Enrichment of neural-related genes in human mesenchymal stem cells from neuroblastoma patients
- Authors:
- Published online on: May 22, 2012 https://doi.org/10.3892/ijmm.2012.1008
- Pages: 365-373
Abstract
Introduction
Neuroblastoma (NB), a poorly differentiated tumor derived from neural crest cells that affects mainly children, is the most common extracranial pediatric solid tumor. The origin of stroma in primary NB tumors, formed by Schwann cells, and whether the stroma is the cause or consequence of the maturation potential of tumor cells remain controversial. It has been hypothesized that crosstalk between Schwann cells and neuroblasts influences the biology and clinical behavior of NB tumors. However, little is known about the role of the NB microenvironment in metastasis localizations, especially in bone marrow (BM). Recent articles suggest that mesenchymal stem cells (MSCs) have a major role in maintaining stem cells niches in the BM (1), as well as in creating the tumor microenvironment (2). In this study we focused in BM-derived MSC from NB patients as a key factor in the development of metastasis.
During the past years several studies have used microarray-based high-throughput technologies to identify biological processes altered in NB cells. Hiyama et al (3) surveyed the differences in gene expression between unfavorable and maturing/regressing NB. Interestingly, in favorable NB, neuronal differentiation signals such as CD44, IGF2, NTRK1 and ANK1 were overexpressed in maturing tumors. Similarly, Kamei et al (4) identified genes that exhibited altered gene expression in NB tumors associated with a favorable outcome. More recently, Chen et al (5) performed parallel global protein and mRNA expression profiling on NB tumors and identified that cell adhesion, nervous system development and cell differentiation processes were downregulated in stage 4 MYCN-amplified NB tumors, suggesting a less mature neural and a more invasive phenotype of this type of cells. Finally, Abel et al (6), based on expression profiling, identified 4 molecular subgroups of NB that can be distinguished by a 6-gene signature. These studies show that microarray techniques are useful tools for gene expression profiling in NB tumors.
It is well known that the BM is a common site for metastasis in patients with high-risk NB. Although little is known about the control of NB tumor growth by the BM microenvironment, there has been an increasing interest in the role of MSCs and their BM niches in cancer (7,8). Some reports implicate MSCs having tumor-promoting effects whereas others show inhibition of tumor growth. Regarding the relationship between MSCs and NB, Ma et al (9) reported that MSCs in BM may enhance metastasis of NB via SDF-1/CXCR4 and SDF-1/CXCR7 signaling. Moreover, CXCR5 may be involved in the attraction of human metastatic NB cells to the BM (10). Despite these findings, the relationship between MSCs and NB cancer cells is still unknown, and research in this area would add new scientific knowledge and provides new therapeutic ideas and targets. Thus, we isolated MSCs from BM of NB patients and control donors and compared their global expression patterns using microarrays. Our findings provide preliminary genetic evidence of the interaction between MSCs and NB cancer cells in BM as well as identify relevant biological processes potentially altered in MSCs in response to NB.
Materials and methods
Mesenchymal stem cell isolation, culture and characterization
MSCs were isolated from 4 NB pediatric patients (with no amplification of the N-Myc gene) and 4 healthy donors. The study protocol was approved by the Ethics Committee of the Hospital. All patients and volunteers were informed about the purpose of the study and provided written consent, by the parents or legal guardians, regarding their participation in the study. BM-derived MSCs were obtained by adherence to plastic. Mononucleated cells were obtained after centrifugation using a Ficoll-Paque gradient. Cells were cultured at 37°C with 5% CO2 in DMEM (Lonza) supplemented with 10% fetal bovine serum (PAN-Biotech GmbH). MSCs cultures were characterized according the International Society for Cellular Therapy criteria (11).
