Objectives:Chondrogenic mesenchymal cells have been developed from human pluripotent stem cells (hPSCs) as a human model of embryonic chondrogenesis. These cells are expandable in the presence of fibroblast growth factor (FGF), and the types of cartilage they tend to generate have been characterized in vitro and in vivo. Resistance to endochondral ossification is one of the most important characteristics of the tissue engineered cartilage for cartilage regenerative therapy. Despite some reports on directed derivation of endochondral ossification resistant chondrocytes from hPSCs, their precursors (chondrogenic mesenchymal cells) have been poorly characterized. We have developed methods to generate two types of chondrogenic mesenchymal cells that lead to cartilage pellets expressing high or very low levels of hypertrophic chondrocyte markers in vitro. These cartilage pellets, when transplanted, became either fully mineralized bony tissue or remained unmineralized, respectively. Here, we present results from cell-type analyses of the two types of mesenchymal cells using the genome-wide RNA-seq technology, aiming to provide mechanistical insights into how the two types of mesenchymal cells prefer to form different types of cartilage.Methods:Human PSC lines were differentiated toward paraxial mesodermal progeny in a chemically-defined medium (CDM) as previously described (1, 2), and mesodermal cells were isolated by cell sorting (2, 3). The isolated cells were then cultured in CDM containing FGF2, platelet-derived growth factor (PDGF), transforming growth factor-beta (TGF-beta) inhibitor and glycogen synthase kinase (GSK) inhibitor (1), which were then subjected to CDM containing FGF2, TGFbeta inhibitor and GSK inhibitor (FSbC medium) as well as to CDM containing PDGF and BMP inhibitor (PN medium) to generate mesenchymal cells of distinct size and shape. The resulting cells were subjected to RT-PCR and bulk RNA-seq analyses for cell-type determination. Chondrogenesis was done by pellet culture using PDGF, TGF-beta3 and bone morphogenetic proteins, as described previously (1, 2). Some cartilage pellets were fixed, sectioned, and stained with Toluidine Blue and von Kossa, or immunostained with anti-collagen type I, II and X (COL1, 2, 10) antibodies. DNA, RNA, and sulfated glycosaminoglycan were isolated and quantified from unfixed cartilage pellets. Some cartilage pellets were also subcutaneously transplanted into NSG (severely immunocompromised) mice and the recovered cartilage pellets were similarly analyzed. All the methods involved were previously described in detail (1, 2).Results:The PN cultured mesenchymal cells (Fig. 1) developed COL2+ cartilage pellets that expressed COL10A1 at 10 to 100-fold lesser levels and PRG4 (Lubricin gene) at higher levels than those derived from the FSbC cultured mesenchymal cells, and were resistant to endochondral ossification after subcutaneous transplantation for 8 weeks (Fig. 2) (1, 2). In contrast, the FSbC cultured mesenchymal cells (Fig. 1) developed cartilage pellets that expressed COL10A1 at higher levels than those derived from the PN cultured mesenchymal cells, and were readily mineralized after subcutaneous transplantation (Fig. 2) (1, 2). The comparative RNA-seq analyses of the two mesenchymal cell populations showed that the FSbC cells resembled ectomesenchymal cells derived from neural crest progeny of hPSCs (1, 2) and the PN cells consisted of cells resembling mesenchymal stromal cells and ligament/tendon progenitors (Fig. 3).Conclusions:From hPSCs, chondrogenic mesenchymal cells of distinct chondrogenesis activity can be generated in vitro in a controlled fashion. One type leads to chondrocytes prone to endochondral ossification, and the other type leads to rather stable cartilage that seems to resist endochondral ossification. Our results suggest that the fate of hPSC-derived chondrocytes can be controlled at the mesenchymal cell (i.e., chondrocyte precursor)-stage, which seems to be dependent on how the mesenchymal cells are generated from hPSCs and maintained in culture. Further biological studies on these cells will not only enable hPSC-derived chondrocytes/chondroprogenitors to be used directly for cartilage regenerative therapy, but also may lead to a critical mechanism that allows therapeutically relevant adult chondrogenic cells such as skeletal stem cells to reproducibly regenerate hyaline permanent cartilage during cartilage repair.Fig. 1Fig. 2Fig. 3