inflammation cellular response acting as a defense mechanism of damaged cells for preventing cell death. AKT node, connected with RAC2 and PTPRF genes, is involved in protein phosphorylation, regulating cellular growth and differentiation, and their over expression appears to be mainly related to cell differentiation and maturation [25]. Other up-regulated genes in this network, although not related to any evident node, are anyway worth of interest due to their specific functions. Among these, four genes (CREBBP, PARP1, HNRNPAB, NONO) are well known to be involved in the regulation of transcription and play an important role in various cellular processes such as differentiation and proliferation [30,32]. In particular, CREBBP gene is critical in embryonic neural development [29] and HNRNPAB gene, interacting with Oct4, is involved in embryonic stem cells differentiation [31]. On the other hand, cytoskeleton organization is the main function played by the other two genes present in the network, complex Arp2/ 3 and Nischarin (ACTR3, NISCH). In the top upregulated network are also presents 3 genes (EPHA4, MT2A, CDH2) representing neural marker related to MSC differentiation. Interestingly, the contemporary activation of RAC2 and CDH2 has been reported as related to MSC differentiation [37]. The last three genes with interesting function detected in this network are ERRFI1, RNF7 and LYN. The up-regulation of these genes is associated with cell growth and proliferation [40] and ERRFI1 activity is induced during cell stress [38,72]. RNF7 has an antiapoptotic activity mediated by the activation of the transcription factor JUN [39]. Among the down-regulated gene network, the most interesting node is centered around the TP53 gene, encoding for the major suppressor protein that can recognize DNA damage and subsequently arrest the cell cycle and trigger the repair process. It has been reported that TP53 functional activity is reduced during the aging process [41,42], probably due to a reduced transcriptional activity [73]. In the present study, the downregulation of the TP53 in older cells was associated to a contemporary down regulation of genes encoding for both transcriptional factors and proteins involved in post-translational modifications, such as UBE2D3, MAP2K3, ZHX1, XPO1, PIN1, CDC27, ANAPC2 [74]. In the same network are also present other genes related to specific cell functions not linked to any evident node gene. Some of these genes, listed in results section, are of particular interest for further discussion. The down regulation of genes related to oxidative stress response (RPS3, NFE2L2, HSP1 and HAT1), has been reported to indicate the cell inability to react to oxidative and DNA damage [59,60,75]. The reduced capacity in protein folding, related to HSP1 down-expression, causes incapacity to react to stress and accelerates aging process [61][62][63]. The down-expression of IKBKG and MYL5, involved in NF-κB signaling [65], can be related to abnormal growth of cells while MITF down-regulation is associated with loss of pluripotency [56]. In the down-expressed dataset are present two genes (HIST1H3C and HAT1) involved in histone modifications. The down expression of HIST1H3C (encoding a member of the histone H3 family) is related to loss of pluripotency and represents a marker of differentiation in stem cells [76]. The highest expression of HAT1 (encoding a protein involved in rapid acetylation of newly synthesized cytoplasmic histones) occurs during embryogenesis and its down expression is related to aging and differentiation processes [68]. Taken together, all data obtained in this study indicate that WJ-MSCs appear to undergo a process of aging rather than senescence during the in vitro expansion from the 4 th to the 12 th culture passage. In fact, cellular aging can be defined as a progressive decline in the physiological properties of tissues, characterized by a decreased replication capacity and an increase of cell-cycle -arrested cells, while senescence is the state in which cells have irreversibly lost their proliferation ability [41]. This is mainly demonstrated by the observed low frequency of cells staining positive for β-galactosidase found in all different passages, indicating that even at 12 th passage cells cannot be considered as senescent. In addition, further evidences are provided by the presence of intact G1 and G2 cell-cycle checkpoints and by the presence of long telomeric end at all examined passages, which confirm that WJ-MSCs preserve their capability to undergo a high number of cellular divisions up to the 12 th passage. Cellular aging can be considered as an evolutionary conserved defence representing an alternative to cell death in the presence of chronic low stress conditions increasing resistance to apoptosis and thus allowing the survival of post-mitotic cells damaged in their central functions. These figures reflects the cell LYN v-yes-1 Yamaguchi sarcoma viral related oncogene Tyrosine kinase activity; Cellular proliferation [40] biological condition and confirms previously reported data about the presence of an aging phenotype of these cells as evidenced by proteomic analysis [15]. Salminen et al. [41], have recently described that the molecular basis of increased resistance to apoptosis in aging cells involves several mechanisms such as alterations in p53 and NF-κB networks and pathways, protein folding, and increased presence of pro-inflammatory mediators. Our results match with these mechanisms, demonstrating a resistance to apoptosis related to the above described down-expression of p53 and the over-expression of TWIST2, RNF7 and ILB1 genes, present in the first top network, as well as of SOD2, RPS27L and STAMBP genes, included in the up-expressed gene cluster. TWIST2 and ILB1 are able to induce the resistance to apoptosis mediated by NF-κB [77], which activates antiapoptotic survival genes such as SOD2, the cytokine STAMBP, and the caspase activator RPS27L, all positive regulator of anti-apoptotic signaling acting on different pathways. Feng et al. [74], convincingly demonstrated that the functional activity of p53 declines in several murine tissues during aging. Moreover, a reduced capacity in protein folding is related to apoptosis and cell death. Aging process and incapacity to react to stress are also related to down expression of the NFE2L2-NFE2-NFE2L gene complex [60]. The resistance to apoptosis is also enhanced by persistent type of stress, e.g. oxidative stress [41] and in this view, our data show the over-expression of IL1B, LTBP1 AND RAC2 genes, related to the production of oxidative species. The aging process denotes a reduced capacity to maintain intact cytoskeleton structures, affecting cellular processes such as motility, interaction with neighbouring cells and mitosis [34], and in the present study, we demonstrated the over-expression in older cells of NISCH gene. ACTR3 instead is demonstrated to be upregulated during stem cell differentiation [33], in particular the neural differentiation. Moreover, as confirmed by data from cell cultures, the cells at the 12 th passage show decreased proliferation likely related to the above described down-expression of genes such as MAP2K3, PIN1, CDC27, ANAPC2, TYMS, FGFR3, NCAPD2, MYL5. The same cells also show altered modulation of HIST1H3C, PTPRF, EPHA4, MT2A, CDK5R2, CDH2, genes related to neuronal differentiation indicating a loss in WJ-MSCs plasticity.