RNA isolation and cRNA labeling
MSCs were stabilized in PrepProtect™ (Miltenyi Biotec) and total-RNA was isolated using standard RNA extraction protocols (NucleoSpin® RNA II, Macherey-Nagel). RNA integrity and overall quality was checked via the Agilent 2100 Bioanalyzer expert software (Agilent Technologies). All RNA samples revealed an RNA Integrity Number (RIN) between 7.3 and 10. For the linear T7-based amplification step, 1 μg of each total-RNA sample was used. To produce Cy3-labeled cRNA, the RNA samples were amplified and labeled using the Agilent Low RNA Input Linear Amp kit (Agilent Technologies) following the manufacturer’s protocol. Yields of cRNA and the dye-incorporation rate were measured with the ND-1000 Spectrophotometer (NanoDrop Technologies).
Microarray hybridization
The hybridization procedure was performed according to the Agilent 60-mer oligo microarray processing protocol using the Agilent Gene Expression Hybridization kit (Agilent Technologies). Briefly, 1.65 μg Cy3-labeled fragmented cRNA in hybridization buffer was hybridized overnight (17 h, 65°C) to Agilent Whole Human Genome Oligo Microarrays 4×44K using Agilent’s recommended hybridization chamber and oven. Finally, the microarrays were washed once with 6X SSPE buffer containing 0.005% N-lauroylsarcosine for 1 min at room temperature followed by a second wash with preheated 0.06X SSPE buffer (37°C) containing 0.005% N-lauroylsarcosine for 1 min. The last washing step was performed with acetonitrile for 30 sec. Fluorescence signals of the hybridized Agilent microarrays were detected using Agilent’s Microarray Scanner System (Agilent Technologies). The Agilent Feature Extraction Software (FES) was used to read out and process the microarray image files.
Microarray data analysis
The microarray raw data have been deposited at the NCBI Gene Expression Omnibus under the accession number GSE35133 (http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?token=rdszfaigqkmgqdw&acc=GSE35133). Intensity values of flagged spots below background were filtered out, data were normalized using the array median and the mean values of 3 replicates for each biological sample were calculated. Statistical analysis of microarray significance (SAM) was performed to identify genes with significant changes in expression, and permutations were used to estimate the false discovery rate (FDR) (12). Genes were selected if they showed a 2-fold regulation and a FDR <0.05%. These statistical analyses were performed using TMEV and default parameters were used unless specified (13).
Gene ontology (GO) analysis
To better characterize the functionally-related genes which showed at least a 2-fold regulation in the microarray analysis, the genes were assigned to Gene Ontologies using a database for annotation, visualization and integrated discovery (DAVID) (14). We used the 3 ontologies produced by the GO consortium, namely ‘biological process’, ‘cellular component’ and ‘molecular function’. GO terms were collected, redundant terms were excluded, and P-values were used to evaluate the significance of the terms.
Results
We isolated MSCs from NB patients (NB-MSCs) as well as from normal donors. All MSC cultures were characterized according to the International Society for Cellular Therapy criteria: morphology, positive/negative markers and differentiation properties (data not shown) (11). We analyzed the expression profiles of RNA from NB-MSCs compared to those from normal donors, using the Agilent Whole Human Genome Oligo Microarrays. The results from the SAM analysis (≥2-fold regulation and FDR <0.05%) revealed that 454 genes had transcript levels significantly higher in NB-MSCs, whereas 42 genes had transcript levels significantly lower in NB-MSCs (Fig. 1). A list of the genes showing the strongest upregulation in NB-MSCs is shown in Table I whereas the downregulated genes are shown in Table II.
To illustrate the differences between NB and normal samples the genes whose expression was induced or repressed by at least 2.5-fold in NB samples are shown as heat map in the Fig. 2, where NB and normal samples are clearly differentiated. Interestingly, several of the genes are known to play roles in NB (ANXA10, ITGA2, COL4A5 and SHC3) or other types of cancer (TMEFF2, TSPAN8, DCBLD2, PSCDBP and BCAN) (Table I). Table V lists the genes involved in neuronal processes that were >2.0-fold upregulated.
