br Experimental Procedures br Acknowledgments br Introductio
Introduction Articular cartilage covers the ends of bones and provides shock RG7388 and lubrication to diarthrodial joints. Articular cartilage is a highly specialized tissue composed of chondrocytes and a specific extracellular matrix (ECM) that consists of types II, IX, and XI collagen and proteoglycans, but not type I collagen. Such cartilage is called hyaline cartilage. Focal defects or degeneration of articular cartilage due to trauma or regional necrosis can progressively degenerate large areas of cartilage owing to a lack of repair capacity. These conditions ultimately result in a loss of joint function, inducing osteoarthritis. Autologous chondrocyte transplantation is a successful cell therapy for repairing focal defects of articular cartilage. However, this approach suffers from the need to sacrifice healthy cartilage for biopsies and the formation of fibrocartilaginous repair tissue containing type I collagen (Roberts et al., 2009), because the required in vitro expansion induces the dedifferentiation of chondrocytes toward fibroblastic cells. In addition, it is difficult to achieve the integration of repair tissue into the adjacent native cartilage. Other attractive cell sources for repairing cartilage defects include mesenchymal stem cells (MSCs). However, MSCs can differentiate into multiple cell types, resulting in a mixture of cartilaginous tissue, fibrous tissue (as indicated by the expression of type I collagen), and hypertrophic tissue (as indicated by the expression of type X collagen) (Mithoefer et al., 2009; Steck et al., 2009). Despite the ability to achieve short-term clinical success, non-hyaline repair tissue is eventually lost, because it does not possess the proper mechanical qualities. Currently, a new option for repairing defects in cartilage has become available by applying human induced pluripotent stem cells (hiPSCs) with self-renewal and pluripotent capacities without ethical issues. It has been reported that both human embryonic stem cells (hESCs) and hiPSCs can be differentiated into chondrogenic lineages (Barberi et al., 2005; Vats et al., 2006; Koay et al., 2007; Hwang et al., 2008; Bigdeli et al., 2009; Nakagawa et al., 2009; Bai et al., 2010; Oldershaw et al., 2010; Toh et al., 2010; Medvedev et al., 2011; Umeda et al., 2012; Wei et al., 2012; Koyama et al., 2013; Cheng et al., 2014; Ko et al., 2014; Zhao et al., 2014). However, the purity and homogeneity of the resultant cartilage vary, and in vivo transplantation studies have not investigated the risk of teratoma formation systematically. The transplantation of inappropriately differentiated embryonic stem cells (ESCs) results in teratoma formation and tissue destruction at implanted sites, as shown in experiments using murine ESCs (Wakitani et al., 2003; Taiani et al., 2010). The transplantation of hiPSC-derived cells also carries the risk of tumor formation in association with the artificial reprogramming process (Okita et al., 2007; Yamashita et al., 2013). Therefore, an optimized protocol for driving hiPSC differentiation toward chondrocytes that generates pure cartilage without tumor formation in vivo is needed. In this study, we aimed to generate hiPSC-derived cartilage that exhibits the ability to (1) generate pure cartilage in vivo, (2) integrate neocartilage into the adjacent native articular cartilage, and (3) produce neither tumors nor ectopic tissue formation when transplanted in immunodeficiency animals. We therefore developed a chondrogenic differentiation method by taking advantage of real-time monitoring of the chondrocytic phenotype of cells derived from COL11A2-EGFP hiPSCs. We then examined whether the resultant hiPSC-derived cartilage met the above specifications using an animal transplantation model.
Discussion BMP2, TGF-β1, and GDF5 were required to obtain GFP-positive cells from COL11A2-EGFP hiPSC-derived mesendodermal cells. Plural receptors for BMPs (BMPRs) have been identified, and the affinity for these receptors has been shown to differ between BMPs and GDF5 (Nishitoh et al., 1996). In addition, BMPRIA and BMPRIB regulate distinct processes in the formation and differentiation of cartilage (Zou et al., 1997), and BMP and GDF family members have distinct functions in cartilage formation when overexpressed in transgenic mice (Tsumaki et al., 2002). Furthermore, both BMPRIA and BMPRIB are necessary for cartilage formation (Yoon et al., 2005). These findings are consistent with our results showing that both BMP2 and GDF5 are necessary for the differentiation of hiPSCs toward mature chondrocytes. It is also possible that BMP2, TGF-β1, and GDF5 were each required in our protocol, because we used iPSC-derived mesendodermal cells instead of more mature mesodermal tissues to induce chondrogenesis.