In a healthy joint, mechanical loading increases matrix synthesis and maintains cell phenotype, while reducing catabolic activities. It activates several pathways, most of them yet largely unknown, with integrins, TGF-β, canonical (Erk 1/2) and stress-activated (JNK) MAPK playing a key role. Degenerative joint diseases are characterized by Wnt upregulation and by the presence of proteolytic fibronectin fragments (FB-fs). Despite they are known to impair some of the aforementioned pathways, little is known on their modulatory effect on cartilage mechanoresponsiveness. This study aims at investigating the effect of mechanical loading in healthy and in vitro diseased cartilage models using pro-hypertrophic Wnt agonist CHIR99021 and the pro-catabolic FB-fs 30 kDa.

Human primary chondrocytes from OA patients have been grown in alginate hydrogels for one week, prior to be incubated for 4 days with 3μM CHIR99021 or 1 μM FB-fs. Human cartilage explants isolated from OA patients have incubated 4 days with 3 μM CHIR99021 or 1 μM FB-fs. Both groups have then been mechanically stimulated (unconfined compression, 10% displacement, 1.5 hours, 1 Hz), using a BioDynamic bioreactor 5270 from TA Instruments. Expression of collagen type I, II and X, aggrecan, ALK-1, ALK-5, αV, α5 and β1 integrins, TGF-β1 have been assessed by Real Time-PCR and normalized with the expression of S29. Percentage of phosphorylated Smad2, Smad1 and JNK were determined through western blot. TGF-β1 content was quantified by sandwich ELISA; MMP-13 and GAG by western blot and DMMB assay, respectively. At least three biological replicates were used. ANOVA test was used for parametric analysis; Kruskal-Wallis and Mann-Whitney post hoc test for non-parametric.

Preliminary data show that compression increased collagen II expression in control, but not in CHIR99021 and FB-fs pre-treated group (Fig. 1A-B). This was associated with downregulation of β1-integrin expression, which is the main collagen receptor and further regulates collagen II expression, suggesting inhibition of Erk1/2 pathway. A trend of increase expression of collagen type X after mechanical loading was observed in CHIR and FB-fs group. ALK-1 and ALK-5 showed a trend toward stronger upregulation in CHIR99021 group after compression, suggesting the activation of both Smad1/5/8 and Smad 2/3 pathways. To further investigate pathways leading to these different mechano-responses, the phosphorylation levels of Smad1 and Smad2, Erk1/2 and JNK proteins are currently being studied. Preliminary results show that Smad2, Smad1 and JNK protein levels increased in all groups after mechanical loading, independently of an increase in TGF-β1 expression or content. Compression further increased phosphorylation of Smad2, but not of Smad1, in all groups.

Mechanical loading regulates the metabolism of chondrocytes in cartilage1. Nowadays, studies exploring the in vitro response of cartilage towards loading often rely on bioreactor experiments applying only compressive loading. This is likely not sufficiently representative for the complex multi-directional loading profile in vivo (i.e. where typical compressive and shear loading are both present). The impact of multi-axial loading is specifically relevant in the context of the onset of osteoarthritis (OA) due to joint destabilization. Here, alterations in the 3D loading profile, and in particular increased shear forces, are suggested to initiate catabolic molecular responses leading to cartilage degeneration3. However, in vitro/ex vivo data confirming this hypothesis are currently lacking. Therefore, we aim to investigate how increased shear loading affects the metabolism and ECM deposition of a healthy chondrogenic cell line and if this response is different in osteoarthritic primary chondrocytes.

A murine chondrogenic precursor cell line (ATDC5) and primary human osteoarthritic articular chondrocytes (hOACs) were encapsulated in 2.2% alginate disks and cultured in DMEM medium for three days. Hydrogels seeded with the different cell groups were loaded in the TA ElectroForce BioDynamic Bioreactor and subjected to following loading conditions: (a) 10% compression at 1Hz for 1h, (b) 10% compression and 10° shear loading at 1Hz for 1h. Unloaded constructs were used as control. After loading, hydrogel constructs were stabilized in culture medium for 2 hours, to facilitate adequate gene expression responses, before being dissolved and snap frozen. RNA was isolated and gene expression levels specific for anabolic pathways, characterized by extracellular matrix (ECM) genes (Col2a1, Aggrecan and Perlecan), catabolic processes (MMP-3 and MMP-13) and chondrogenic transcription factor (Sox9) were evaluated using RT-qPCR. The TA ElectroForce BioDynamic Bioreactor was successfully set-up to mimic cartilage loading.

In ATDC5 cells, compression elicits an increase in all measured ECM genes (Col2a1, Aggrecan and Perlecan) compared to unloaded controls, suggesting an anabolic response. This upregulation is decreased when adding additional shear strain. In contrast to ATDC5 cells, the anabolic response of proteoglycans Aggrecan and Perlecan to compressive loading was lower in osteoarthritic chondrocytes, and Col2a1 expression appeared decreased. Adding shear strain reversed this effect on Col2a1 expression. Multi-directional loading increased transcription factor Sox9 expression compared to compression in both ATDC5 and OA chondrocytes. In OA chondrocytes, both loading regimens increased MMP-3 and MMP-13 expression. Shear loading reduces the anabolic effect of compressive loading in both cell types. OA cells presented more catabolic response to mechanical loading compared to precursors, given the increase in catabolic enzymes MMP-3 and MMP-13.

The main limitation of autologous chondrocyte implantation techniques is the necessity for in vitro cell expansion, which is associated with phenotypic drift and loss of extracellular matrix synthesis. Although media supplements (e.g. TGF-β) are extensively used to mitigate the tendency of de-differentiation, the lack of extracellular matrix is still one of the major obstacles to obtaining engineered cartilage substitutes with long-term clinical efficacy. Macromolecular crowding (MMC) is a biophysical phenomenon that increases tissue-specific extracellular matrix deposition. This study aimed to test whether MMC can be used to enhance hyaline-like ECM deposition in human chondrocyte culture: this hypothesis was tested in cells at P2 and at P7. Cells at P2 were cultured using a standard medium (DMEM/F12) in monolayer or alginate beads, whilst cells at P7 were cultured and re-differentiated using the system Clonetics™ of Lonza in the presence of 5 % HS or 5 % FBS, in monolayer and alginate beads. Macromolecular crowding medium was added 14 days after the start of re-differentiation. Collagen deposition was evaluated after 2, 5 and 10 days using SDS-PAGE and immunocytochemistry. MMC enhanced matrix deposition in all the conditions tested. However, although cells at P7 were cultured using a commercially available system, their deposited matrix was richer in collagen type I, whilst collagen type II was barely detectable. This was even more evident for cells in monolayer in HS and indicates that cells acquired a fibroblastic phenotype. To conclude, we showed that MMC increased matrix deposition in chondrocyte culture and that, unfortunately, commercially available systems are not always able to maintain chondrogenic phenotype. Since ECM produced is often undetectable and collagen expression and synthesis are not always correlated with its secretion, we propose to use MMC to assess chondrocyte phenotype maintenance and effectiveness of re-differentiation media.