Impact of source tissue and ex vivo expansion on the characterization of goat mesenchymal stem cells

Abstract

Background: There is considerable interest in using goats as models for genetically engineering dairy animals and also for using stem cells as therapeutics for bone and cartilage repair. Mesenchymal stem cells (MSCs) have been isolated and characterized from various species, but are poorly characterized in goats.

Results: Goat MSCs isolated from bone marrow (BM-MSCs) and adipose tissue (ASCs) have the ability to undergo osteogenic, adipogenic and chondrogenic differentiation. Cytochemical staining and gene expression analysis show that ASCs have a greater capacity for adipogenic differentiation compared to BM-MSCs and fibroblasts. Different methods of inducing adipogenesis also affect the extent and profile of adipogenic differentiation in MSCs. Goat fibroblasts were not capable of osteogenesis, hence distinguishing them from the MSCs. Goat MSCs and fibroblasts express CD90, CD105, CD73 but not CD45, and exhibit cytoplasmic localization of OCT4 protein. Goat MSCs can be stably transfected by Nucleofection, but, as evidenced by colony-forming efficiency (CFE), yield significantly different levels of progenitor cells that are robust enough to proliferate into colonies of integrants following G418 selection. BM-MSCs expanded over increasing passages in vitro maintained karyotypic stability up to 20 passages in culture, exhibited an increase in adipogenic differentiation and CFE, but showed altered morphology and amenability to genetic modification by selection.

Conclusions: Our findings provide characterization information on goat MSCs, and show that there can be significant differences between MSCs isolated from different tissues and from within the same tissue. Fibroblasts do not exhibit trilineage differentiation potential at the same capacity as MSCs, making it a more reliable method for distinguishing MSCs from fibroblasts, compared to cell surface marker expression.

Keywords: Adipose; Bone marrow; Characterization; Differentiation; Goat; Mesenchymal stem cells.

Figure 1

Colony-forming efficiency of MSCs. Colony-forming efficiency (CFE) of P5 MSCs from different sources (A) and 9004 BM-MSC cultured to passages 10, 15 and 20 (B). CFE is expressed as percentage of colonies formed per total cells plated and presented as means (± SEM). Statistically significant groups by Tukey’s test are indicated by letters (P < 0.05).

Figure 2

Osteogenic differentiation of MSCs and fibroblasts. P5 MSCs from 9004 BM-MSC (A, B), 9003 BM-MSC (C, D), 9003 ASC (E, F) and 1014 EF (G, H) were cultured in osteogenic differentiation or control medium, respectively and stained with Alizarin Red S, which stains calcium deposits red. Calcium mineralization in 9003 ASC (E) appeared less extensive than that of the BM-MSCs (A and C). 1014 EF failed to undergo osteogenesis, as evident from the absence of staining. (G, H;). Representative images are shown at 200X magnification. Scale bars represent 100 μm.

Figure 3

Chondrogenic differentiation of MSCs and fibroblasts. P5 MSCs from 9004 BM-MSC (A, B), 9003 BM-MSC (C, D), 9003 ASC (E, F) and 1014 EF (G, H) were cultured in chondrogenic differentiation or control medium, respectively and stained with Alcian Blue, which stains cartilage blue. Cellular condensation, as well as ridge and micromass formations that stain positive were observed in 9004 BM-MSC, 9003 BM-MSC, 9003 ASC and 1014 EF cultured in chondrogenic medium (A, C, E, G respectively). Some staining was observed in cells cultured in control medium, but cells generally remained in monolayer (B, D, F, H respectively). Representative images are shown in phase contrast at 40X magnification. Scale bars represent 500 μm.

Figure 4

Oil Red O staining of MSCs and fibroblasts differentiated using two methods of adipogenic induction. P5 9004 BM-MSC (A, E, I), 9003 BM-MSC (B, F, J), 9003 ASC (C, G, K) and 1014 EF (D, H, L) were differentiated by the Adipo I method (A-D), Adipo II method (E-H), or cultured in expansion medium (I-L). Differentiated adipocytes accumulated lipid droplets in the cytoplasm that stain red with Oil Red O. Cells cultured in control medium (I-L) and 9003 BM-MSC induced by Adipo II (F) did not yield lipid-filled adipocytes. Representative images are shown in phase contrast at 200X magnification. Scale bars represent 100 μm.

Figure 5

Adipogenic differentiation capacity of P5 MSCs and fibroblasts. The percentage of lipid-positive cells over total cells counted (± SEM) is shown for each cell line differentiated with Adipo I (A) and Adipo II (C). Quantitative RT-PCR analysis of PPARG and FABP4 expression was performed on each cell line differentiated with Adipo I (B) and Adipo II (D), and data is presented as fold change (± SEM) in expression relative to expression levels in 1014 EF (fold change ~1, indicated by the dotted line). Statistically significant groups by Tukey’s multiple means comparison test are indicated by letters (P < 0.05 for A-D).

Figure 6

Adipogenic differentiation capacity of BM-MSCs at higher passages. The percentage of lipid-positive cells over total cells counted (± SEM) for 9004 BM-MSC at passages 10, 15 and 20 (A) and quantitative RT-PCR analysis of PPARG and FABP4 expression for each passage presented as fold change (± SEM) in expression relative to control cells at corresponding passages (B). Statistically significant differences are indicated in letters according to analysis by Tukey’s multiple means comparison (P < 0.05).

Figure 7

RT-PCR analysis of gene expression in MSCs and fibroblasts. RT-PCR was performed using primers that amplify the cell surface markers CD90, CD105, CD73, CD45 and CD34, and the pluripotency markers OCT4, SOX2 and NANOG. Primers amplifying GAPDH were used as a positive control for the cDNA. PCR products were visualized in 2% agarose.

Figure 8

Cell surface marker expression in MSCs and fibroblasts. Quantitative RT-PCR was used to measure the expression of the cell surface markers CD90, CD105 and CD73 in P5 MSCs (A) and in 9004 BM-MSCs at passages 10, 15 and 20 (B). Expression levels are presented as fold change (± SEM) relative to 1014 EF (A) or expression levels in P5 9004 BM-MSC (B) (fold change ~1, indicated by the dotted line). Statistically significant groups by Tukey’s multiple means comparison test are indicated by letters (P < 0.05). Significant differences between the cell lines were observed in CD90 and CD105 expression. Significant difference between the different passages was observed in CD105, but not in CD90 and CD73.

Figure 9

Immunofluorescent staining of OCT4 in MSCs and fibroblasts. P5 cells were stained with antibodies against OCT4. OCT4 was localized in the cytoplasm of MSCs and fibroblasts, whereas staining for OCT4 was specific to the nuclei in the bovine blastocyst. Representative images are shown at 200X magnification.

Figure 10

Transfection and integration efficiency in MSCs. P5 MSCs (A) and 9004 BM-MSC at passages 10, 15 and 20 (B) were transfected with pEGFP-N1. Integrant number were expressed as the number of discrete colonies that were resistant to G418 selection. Statistically significant groups by Tukey’s test is indicated by letters (P < 0.05). 9004 BM-MSC showed the highest number of integrants, followed by 9003 ASC and 9003 BM-MSC. 9004 BM-MSC showed a decreasing trend in number of integrants with increasing passages.