br Materials and methods br Results and discussion In order

Materials and methods
Results and discussion In order to evaluate the staining properties of CDy1 in a format compatible with high-throughput screening, we transduced MEFs with different combinations of Oct4, Sox2, Klf-4 and c-Myc and seeded them in 96 or 384-well plates using a 4:1 (transduced:feeder) cell ratio. After 7days post infection (dpi), we incubated the bet inhibitor with CDy1 and observed them with an ImageXpressMICRO system for high-throughput fluorescence cell-imaging data acquisition after washing (Fig. 1A). As shown in Fig. 1A, the average fluorescence intensity in cells infected with Oct4, Sox2, Klf-4 and c-Myc (OSKM) was significantly higher than in cells infected with different combinations of Oct4, Sox2 and Klf-4 (OSK, OS, OK, and KS) and non-infected cells. OSKM-transduced cells may render a significant population of non-iPSC colonies (Nakagawa et al., 2008) so we analyzed the fluorescence cell images and observed that CDy1 stained cells resembling early iPSC colonies but not all clumps of cells or colony-like structures (Fig. 1B). These results confirmed that CDy1 could be efficiently used as a fluorescent reporter to discriminate early stages of reprogramming in a high-throughput screening format. The evaluation of small molecules that can substitute exogenous transcription factors in iPSC generation is focused on the replacement of c-Myc, as its oncogenic role hampers the potential therapeutic application of iPSCs (Huangfu et al., 2008; Shi et al., 2008a, 2008b). Since CDy1 proved to discriminate well the reprogramming states of OSK and OSKM-infected cells at 7dpi (Fig. 1A), we designed a CDy1-based screening to identify small molecules that could replace c-Myc in reprogramming (Okita et al., 2007) (Fig. 1C). OSK-transduced cells were seeded at 4000 cells/well into 384-well plates with mitomycin-treated MEF as feeder cells (1000 cells/well) and incubated in mouse ESC (mESC) media containing leukemia inhibitory factor (LIF, 100U/mL) in the presence of a collection of newly synthesized hydroxamic acids at a final concentration of 5μM (0.5% DMSO). After 5days of incubation with the hydroxamic acids (i.e. 7dpi), cells were treated with CDy1, washed with mESC media and analyzed by high-throughput fluorescence cell-imaging. OSK-transduced cells treated with 0.5% DMSO and OSKM-transduced cells were included as controls for non- and full-replacement of c-Myc respectively. Upon evaluation of 240 hydroxamic acids, we identified 5 compounds inducing higher fluorescence intensities in OSK-infected cells than DMSO and with comparable responses to various reported chemicals enhancing reprogramming (e.g. epigenetic modulators: 5′-azacytidine (5′-azaC) and trichostatin A (TSA) as well as TGFβ inhibitors: SB431542) (Figure S1 in Supporting Information (SI)). Secondary screenings showed that 1–26 yielded the largest population of CDy1-stained cells in a reproducible manner and suggested that it could partially replace c-Myc to a similar extent than reported epigenetic modulators, such as 5′-azaC (Huangfu et al., 2008) (Figure S2 in SI). After identifying 1–26 from the CDy1-based primary screening, we aimed to confirm the effect of 1–26 (Fig. 2A) in the reprogramming efficacy using an alternative reported method. The Oct4-GFP transgene has been extensively used to monitor the generation of miPSCs, so we examined OSK-transduced cells harboring an Oct4-GFP reporter upon continuous incubation with 1–26. The treatment of Oct4-GFP OSK-transduced cells with 1–26 or 5′-azaC enhanced the formation of GFP+ colonies at an earlier time point than DMSO-treated cells with an overall 2.5-fold increase in the reprogramming efficiency (Fig. 2B). Furthermore, we investigated the biological activity of 1–26 that may be responsible for this effect. Some hydroxamic acids (e.g. TSA, suberoylanilide hydroxamic acid) have been reported as HDAC inhibitors (Finnin et al., 1999; Biel et al., 2005), which can improve the generation of iPSC reprogramming by inducing chromatin modifications (Huangfu et al., 2008). 1–26 behaved as a HDAC inhibitor at micromolar concentrations (Fig. 2C), which correlated well with its reprogramming effect in OSK-transduced cells.