Archives
br Materials and methods br Results br Discussion
Materials and methods
Results
Discussion
While ASCs are considered a more accessible source of ddr1 than MSCs, both cells exhibit similar properties and are considered as promising tools for tissue engineering and cell therapy applications (Maumus et al., 2011a). Previously, a number of studies investigated the cross-talk between MSCs and chondrocytes and most of the studies reported the role of chondrocytes on the differentiation of MSCs. Differentiation of MSCs, showed by up-regulation of Col II and aggregan expression as well as GAG production, occurred via the secretion of factors such as Parathyroid Hormone related Peptide (PTHrP), TGFβ, Insulin Growth Factor (IGF)-1 or Bone morphogenetic protein (BMP)-2 by chondrocytes (Acharya et al., 2012; Aung et al., 2011; Fischer et al., 2010; Liu et al., 2010). Conversely, other studies reported that the addition of MSCs increased cartilage matrix production by chondrocytes in co-culture (Giovannini et al., 2010; Mo et al., 2009; Wu et al., 2011, 2012). However, all these reports used MSCs in direct contact with chondrocytes cultured in pellet and under chondroinductive conditions. In these studies, the addition of fat- or BM-derived MSCs resulted in the absence or slight increase of cartilage matrix production by chondrocytes or enhanced proliferation. We therefore investigated the role of ASCs on chondrocytes using a co-culture assay in which OA chondrocytes are cultured in monolayer which is a condition close to the altered phenotype occurring during OA and in a minimal medium to see a direct effect of ASCs, independently of FCS or growth factors contained in chondroinductive medium. We also cultured chondrocytes in a transwell culture system to avoid direct contact with ASCs and evaluated the effect of diffusible molecules. Contrary to previous reports, in our conditions, we did not observe an increased proliferation rate or enhanced production of cartilaginous matrix by chondrocytes, likely due to the absence of chondro-inductive components. However, we were able to rather demonstrate a dual protective role of ASCs on chondrocytes, which was not described previously. ASCs decrease the expression of fibrotic markers, indicative of dedifferentiation towards fibroblast-like cells, and of hypertrophic markers, both hallmarks of OA chondrocytes. These properties were broadly shared by mesenchymal stem cells isolated from bone marrow or fat from different locations. ASCs from diverse fat tissue locations were tested because they display distinct secretomes and may differ in terms of functional properties (Jurgens et al., 2008, 2009; Schipper et al., 2008). Our findings therefore validate the interest of using mesenchymal stem cells, in particular from adipose tissue, not only for chondroinductive effect as shown by some other studies but also for a chondroprotective role decreasing the propensity of OA chondrocytes to become fibrotic or hypertrophic. Moreover, the present study provides mechanistic insight with which MSCs or ASCs may exert their therapeutic effect in vivo when implanted in the joints of animals with osteoarthritis (Diekman et al., in press; Murphy et al., 2003; Ter Huurne et al., 2012).
We showed that the mechanism of chondroprotection is associated with the down-regulation of TGF-β1 in chondrocytes. Although not investigated and not demonstrated here, the reduced TGF-β1 secretion by chondrocytes when co-cultured with ASCs might down-regulate ALK1-Smad1/5/8 signaling that has been described to induce Runx2 activation and MMP13 secretion in OA chondrocytes and therefore explain the lower expression of hypertrophic markers (van der Kraan et al., 2012). Since TGF-β1 is also known to be pro-fibrotic, its down-regulation might also explain the anti-fibrotic effect of ASCs (Biernacka et al., 2011). Together with the decrease of TGF-β1 in chondrocytes, we observed the increase of HGF secretion by ASCs. HGF, which is mainly secreted by stromal cells, is known to participate in tissue protection and regeneration, promoting cell survival, proliferation and stimulating extracellular matrix degradation (Nakamura et al., 2011). The anti-fibrotic effect of MSCs through HGF secretion is documented in several diseases such as heart failure, bladder or kidney fibrosis but was never reported in OA (Liu et al., 2011; Shabbir et al., 2009; Song et al., 2012). Although it was reported that chondrocytes do not produce functional HGF (Bau et al., 2004; Guevremont et al., 2003; Pfander et al., 1999), the protein is detected in synovial fluid and its concentration is positively correlated with the severity of OA (Dankbar et al., 2007) while HGF increases proteoglycan synthesis in rabbit chondrocytes (Bau et al., 2004; Takebayashi et al., 1995). In our settings, we showed that HGF mediates the decrease of Col type I and III expression by OA chondrocytes. Moreover, up-regulation of HGF secretion by ASCs co-cultured with chondrocytes argues for a cross-talk between the two cell types. Even though the cross-talk clearly influences the range of HGF secretion, it is likely not required for a measurable effect on chondrocytes, as suggested by another study in a model of heart failure where the intramuscular injection of MSC conditioned medium allowed a regenerative process on the infarcted myocardium (Shabbir et al., 2009). The mechanism by which HGF mediates an anti-fibrotic role in chondrocytes has still to be investigated. However, possible explanation may be through the induction of connective tissue growth factor (CTGF), which suppresses collagen I synthesis (Inoue et al., 2003). It may also be explained by the induction of various matrix metalloproteinases, in particular MMP-1, -2, -9 which are crucial in matrix remodeling by denaturing collagens type I, III or IV (Lee et al., 2011; Sherriff-Tadano et al., 2006). While we identified HGF as a secreted factor playing a key role in the anti-fibrotic effect of ASCs on OA chondrocytes, we cannot however exclude that other released factors may contribute to this effect. Proteomic analysis of supernatants from co-culture and that of cells cultured alone might help identifying other molecules potentially involved in the chondroprotective role of ASCs on chondrocytes (Skalnikova et al., 2011).