15502-m-pleumeekers

vitro. After subcutaneous implantation, only constructs containing ECs were able to produce elastin. Most elastin fibres were found around the cell. Elastin = Immunohistochemical staining for elastin ; RF = Resorchin Fuchsin, chemical staining for elastin. (E) All cartilage constructs were of human origin. Biochemical data are shown as mean ± SD. For statistical evaluation, a mixed model was used. Histological data are shown as the median of individual data points. For statistical evaluation, a Kruskal-Wallis test was used followed by Mann-Whitney-U comparison. *, ** or *** indicates p -values smaller than 0.05, 0.01 or 0.001 respectively compared to the other cell sources. EC = Ear Chondrocytes ( n=3 donors) ; NC = Nasal Chondrocytes ( n=3 donors) ; AC = Articular Chondrocytes ( n=3 donors) ; BMSC = Bone-marrow-derived Mesenchymal Stem Cells ( n=3 donors) ; AMSC = Adipose-tissue-derived Mesenchymal Stem Cells ( n=3 donors). Per donor, 2 samples were used for analyses. Cartilage stability Hypertrophic differentiation is an unwanted phenomenon in cartilage regeneration, resulting in cartilage that can remodel into bone when implanted in vivo . To evaluate hypertrophy in vitro , we have studied gene expression of a panel of three hypertrophic markers during five weeks of cell culture (i.e. COL10 , ALP and MMP13 ; Figure 4A). Cultured with TGFβ1 COL10 expression was highest in NCs ( p <0.05) and BMSCs ( p <0.001), and was minimally expressed by ACs. MMP13 was expressed by all cells and significantly highest in NCs. ALP was significantly higher in constructs with BMSCs compared to the other cell sources (p <0.05). In addition, constructs with BMSCs already expressed high COL10 and ALP after two weeks of culture indicating early hypertrophic differentiation. (Data not shown) Although BMSCs expressed all hypertrophic markers in vitro , they did not mineralize or form bone after eight weeks of subcutaneous implantation. Also, no signs of tissue calcification or bone formation were observed in construct containing AMSCs. On the contrary, 100% (3/3) of the cell-free constructs and, unexpectedly, 58.3% (7/12) of constructs encapsulating ACs did calcify in vivo . Also, calcification was more often seen in constructs pre-cultured in control medium (without TGFβ1) compared to constructs cultured in chondrogenic medium (with TGFβ1). (Figure 4B) Cartilage structure and functionality The elastic modulus of constructs was low in vitro , irrespective of the cell source used, being on average 7.42 ± 2.10 kPa. However, after subcutaneous implantation, mechanical properties improved in constructs containing either ECs (23.68 ± 10.20) or NCs (55.12 ± 59.25), but was not perceived in constructs containing ACs, BMSCs or AMSCs. (Figure 5) Since tissue calcification misrepresents the biomechanical properties of the cartilage matrix; we excluded calcified cartilage constructs from further analyses. To determine whether the mechanical properties were enhanced by the deposition of matrix components, a multiple regression analyses was performed for all cell sources separately using sGAG and collagen deposition as independent variables. Only for constructs containing NCs, matrix components significantly associated with the biomechanical functionality of the constructs ( R 2 =0.477, F =4.558, p =0.039). For these constructs, only sGAG- deposition associated significantly with the biomechanical properties of the cartilage constructs independently (sGAG: β =0.689, p =0.013; collagen: β =0.044, p =0.851). 75 CARTILAGE-FORMING CAPACITY OF SEVERAL CELL SOURCES 4