Genetic comparison of mouse lung telocytes with mesenchymal stem cells and fibroblasts

Telocytes (TCs) are interstitial cells with telopodes – very long prolongations that establish intercellular contacts with various types of cells. Telocytes have been found in many organs and various species and have been characterized ultrastructurally, immunophenotypically and electrophysiologically (http://www.telocytes.com). Telocytes are distributed through organ stroma forming a three-dimensional network in close contacts with blood vessels, nerve bundles and cells of the local immune system. Moreover, it has been shown that TCs express a broad range of microRNAs, such as pro-angiogenic and stromal-specific miRs. In this study, the gene expression profile of murine lung TCs is compared with other differentiated interstitial cells (fibroblasts) and with stromal stem/progenitor cells. More than 2000 and 4000 genes were found up- or down-regulated, respectively, in TCs as compared with either MSCs or fibroblasts. Several components or regulators of the vascular basement membrane are highly expressed in TCs, such as Nidogen, Collagen type IV and Tissue Inhibitor of Metalloproteinase 3 (TIMP3). Given that TCs locate in close vicinity of small vessels and capillaries, the data suggest the implication of TCs in vascular branching. Telocytes express also matrix metalloproteases Mmp3 and Mmp10, and thus could regulate extracellular matrix during vascular branching and de novo vessel formation. In conclusion, our data show that TCs are not fibroblasts, as the ultrastructure, immunocytochemistry and microRNA assay previously indicated. Gene expression profile demonstrates that TCs are functionally distinct interstitial cells with specific roles in cell signalling, tissue remodelling and angiogenesis.


RNA isolation and preparation
Mouse lung telocytes were isolated after 5 days of culture. Mouse MSCs and fibroblasts were cultured and collected on days 5 and 10 respectively. RNA preparation was performed using TRIzol reagent (Invitrogen Life Technologies, Carlsbad, CA, USA) and the RNeasy kit (Qiagen, Valencia, CA, USA) according to the manufacturer's instructions, including a DNase digestion treatment. The amount and quality of RNA were measured by NanoDrop-1000 spectrophotometer and with the Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA).

Data analysis
The acquired array images were analysed with Agilent Feature Extraction software (version 10.7.3.1). Quality normalization and subsequent data processing were performed with the GeneSpring GX v11.5.1 software package. The genes detected in all samples were chosen for further data analysis. Differentially expressed genes were identified through Fold Change filtering and hierarchically clustered by the Agilent GeneSpring GX software (version 11.5.1). Gene ontology and String Network analyses were performed with the standard enrichment computation method to study the relation among variant proteins expressed by variant genes.
Fisher's exact test was used to find more overlaps between the descriptive list and the GO annotation list than would be expected by chance. The Pvalue denoted the significance of GO terms enrichment in the descriptive genes.

Results and discussions
The quality of gene data after filtering and the distribution of data sets were assessed and visualized by Box-Plot. There was no significant difference in distributions of log2 ratios among TCs, MSCs and fibroblasts ( Figure S1).

Gene expression analysis
Gene expression array data show that more than 500 genes are at least 10 times higher expressed in TCs comparing with either MSCs or fibroblasts (Table 1). Several genes are found 100 times up-regulated in TCs versus fibroblasts (Cdh2, Cyba, Rnf128, Dpysl3, Fstl1, Rbp1, Gm12892, Cdh2, Aldh1a1, Gm5864) or MSCs (Rbp1 and Glipr1; Table 1A). Additional genes are significantly overexpressed in TCs comparing with MSCs or fibroblasts (Table 1B). Table 2 is a summary of genes found to be down-regulated in TCs. Although many genes are less expressed in TCs comparing with MSCs or fibroblasts, very few are found at least 100 times down-regulated in TCs.

Hierarchical cluster and gene ontology analyses
The hierarchical cluster of the genes with more than twofold changes among telocytes, MSCs and fibroblasts is shown in Figure 1.
Remarkably, the MCSs and fibroblast gene expression profiles relate each other to higher extent than to TCs supporting the view that TCs have a distinct gene expression pattern. In fact this is an important additional proof that TCs and fibroblasts are different cells. The GO analysis indicates that the genes differentially expressed in TCs are mainly involved in development, in tissue and organ morphogenesis and in transport and maintenance of a biological compound to a specific location ( Fig. 2A). In addition, many of the differentially expressed genes likely function in extracellular compartments (Fig. 2B) and may play roles in cell survival, growth and differentiation through autocrine and paracrine activity (Fig. 2C). The relationships, including direct (physical) and indirect (functional) associations, of those genes were analysed by String Network analysis (www.string-db.org). Among the 156 co-expressed genes, 46 genes were found to have certain interactions (Fig. 3). TCs are potentially involved in tissue remodelling and basement membrane homeostasis A set of genes are specifically up-or down-regulated in TCs comparing with both fibroblasts and MSC (Table 3). As last two cell types are developmentally and functionally quite different, one being progenitors and the other differentiated, specialized cells, this set of genes should connect to the specific biological activities of TCs among the other stromal cells. Thus, we have found that several genes with roles in tissue remodelling and repair are significantly up-regulated in TCs (Tables 1A and 3): connective tissue growth factor (CTGF) [24,25], Transgelin (Tagln) [26], Nidogen 1 (Nid1) [27,28], tissue inhibitor of metalloproteinase 3 (TIMP3) [29], collagen type IV, alpha (Col4a4, Col4a6, Col4a5) [28,30], Matrix Metallopeptidase 10 (Mmp10) [31][32][33], Matrix Metallopeptidase 3 (Mmp3) [31][32][33] and Retinol-binding protein 1 (RBP1). RBP1 (also known as CRABP-I, CRBP, CRBP1, CRBPI, RBPC) is required in tissue remodelling [34]. Regarding the molecular mechanisms, RBP1 delivers vitamin A to other cells through the plasma membrane protein STRA6 involved in JAK/STAT signalling and the intracellular metabolism of the vitamin [35]. Remarkably, two main components of basement membrane, Collagen type IV and Nidogen 1 are up-regulated in the cultured TCs comparing with both MSCs and fibroblasts. Moreover, TIMP3 is an extracellular matrix-anchored metalloproteinase inhibitor that acts specifically to increase vascular (endothelial) basement membrane stability [36,37]. As TCs express Matrix Metalloproteases Mmp3 and Mmp10 also, it is likely that TCs are involved in both basement membrane assembly (stability) and surrounding extracellular matrix remodelling.

Concluding remarks
Overall, the data indicate that TCs are clearly distinct from both MSCs and fibroblasts, and the gene signature of TCs suggests specific biological functions in (a) development and tissue morphogenesis, (b) biological compound transport and (c) extracellular matrix remodelling. It has been proposed that TCs play essential roles in angiogen-esis given that TCs are frequently found in close vicinity of small vessels and express angiogenesis-related factors (VEGF, NO) and pro-angiogenic microRNAs [22]. The data presented here bring additional support to this view suggesting that TCs may also regulate vascular basement membrane remodelling as key step in vascular branching and de novo vessel formation.