Abstract
Introduction
Cell-based tendon engineering is an attractive alternative therapeutic approach to established treatments of tendon injuries. Numerous cell types are promising source of tendon engineering; however, there are certain disadvantages for each cell type. Interestingly, dermal fibroblasts (DFs) are able to transdifferentiate into other cell types, they are widely distributed in dermis and easy to harvest and isolate. Furthermore, pilot clinical studies suggested a promising therapeutic potential of autologous DFs for discorded tendons (Connell et al., 2009&2011), but the underlining repair mechanisms remain unclarified. To investigate tenogenic differentiation process in great detail, we have previously established a three-dimensional (3D) cell sheet model, comprising of three consecutive step (expansion, stimulation and maturation) leading to the formation of 3D tendon-like tube (Hsieh et al., 2018; Yan et al., 2020). Hence, the aim of this study was to carry out pilot examination of the tenogenic potential of human DFs (hDFs) by implementing the 3D cell sheet model.
Methods
hDFs (company purchased, n=2), hBMSCs (human bone marrow mesenchymal stem cells, n=1) and hTSPCs (human tendon stem/progenitor cells, n=1) were used and subjected to the 3D model. In 2D culture, semi-qPCR was performed to validate the expression of DF markers in hDFs, namely NTN1, PDPN and CD26 for papillary dermis layer, and PPARG, ACTA2 and CD36 for reticular dermis layer). FACS analysis and immunofluorescence were employed to validate expression of CD73, CD90, CD105 and vimentin (mesenchyme marker), respectively. After harvesting the 3D cell sheets, wet weigh measurements, H&E and collagen type I stainings, and semi-qPCR for Scleraxis and tenomodulin were executed.
Results
Semi-qPCR of DF markers validated the dermal origin of both donor-derived hDFs; however, the data suggested that donor 1 was mixed cell pool of papillary and reticular dermal cells, whilst donor 2 was reticular dermal cells. In FACS analysis, the expression levels of CD73 and CD90 were comparable among all cell types. For CD105, ca. 20% of the cells were negative in both hDF and hTSPC cultures, but only 2% in hBMSCs. As expected, all three cell types were vimentin-positive. 3D cell sheet formation was successful for all cell type. Interestingly, the hDF cell sheets were thicker and ca. 2-fold heavier than that of hBMSCs and hTSPCs. Next, H&E and collagen I analyses revealed higher cellularity as well as higher collagen I deposition in the hDF sheets compared to the other two cell types. Last, semi-qPCR for Scleraxis and tenomodulin suggested upregulation of both genes in hDF and hTSPC sheets versus 2D culture.
Discussion
Our pilot data suggests that hDFs perform well and even exceed hBMSCs and hTSPCs in the 3D model in terms of size, cellularity and collagen I expression. However, increase of cohort size and further detailed molecular and histomorphometric analyses are necessary to conclude on the promising tenogenic potential of hDFs.
