Indirect co-culture with tendons or tenocytes can program amniotic epithelial cells towards stepwise tenogenic differentiation

Abstract

Background: Amniotic epithelial cells (AEC) have potential applications in cell-based therapy. Thus far their ability to differentiate into tenocytes has not been investigated although a cell source providing a large supply of tenocytes remains a priority target of regenerative medicine in order to respond to the poor self-repair capability of adult tendons. Starting from this premise, the present research has been designed firstly to verify whether the co-culture with adult primary tenocytes could be exploited in order to induce tenogenic differentiation in AEC, as previously demonstrated in mesenchymal stem cells. Since the co-culture systems inducing cell differentiation takes advantage of specific soluble paracrine factors released by tenocytes, the research has been then addressed to study whether the co-culture could be improved by making use of the different cell populations present within tendon explants or of the high regenerative properties of fetal derived cell/tissue.

Methodology/principal findings: Freshly isolated AEC, obtained from ovine fetuses at mid-gestation, were co-incubated with explanted tendons or primary tenocytes obtained from fetal or adult calcaneal tendons. The morphological and functional analysis indicated that AEC possessed tenogenic differentiation potential. However, only AEC exposed to fetal-derived cell/tissues developed in vitro tendon-like three dimensional structures with an expression profile of matrix (COL1 and THSB4) and mesenchymal/tendon related genes (TNM, OCN and SCXB) similar to that recorded in native ovine tendons. The tendon-like structures displayed high levels of organization as documented by the cell morphology, the newly deposited matrix enriched in COL1 and widespread expression of gap junction proteins (Connexin 32 and 43).

Conclusions/significance: The co-culture system improves its efficiency in promoting AEC differentiation by exploiting the inductive tenogenic soluble factors released by fetal tendon cells or explants. The co-cultural system can be proposed as a low cost and easy technique to engineer tendon for biological study and cell therapy approach.

Figure 1. Molecular characterization of freshly isolated AEC.

A) Levels of surface and intracellular stemness markers analyzed by flow cytometry and expressed as Mean Fluorescence Intensity (MFI) ratio. B) The mRNA content of tendon-related genes analyzed by RT-PCR. The bars show the standard error calculated on 3 independent experiments. C) Two representative images of cytokeratin 8 (epithelial marker) and α-SMA (mesenchymal marker) proteins detected in AEC by using an immunocytochemistry approach. The images show the blue nuclei counterstained with DAPI, and both proteins in red (Cy3). Scale bar for all images = 50 µm. D) The in vitro differentiation of AEC into endoderm (liver: bottom image) and ectoderm (neural cells: top image) cell lineages are documented by the immunocytochemistry detection of nestin and albumin, respectively. Nuclei were counterstained with DAPI. The mesodermal osteogenic in vitro differentiation (central image) was documented by the Alizarin Red staining. Scale bar for all images = 50 µm.

Figure 2. Morphology of differentiated AEC toward tenocyte lineage after co-culture.

The four top images reproduce the more representative phenotypes acquired by AEC (monolayer, circular aggregates, elongated, and tendon-like structures) during the 4 weeks of co-incubation. The bigger images were obtained with the aid of a contrast microscope, while the small ones, inserted in the corner, represent a low magnification recorded under a stereomicroscope. In the lower part of the Table, the incubation intervals (expressed in weeks) required to obtain these different phenotypes are indicated. Scale bar for all images = 50 µm.

Figure 3. Molecular, genomic, and functional characterization of co-cultured AEC.

