Mixed populations of PSCs were also converted to naive

Mixed populations of PSCs were also converted to naive PSCs in 2i + LIF medium without serum or to primed EpiLCs in bFGF + activin medium. EpiLCs induced from ESCs expressed only RFP, and the RFP+ EpiLCs were indistinguishable from EpiSCs in terms of morphology, gene expression patterns, epigenetic status, and defective incorporation into ICM of blastocysts. Therefore, GFP−RFP+ state constitute a definitive marker of primed PSCs; GFP−RFP+ ack1 inhibitors were not observed in any of the three different ESC culture media. The conversion from GFP+RFP+ to GFP+RFP− cells and vice versa takes place within 2 days of changing the culture medium (Figure 2), indicating that the conversion of enhancer activity occurs within 2 days. It has been suggested that genome-wide demethylation and transcriptional changes occur in mouse ESCs within the first 24 hr of 2i addition (Ficz et al., 2013). Therefore, the regulatory machinery of Oct4 enhancers seems to change rapidly in response to external cues accompanying epigenetic modification. GFP−RFP+ EpiLCs derived from transgenic ESCs showed the molecular signature of EpiSCs, including expression of EpiSC markers and hypermethylation of germ cell markers (Stella and Dppa5 region), LINE1, and IAP. Primed PSCs are derived from post-implantation stage epiblasts and express Oct4 for self-renewal under the control of the PE (Brons et al., 2007; Tesar et al., 2007). RFP+ cells were not incorporated into the ICM by aggregation with normal embryos. Although a single aggregated embryo with RFP+ cells was observed, the GFP−RFP+ cells did not contribute to the ICM (Figure S6). This result could support the observation that GFP−RFP+ cells formed a pure population of primed PSCs. Finally, we showed that the activity of the two Oct4 cis-regulatory elements was controlled by DNA methylation and histone modification concurring with the naive or primed pluripotency state. Consequently, the differential activity of DE and PE was closely related to binding affinity of Nanog on the two enhancers; Nanog was highly enriched on the DE in naive PSCs, but highly enriched on the PE in primed PSCs (Figures 6C and 7B). We also found that the activity of Oct4 DE and PE is affected by the repressive histone marks, H3K9me3 and H3K27me3, and DNA methylation in a cell-type-specific manner. The enhancers of Oct4 in primed PSCs is regulated by DNA methylation as well as H3K9me3, but those in naive PSCs is regulated by H3K9me3 and H3K27me3 but not by DNA methylation; DNA methylation was only correlated with inactive DE in primed PSCs and not with inactive PE in naive PSCs (Figure 7B). Thus, demethylation of the PE region seems to be the default state for PSCs and methylation of DE could constitute an epigenetic marker for primed PSCs. In contrast, H3K27ac was enriched in both the DE and PE of naive and primed PSCs, indicating that H3K27ac cannot be a histone mark for the active state of the Oct4 DE and PE. Another explanation is that although one of the enhancers is not active based on fluorescent reporter expression, epigenetically these enhancer elements of Oct4 are basically not completely silenced as H3K27ac remains enriched in both naive and primed PSCs. Similarly the PE is not methylated in 2i/LIF conditions, although the fluorescent mark is not on. These results suggest that overall both enhancers are active and never fully silenced in any of the states, although their level of activity is different under the tested conditions. Thus we should be more careful against using of terms “off” and “on” when describing enhancer activity of Oct4, the more accurate term rather being “dominant activity.” This fact could also be the reason for the need of a dual reporter system for accurate separation of naive and primed PSCs. In this study, we showed that enhancer-specific regulation of Oct4 could constitute a determinant for distinguishing between naive and primed pluripotency. This is based on the ability to accurately separate naive and primed PSCs using the dual reporter system, thereby obtaining pure populations of naive and primed pluripotent cells, which could be used for accurate analysis of distinct cell states. The dual reporter system could also be a useful tool for monitoring cellular reprogramming to naive or primed states as well as embryonic development under live-cell conditions.