To further examine the differences in the expression profiles between NB-MSCs and normal donors, the 496 significantly up or downregulated genes were analyzed with the DAVID software and classified into the 3 main GO domains. For the upregulated genes, in the gene ontology ‘biological process’ we identified 138 terms, in ‘cellular component’ 23 terms, and in ‘molecular function’ 17 terms. The highest ranked terms are shown in Table III. For the downregulated genes, in the gene ontology ‘biological process’ we identified 24 terms, in ‘cellular component’ 1 term, and in ‘molecular function’ 9 terms. The highest ranked terms are shown in Table IV.
 | Table IIIMain GO terms of the different GO categories enriched in upregulated genes in hMSCs from NB patients. |
| Table IVMain GO terms of the different GO categories enriched in downregulated genes in hMSCs from NB patients. |
Discussion
The role of MSCs in tumor progression is unclear as it has been suggested that MSCs may promote or suppress tumor growth (8). Therefore, identifying potential genes regulated in BM-derived MSCs by NB cancer cells would be of great importance to assess their role in determining disease outcome. Previously, Hahn et al (15) studied the effect of conditioned medium of BM cultures in NB cell growth in vitro (15). They showed that BM cultures may stimulate the proliferation and differentiation suppression of NB cells. In this model, monocytes seem to be the mediators of these effects. However, there is no data on how NB cells modify the characteristics of BM-resident cell populations. In this study, we analyzed for the first time the expression profiles of BM-derived MSCs from NB patients and report the identification of 496 genes with more than a 2-fold increase or decrease transcript levels. Our findings suggest that NB cancer cells may have an impact on several processes of MSCs localized in BM.
Interestingly, our microarray analysis revealed that some of the top ranked upregulated genes in NB-MSCs (ANXA10, ITGA2, COL4A5 and SHC3) have been previously reported to have a potential role in NB (Table I). For instance, Annexin A10 (ANXA10) has been identified in a microarray analysis of human NB stem cells as a gene associated with malignancy (16). Similarly, integrin upregulation has been reported as a marker of NB cell differentiation (17). The same study identified the overexpression of COL4A5 in unfavorable NB. Finally, a distinct role of ShcC (SHC3) docking protein in the differentiation of NB has been proposed (18). In addition, we identified several genes reported in other studies to be regulated in different types of cancer: TMEFF2 (19,20), TSPAN8 (21,22), DCBLD2 (23,24). Taken together, these observations suggest that the interaction between MSCs and NB cancer cells in the BM microenvironment induces changes in the expression of cancer-related genes in the MSCs. An unlikely explanation would be the cellular fusion of NB cells and MSCs, in a similar manner to Rizvanov et al (25) who observed rare in vitro cell fusion in co-cultures of NB tumor cells and MSCs.
The GO functional classification analysis through DAVID showed a number of mainly affected categories further suggesting that NB-MSCs are altered (Table III). Overall, we noted that in the ‘cellular component’ domain, NB upregulated a large number of genes encoding proteins ‘intrinsic to membrane’ (n=98) and localized to the ‘extracellular region’ (n=54), suggesting that NB cancer cells may exert a large repertoire of changes in these MSCs compartments. It is likely that functional relationships between NB and MSCs are mostly mediated through these proteins. Analysis of the category ‘biological process’ revealed effects on MSCs in terms previously described in NB cancer cells in the literature. Most importantly, we remark on the upregulation of genes in the term ‘neurological system processes’. Despite of the wide variety of proteins encoded by these genes, they may provide insights into potential neurological functions altered in NB-MSCs. Interestingly, regulation of neural-related genes has also been shown in previous microarray analysis of NB tumors. Thus, Hiyama et al (3) reported that in favorable NB neuronal differentiation signals were overexpressed in maturing tumors whereas Chen et al (5) found that in NB tumors of stage 4+, proteins with functions in nervous system development were downregulated, suggesting a less mature neural and a more invasive phenotype of these tumors.