A) The top images are two representative examples of migration testes perfomed in cultured and co-cultured AEC during the first week of incubation. As indicated by the+symbol, only the AEC co-cultured with fetal tendon explants show a migration activity. B) Representative images of immunocytochemistry showing the distribution of cytokeratin 8 and α-SMA in AEC co-cultured with fetal primary tenocytes after 14 or 28 days of incubation. All images show cell nuclei in blue (DAPI) and both the proteins in red (Cy3). At day 14, the co-cultured AEC organized in monolayer display high levels of cytokeratin 8, and undetectable levels of α-SMA. A similar molecular phenotype is displayed by monolayered AEC cultured alone (small insets) or co-cultured with fetal tenocytes (arrows in large figures) for 28 days. By contrast, an opposite behaviour is observed in co-cultured AEC that organized cell-aggregates. Scale bar for all images = 100 µm. C) Q-FISH detection of telomere length in freshly isolated AEC (AEC), in cultured AEC (AEC 28 days), in AEC co-cultured with fetal explants (AEC+fetal explants 28 days), and fetal tenocytes. The top figure is a representative image showing several hybridized red telomeres (Cy3) within two interphase nuclei stained in blue with DAPI. Scale bar = 15 µm. The three box plots indicate the Telomere area (TEA), the feret maximum (TEF), and the mean densitometric value (MEAND) parameters. The horizontal lines express the 5th, 25th, 50th, 75th, and 95th percentile of the distribution. The box stretches from the 25th to the 75th percentile, and therefore contains the middle half of the scores in the distribution. The median is shown as a line across the box, meanwhile the mean value as a black square within the box. ^* indicates data of TEA, TEFmax, and MEAND that resulted significantly different from AEC for p<0.01 after One Way ANOVA test followed by post-hoc Tukey test. D) Three representative normal karyotypes obtained by freshly isolated AEC (day 0), cultured AEC and cells co-cultured in the presence of fetal explants. E) Flow cytometry for the major histocompatibily (MHC) class I and II molecules performed on ovine peripheral blood mononuclear cells (PBMC; top image) to test the ovine reactivity of both the antibodies and on freshly isolated AEC (bottom images) to demonstrate the presence of the MHC class I and the absence of MHC class II antigens.

Figure 4. COL1 distribution in AEC aggregates developed in co-culture.

The images exemplify the COL1 protein distribution recorded by immunohistochemistry in the more representative typologies of cell aggregates obtained during AEC co-cultures. The images show in blue the nuclei counterstained with DAPI and in green the COL1 protein (Alexa Fluor 488). The COL1 protein is undetectable when the AEC are organized in monolayer independently from the cultural conditions adopted (co-cultured AEC, monolayer; AEC cultured alone, small inset). COL1 starts to appear in AEC forming cell aggregates and reaches its highest and widespread distribution within the tendon-like structures. In the early phase, COL1 is localized within the fusiform cells that start to be oriented along the longitudinal axis of the tendon-like structures. Later, the protein was either localized into the AEC or deposited within the extracellular matrix (tendon-like: late phase). Scale bar for all images = 100 µm.

Figure 5. OCN, Cx32 and Cx43 proteins distribution in AEC co-cultured with fetal tenocytes or tendon explants.

Representative images showing osteocalcin (OCN: lower panel) and connexins (Cx32 or Cx43: upper panel) immunolocalization in AEC cultured alone (AEC) and in AEC co-cultured fetal derived cell/tissues that developed 3D structures. The pictures show the cell nuclei in blue (DAPI), OCN or Cx32 proteins in green (Alexa Fluor 488) and Cx43 in red (Cy3). The co-expression of Cx proteins on AEC cell aggregates were analyzed with a double immunostaining. OCN and both the Cx proteins were undetectable in AEC cultured alone (AEC) and in AEC organized within elongated aggregates. By contrast, AEC that differentiated 3-D tendon-like structures co-expressed Cx32 and Cx43. The Cx43 protein shows higher levels in the early tendon-like structures with a clear membrane localization, while Cx32, more abundant in late structures, is localized either on the membrane or into the cytoplasm. OCN appears as a cytoplasmatic protein within the fusiform shaped cells forming the tendon-like structures (early phase). Its intracellular levels progressively increased during the process of in vitro tendon differentiation (late phase). As indicated in the corner box showing a representative image of an ALP assay, tendon-like structures did not display any osteogenic foci. Scale bar for all images = 50 µm.

Figure 6. Expression profile of tendon/ligament-related genes in co-cultured AEC.

The mRNA content of SCXB, COL1, COL3, TNMD, THSB4, OCN was analyzed in cultured AEC or in AEC co-cultured with fetal or adult tenocytes/tendons explants by using RT-PCR. The semi quantitative analyses of mRNA levels were normalized for GAPDH gene and expressed as mean of 3 different replicates ± SD. The data were compared by One Way ANOVA test followed by post-hoc Tukey test. ^* Values significantly different from AEC group for p<0.05; ^a values significantly different from fetal tendon group for p<0.05; ^a′ values significantly different from adult tendon group for p<0.05.