Since the first report of induced

Since the first report of induced pluripotent stem (iPS) purchase CB-5083 generated by the expression of four transcription factors Oct4, Sox2, Klf4 and c-Myc (OSKM) in fibroblasts (Takahashi and Yamanaka, 2006), there has been a variety of technical improvements to generate iPS cells more efficiently. Different combinations of reprogramming transcription factors or addition of synthetic chemicals that influence epigenetic regulation and signaling pathways have been discovered to enhance the reprogramming efficiency and quality of stem cells (Feng et al., 2009b; Han et al., 2010; Lin et al., 2009). However, the mechanism of reprogramming, especially at the early phases, remains elusive. One of the reasons for the scant knowledge about reprogramming mechanisms is the very low efficiency of reprogramming. Only a small number of cells, which is undergoing reprogramming, need to be isolated out of heterogeneous cell populations for the analysis. Although the expression of fluorescent proteins using Oct4 or Nanog promoters has been used to detect and isolate mature iPS cells in 2–4weeks of time, there has been no efficient way to probe the cells undergoing reprogramming at early phases. Profiling gene expression patterns during reprogramming has previously been attempted using unsorted mixed populations of secondary mouse embryonic fibroblasts (MEFs) harboring doxycycline-inducible OSKM genes (Samavarchi-Tehrani et al., 2010) or partially reprogrammed cell lines which were not actively undergoing reprogramming (Mikkelsen et al., 2008). Recently, Buganim et al. identified single cells which turn into iPS cells and analyze expression profiles of 48 genes in those cells (Buganim et al., 2012), while Polo et al. analyzed global gene expression profiles of the cells enriched based on SSEA-1 and Thy1 expressions (Polo et al., 2012) and Hansson et al. analyzed proteome changes during reprogramming under dox-inducible transgenic system (Hansson et al., 2012). The reprogramming mechanisms newly revealed by these studies demonstrate the necessity of more tools with which the rare cells undergoing reprogramming can be isolated. We previously reported a fluorescent chemical probe CDy1, which selectively stains living embryonic stem (ES) cells of both human and mouse origin and enables the isolation of ES cells from a mixture with fibroblasts by fluorescence-activated cell sorting (FACS) (Im et al., 2010; Kang et al., 2011). CDy1 detects iPS cells as well, at both early and late phases of reprogramming as demonstrated using the cells of a transgenic mouse that express GFP under the control of Oct4 promoter. When the mouse fibroblasts were transfected with retroviral vectors encoding OSKM and then incubated with CDy1, some colonies were brightly stained at 10days post infection (dpi) before GFP expression, which eventually turned into GFP-expressing colonies at a later time. In another study, we demonstrated that CDy1 can be used as an iPS cell reporter even at 7dpi in a high throughput screening of chemicals designed for the development of reprogramming enhancer (Vendrell et al., 2012). These findings led us to hypothesize that cells undergoing reprogramming could be isolated at different time points during reprogramming using CDy1 and their stepwise global gene expression analysis would enable the identification of key mechanisms, pathways or molecules that play important roles in cellular reprogramming. Based on this hypothesis, we identified differentially expressed genes (DEGs) in the CDy1-positive cells harvested at several early time points of reprogramming starting from as early as 3dpi and analyzed the data using various bioinformatics tools. A functional transcriptomic analysis of the data revealed an unprecedented sequence of cellular mechanisms of reprogramming and a DEG network analysis identified PDGF-BB as a critical factor for the process.
Materials and methods
Results
Discussion The low efficiency of somatic cell reprogramming by the expression of reprogramming factors delivered by viral vectors may be attributed to variations in transfection efficiency and transgene integration loci. Although the reprogramming efficiency significantly improved by drug-induced transgene expressions in genetically identical secondary MEF, it still remains as low as 4% (Wernig et al., 2008). Stochastic processes governed by factors in epigenetic regulation and cell division have been proposed to account for the different responses of each somatic cell to the reprogramming factors (Hanna et al., 2009). Despite the important role of stochasticity in reprogramming, the highly transient expression dynamics observed from our study are remarkably consistent with the robust modulation of cellular functions at critical time points. The response timings for each gene depend on its specific multi-functional roles, and the combinatorial requirements of concurrent programs. Together, genes participating in simultaneous programs may drive, coordinate, explore and respond to the reprogramming landscape, all at the same time. As such, gene expression involved can exhibit highly stochastic, complex and yet robust dynamics. As an example to illustrate the coordinated nature of reprogramming, we noted the programs regulating signaling, adhesion, cytoskeleton and extracellular matrix formation together with specific proliferation-related ones are likely to be initially activated to reset the plasticity of the transcriptome and proteome (Hansson et al., 2012), in response to OSKM activities. Subsequently, they are turned off and are distinct from the global metabolism and proliferation programs that become inbuilt later with the progression of reprogramming. The delayed proliferation programs include a rapid decline in cell size (Smith et al., 2010) and the distinct expression patterns of cell-cycle proteins (Polo et al., 2012; Ruiz et al., 2011).