The absence of NB markers in our MSC cultures pants to the absence of a tumor cell contamination. Then, it is tempting to speculate whether our findings reflect that MSCs in the BM microenvironment of NB patients redirect toward neuronal lineage. Therefore, MSCs have been proposed to adopt neural cell phenotypes, although this occurs at a very low frequency (26). In this sense, it has been proposed that NB cells would induce MSCs differentiation into Schwann cells (27). However in our data we did not observe an increase of classical Schwann-markers such as S100, Egr-1 or Egr-2 in NB-MSCs. On the other hand, MSCs would suffer a dedifferentiation process since a neuroectodernal origin of fetal MSCs localised in BM has been proposed (28).
The term ‘cell surface receptor linked signal transduction’ included several members of the G protein-coupled receptor family (GPR128, GPR135, GPR132, GPR56 and GPR34). Indeed, GPR128 was the strongest upregulated gene in our microarray analysis (Table I). The G protein-coupled receptor (GPCR) superfamily has long been proposed to have vital dual roles in cellular adhesion and signaling (29). One of the best described GPCRs is GPR56. In addition to its role in neural progenitor cell migration (30), a role in suppression of tumor growth by the microenvironment have been investigated (31). By interacting with an extracellular matrix ligand, TG2 (transglutaminase 2), GPR56 seems to suppress tumor growth and metastasis in vivo; conversely, reduced expression is associated with tumor progression. In addition, it is overexpressed in many human glioblastomas and functions in tumor cell adhesion (32). SHC3, another component of this GO term, has been described in the literature. Miyake et al (18) observed a significantly higher level of ShcC protein in NBs with poor prognostic factors and indicated that the expression of ShcC potentially has a function in inhibiting the differentiation of NB cells (18).
Upregulation of ‘cell adhesion’ genes supports the hypothesis that NB-MSCs may undergo changes in their extracellular matrix and cell adhesion properties. Previous research has shown the modulation of NB cell differentiation by the extra-cellular matrix (33). In this study, the authors showed how extracellular matrix rigidity potentiates NB cell differentiation and decreases cell proliferation; and, as we mentioned above, the receptors of extracellular matrix molecules have been reported as markers of NB cell differentiation (17). Similarly, Chen et al (5) reported that suppression of cell adhesion proteins in NB tumors of stage 4+ indicates the metastatic nature of this kind of NB tumors.
In addition, increased transcript levels of a number of relevant genes involved in cell-cycle regulation or apoptosis (e.g., CCND1, CDKN2A (p14/p16), RET and GDNF) suggest that there may be alterations in these processes in NB-MSCs. Similarly, the number of upregulated genes in the ‘calcium ion binding’ category may indicate that intracellular calcium is likely involved in the response of MSCs to NB. ANXA10 belongs to the annexin super-family of closely related calcium and membrane-binding proteins, and many studies have shown their potential role in tumor development and progression (34). An alternative explanation would be based in the evidence that NB cells stimulate osteoclasts to generate osteolytic lesions and set free calcium, in which interactions of NB cells with BM-derived MSCs play a critical role (35).
Finally, most of the downregulated GO terms contained genes encoding immune-related proteins. However, our results also showed upregulation of HLA-DRB5 and HLA-DOA1, which encode MHC class II molecules. In this sense, Johann et al (36) showed that NK cell cytotoxicity was significantly impaired after co-culturing NB cells with NB-MSCs, compared with MSCs of normal donors. Further study is needed to assess the impact of NB cancer cells on the immune response of MSCs.
In summary, we present initial data of a genome-wide analysis of MSCs from NB patients. Our data suggest that the microarray approach is a useful tool to identify deregulated genes in cultured MSCs isolated from NB patients. We provide preliminary genetic evidence of the interaction between MSCs and NB cancer cells in BM. Furthermore, we identifed relevant biological processes potentially altered in MSCs in response to NB. Future studies are necessary to connect these and other differentially expressed genes into their biological roles.
Acknowledgements
This study was supported by grants from the Fondo de Investigaciones Sanitarias (FIS; PI05/2217 and PI08/0029 to J.G.C.), MICINN (PLE2009-0115) and the Madrid Regional Government (S-BIO-0204-2006 and P2010/BMD-2420) in Spain. The experiments were approved by the appropriate committees.