Conclusion
In conclusion, data provided by cell culture experiments and results obtained by profiling studies, all together demonstrate the impairment WJ-MSCs expansion abilities and their resistance to apoptosis, two hallmarks of cell aging. On the other hand, all these data show the need to develop novel culture protocols able to preserve stem cell plasticity. Moreover, the genes identified as impaired in the present study, could be useful biomarkers to evaluate cell culture quality when comparing different in vitro expansion methods.

Availability of supporting data
Raw data of the performed microarray experiments have been recorded in the GEO public database (accession number: GSE34929).

Additional file
Additional file 1: Figure S1. Additional file 3: Table S1. List of transcripts resulting up-expressed in the cluster 1.
Additional file 4: Table S2. List of transcripts resulting down-expressed in the cluster 2.
Additional file 5: Figure S3. Western blot analysis. The intensity of immune-reactivity bands (10 μg of protein) of HIST1H3C, P53 and IL1β was measured by densitometry analysis, normalized respect to the corresponding β-Actin bands and expressed as arbitrary units (A.U.) (bottom histograms). Data are expressed as the averages ± SD of three independent experiments.
Additional file 6: Western Blot Method.

Competing interests
No competing financial interests exist.
Authors' contributions VG participated in the design of the study and performed the assembly of data, data analysis and interpretation and drafted the manuscript. MDA participated in assembly of data, microarray experiments, manuscript writing. PL performed cell isolation and culture and telomere length assay. LP performed cell isolation and culture and cell differentiation experiments. SS carried out microarray experiments and participated in assembly of data. GP participated to data analysis and interpretation. MG participated to data analysis and interpretation. MM conceived of the study and participated to samples collection and data interpretation. SM participated in study design and coordination and revising the manuscript critically for important intellectual content. LS participated in study design and coordination, data interpretation and helped to final approval of the manuscript. All authors read and approved the final manuscript.

Authors' information
Sebastiano Miscia and Liborio Stuppia senior investigators contributed equally to this work.