Differential marker expression by cultures rich in mesenchymal stem cells

Abstract

Background: Mesenchymal stem cells have properties that make them amenable to therapeutic use. However, the acceptance of mesenchymal stem cells in clinical practice requires standardized techniques for their specific isolation. To date, there are no conclusive marker (s) for the exclusive isolation of mesenchymal stem cells. Our aim was to identify markers differentially expressed between mesenchymal stem cell and non-stem cell mesenchymal cell cultures. We compared and contrasted the phenotype of tissue cultures in which mesenchymal stem cells are rich and rare. By initially assessing mesenchymal stem cell differentiation, we established that bone marrow and breast adipose cultures are rich in mesenchymal stem cells while, in our hands, foreskin fibroblast and olfactory tissue cultures contain rare mesenchymal stem cells. In particular, olfactory tissue cells represent non-stem cell mesenchymal cells. Subsequently, the phenotype of the tissue cultures were thoroughly assessed using immuno-fluorescence, flow-cytometry, proteomics, antibody arrays and qPCR.

Results: Our analysis revealed that all tissue cultures, regardless of differentiation potential, demonstrated remarkably similar phenotypes. Importantly, it was also observed that common mesenchymal stem cell markers, and fibroblast-associated markers, do not discriminate between mesenchymal stem cell and non-stem cell mesenchymal cell cultures. Examination and comparison of the phenotypes of mesenchymal stem cell and non-stem cell mesenchymal cell cultures revealed three differentially expressed markers - CD24, CD108 and CD40.

Conclusion: We indicate the importance of establishing differential marker expression between mesenchymal stem cells and non-stem cell mesenchymal cells in order to determine stem cell specific markers.

Figure 1

Tri-lineage differentiation potential of tissue cells. Cells from bone marrow (BM) (A-C), breast adipose (BA) (D-F), foreskin fibroblasts (FF) (G-I) and olfactory tissue (OT) (J-L), and were exposed to conditions known to induce the differentiation of mesenchymal stem cells into chondrocytes, adipocytes and osteocytes. Cytochemical stains were carried out to indicate differentiation; Safranin O (chondrocytes: A, D, G, J-Scale bar (J) = 200 μm), Oil-red-O (adipocytes: B, E, H, K whole well scans (5×5 montage) Scale bar (K) = 1 mm: inset-Scale bar (K) = 40 μm) and Alizarin Red (osteocytes: C, F-Scale bar (F) = 100 μm: I, L-Scale bar (L) = 50 μm). All images are representative except for F and L which portray very rare osteocyte staining in FF and OT cultures. M: Illustrates confirmation and quantitation of differentiation by qPCR. Taqman probes for aggrecan, adiponectin and osteopontin were used to quantitate chondrocyte, adipocyte and osteocyte differentiation respectively amongst our tissue samples. ΔΔCt values were calculated relative to GAPDH and were normalized relative to BM (N = 3/cell type). Standard error is displayed and significant differences between tissue types (P < 0.05) are indicated (*). RNA was extracted on the same day as cytochemical analysis; chondrocytes (day 21), adipocytes (day 25) and osteocytes (day 21).

Figure 2

Phenotypic analysis of tissue cells by flow-cytometry and immuno-fluorescence. A: Bone marrow (BM), breast adipose (BA), olfactory tissue (OT) and foreskin fibroblast (FF) cells were analyzed by flow-cytometry for the expression of 17 extracellular markers and the percentage of positive cells calculated compared with unstained controls (N = 3/cell type). Standard error is presented and significant differences between tissue types (P < 0.05) are indicated (*). B: Tissue cells were analyzed for the expression of fibroblast-associated markers; fibroblast surface protein (FSP) (C -olfactory tissue cells), Collagen 3 (Coll 3) (D – bone marrow cells ) prolyl-4-hydroxylase, beta polypeptide (P4HB) (E – breast adipose cells), and Collagen 1 (Coll 1) (F – foreskin fibroblast cells) using immuno-fluorescence. Scale bar (F) = 50 μm - applicable to C, D and E. The percentage of positive cells was calculated compared with no primary antibody controls (N = 3/cell type). Standard error is demonstrated - no significant differences between tissue types were found.

Figure 3

Proteomic and antibody array analysis of tissue cells. A-E: Proteomic analysis of 2-DE gels, visualized with Sypro ruby, derived from bone marrow (BM) (A), breast adipose (BA) (B), olfactory tissue (OT) (C) and foreskin fibroblast (FF) (D) cells. E: Table indicating correlation analysis of protein expression profiles of 2-DE gels derived from BM, BA, OT and FF cell samples (N = 2 samples per tissue). Quantitative expression levels of matched protein spots on 2-DE gels were compared pair-wise. Note; multiple runs of the same sample generated a correlation coefficient of 0.88 (N = 3). F-H: Antibody array analysis of BM, BA, OT and FF cells. Three samples per tissue were pooled and stained for extracellular antibodies (228) using the BD lyoplate kit and images captured (3x3 montage). Representative images of positive (F: BM – CD105) and negative (G: OT – CD200) wells – Scale bar (G) = 100 μm. H: A pair-wise comparison of positive or negative antibody expression was carried out between all tissue cells and the percentage similarity between samples calculated.

Figure 4

Identification of markers differentially expressed between cultures rich in mesenchymal stem cells and non-stem cell mesenchymal cell cultures. A: Analysis of antibody array and proteomics data identified markers upregulated in cultures rich in mesenchymal stem cells (bone marrow – BM; and breast adipose - BA), compared with non-stem cell mesenchymal cell cultures (olfactory tissue – OT). Antibody array and proteomics data also identified markers upregulated in non-stem cell mesenchymal cultures (OT) compared with cultures rich in mesenchymal stem cell (BM and BA). B: qPCR was carried out on all tissues (N = 3/cell type) for markers identified in A. Only markers found to have expression profiles consistent with A are represented (NFL: neurofilament light polypeptide; END: endoplasmin). ΔΔCt values were calculated relative to GAPDH and were normalized relative to the tissue with the highest expression. Standard error is displayed - only significant differences between tissue types, consistent with A, (P < 0.05) are indicated (*).