br Experimental Procedures br Author Contributions br Acknow

Experimental Procedures
Author Contributions
Acknowledgments
Introduction Induced pluripotent stem order orexin (iPSCs) have similar properties as embryonic stem cells (ESCs), such as self-renewal and differentiation capacity (Park et al., 2008c; Takahashi and Yamanaka, 2006). Reprogramming technique offers tremendous potential for disease modeling, cell-based therapy, and drug screening (Park et al., 2008a). Although the reprogramming process is quite robust and applicable to various types of adult differentiated cells, only a small fraction of donor cells reaches a fully pluripotent state, while the majority are refractory to reprogramming. Imperfect reprograming may carry somatic memory and may contribute to cancer development (Ohnishi et al., 2014). Therefore, efficient selection and generation of bona fide iPSCs are essential for safe uses in regenerative medicine. Serial live cell imaging is one of the tools to distinguish bona fide human iPSCs (hiPSCs) from partially reprogrammed cells. Previously, we identified three distinct types of expandable hESC-like colonies during reprogramming via expression patterns of virus-derived GFP, fibroblast marker CD13 (ANPEP), and two pluripotent markers SSEA4 and TRA160 (Chan et al., 2009). Type I cells are defined by continuous expression reprogramming genes (CD13−GFP+SSEA4−TRA160−). Type II cells express pluripotency marker SSEA4 and continue expressing reprogramming factors (CD13−GFP+SSEA4+TRA160−). Type III cells show expression of TRA160 as well as SSEA4 (CD13−GFP−SSEA4+TRA160+). Among these types of colonies, only type III has similar molecular phenotypes with hESCs and become bona fide hiPSCs. Type I and type II cells are partially reprogrammed cells and display negative nuclear NANOG staining, low expression of several pluripotent genes (e.g., DNMT3B and REX1), and a distinct epigenetic state from type III cells and hESCs. Type I cells remain in their incomplete reprogramed state, while a small population of type II cells may still convert to type III cells and complete hiPSC reprogramming. Reprogramming pathways have been extensively studied. Mesenchymal-to-epithelial transition (MET) occurs in the initial phase of reprogramming and is synergistically activated by OCT4, SOX2, KLF4, and MYC (OSKM) and BMP signaling, but is blocked by the transforming growth factor β (TGF-β) pathway (Li et al., 2010; Samavarchi-Tehrani et al., 2010). Despite the active function of BMP in the initial reprogramming, BMP proteins prevent the transition of pre-miPSCs to fully reprogrammed miPSCs by maintaining H3K9 methylation (Chen et al., 2013). In contrast, ACTIVIN/NODAL signaling pathway, which is a branch of TGF-β signaling, is essential for mESC self-renewal (Ogawa et al., 2007). WNT ligands and a downstream component of WNT signaling pathway, β-catenin, are required to prevent differentiation and maintain self-renewal in mESCs (Lyashenko et al., 2011). Whereas the transcriptional repressor TCF3 inhibits mESC self-renewal, an interaction with β-catenin followed by WNT3A stimulation activates the expression of self-renewal genes by blocking the TCF3 repressive activity (Yi et al., 2011). A recent study further defined the role of WNT, revealing that this pathway is a negative regulator in the early stages, but switches to a positive regulator in the late stage of mouse reprogramming (Ho et al., 2013). Transcription profiling during reprogramming has provided critical insights into understanding reprogramming. Microarray-based transcriptome analysis in miPSCs and partially reprogrammed murine cell populations sorted by a fibroblast marker (THY1) and two pluripotent markers (SSEA1 and Oct4-GFP) revealed that the reprogramming process is composed of two main transcriptional waves (Polo et al., 2012). The first wave is driven by Myc and Klf4 and characterized by the loss of fibroblast identity and a gain in cell proliferation. The second wave is controlled by Oct4, Sox2, and Klf4 and is associated with changes in DNA methylation that facilitate stable pluripotency. A microarray and single-cell qPCR study of cell populations sorted by virus-driven EGFP and TRA160 in hiPSC reprogramming, showed that TRA160+ cell populations at late time points (approximately day 28) exhibit more similar gene expression patterns to hESCs and less heterogeneous than those at early time points (approximately day 11) (Tanabe et al., 2013). However, most of the nascent TRA160+ cells fail to complete reprogramming. These recent reports indicate that transcriptional and signaling regulatory networks are different among intermediate steps.