As shown in Figure D we

As shown in Figure 1D, we propose an optimized strategy for promoting osteoblast differentiation from mESCs under chemically defined conditions. This strategy consists of three phases: the maintenance of mESCs using 2i plus leukemia inhibitory factor (LIF) culture, mesoderm induction by CHIR in combination with Cyc-mediated suppression of neuro-ectoderm differentiation, and osteoblast induction by SAG and TH. Given that Cyc induces SMO translocation to the primary cilium despite its inhibitory effect on Hh signaling (Wang et al., 2009), the accumulated SMO in the cilium may result in increased sensitivity to subsequent SAG treatment, which may also contribute to the osteoblast differentiation by SAG and TH during the osteoblast induction phase. SAG and TH cooperatively induce osteoblast differentiation, possibly through SAG-mediated specification into an osteoblast lineage and TH-mediated promotion of late osteoblast differentiation (Hojo et al., 2013; Maeda et al., 2013; Ohba et al., 2007b). Gene expression patterns in mESCs cultured according to this strategy are shown in Figure 2A. Pou5f1, Nanog, and Sox2 were significantly downregulated on day 19, indicating that the mESCs gradually exited from the pluripotent state as differentiation progressed. T and Mixl1 were upregulated by mesoderm induction (day 5) and in turn were downregulated by osteoblast induction (day 19). Sox1, Sox17, and Gata4, which are ectoderm and endoderm marker genes, were not altered throughout the culture. These data suggested that the present strategy specifically directs pluripotent KPT-185 toward a mesodermal cell fate. The osteoblast-related genes, Runx2, Sp7, Col1a1, and Ibsp, were upregulated during the osteoblast induction phase (day 19) relative to day 0; Runx2, Sp7, Col1a1, and Ibsp were upregulated approximately 128-, 91-, 7,680-, and 147,300-fold, respectively. However, the 19-day culture was not long enough for the cells to express Bglap, a bona fide marker of mature osteoblasts; culturing the cells for an additional 4 days in the absence of SAG or TH induced a 145-fold upregulation of Bglap. Expressions of the osteoblast-related genes in the induced cells were higher than or comparable to those in cultured mouse primary osteoblasts (Figure 2A). Expressions of osteoblast-related transcription factors in the induced cells were comparable to those in freshly isolated mouse osteoblasts (fresh mObs), although the expressions of matrix genes in the induced cells were not as high as those in the fresh mObs (Figure S2A). When gene expressions were compared between the present strategy and an EB-based conventional one (Kawaguchi et al., 2005), the present strategy induced significantly lower expressions of Sox17 and Gata4 on day 19 and higher expressions of Runx2, Sp7, Col1a1, Ibsp, and Bglap on day 19 or 23 than the conventional method (Figure S2B). The present strategy is likely to differentiate mESCs into an osteoblast lineage more specifically than the conventional method, at least partly by avoiding their differentiation into unfavorable lineages. When we examined gene expressions at several time points during the culture (Figure S2C), the expression profile resembled a proposed model for the Hh signaling-mediated specification of skeletal progenitors into Runx2-positive osteoblast precursors and their subsequent differentiation into osteoblasts in skeletal development (Rodda and McMahon, 2006). Thus, we established a strategy to direct mESCs toward a mature osteoblast cell fate by sequentially using four small molecules, CHIR, Cyc, SAG, and TH, under chemically defined conditions. This protocol involves the addition of a fourth osteoblast maturation phase to the three-phase strategy described earlier (Figure 2B). The effectiveness of the strategy was further confirmed by the expression of GFP on days 19 and 23 in mESCs engineered to express GFP under the control of an osteoblast-specific rat 2.3 kb Col1a1 promoter (2.3 kb Col1a1-GFP mESCs) (Figure 2B) (Ohba et al., 2007a).