Scholarly article on topic 'Hierarchical Oct4 Binding in Concert with Primed Epigenetic Rearrangements during Somatic Cell Reprogramming'

Hierarchical Oct4 Binding in Concert with Primed Epigenetic Rearrangements during Somatic Cell Reprogramming Academic research paper on "Biological sciences"

CC BY
0
0
Share paper
Academic journal
Cell Reports
OECD Field of science
Keywords
{Oct4 / reprogramming / "histone modification" / iPSC}

Abstract of research paper on Biological sciences, author of scientific article — Jun Chen, Xiaolong Chen, Min Li, Xiaoyu Liu, Yawei Gao, et al.

Summary The core pluripotency factor Oct4 plays key roles in somatic cell reprogramming through transcriptional control. Here, we profile Oct4 occupancy, epigenetic changes, and gene expression in reprogramming. We find that Oct4 binds in a hierarchical manner to target sites with primed epigenetic modifications. Oct4 binding is temporally continuous and seldom switches between bound and unbound. Oct4 occupancy in most of promoters is maintained throughout the entire reprogramming process. In contrast, somatic cell-specific enhancers are silenced in the early and intermediate stages, whereas stem cell-specific enhancers are activated in the late stage in parallel with cell fate transition. Both epigenetic remodeling and Oct4 binding contribute to the hyperdynamic enhancer signature transitions. The hierarchical Oct4 bindings are associated with distinct functional themes at different stages. Collectively, our results provide a comprehensive molecular roadmap of Oct4 binding in concert with epigenetic rearrangements and rich resources for future reprogramming studies.

Academic research paper on topic "Hierarchical Oct4 Binding in Concert with Primed Epigenetic Rearrangements during Somatic Cell Reprogramming"

Cell Reports

Resource

Hierarchical Oct4 Binding in Concert with Primed Epigenetic Rearrangements during Somatic Cell Reprogramming

Graphical Abstract

Authors

Jun Chen, Xiaolong Chen, Min Li.....

Hong Wang, Cizhong Jiang, Shaorong Gao

Correspondence

czjiang@tongji.edu.cn (C.J.), gaoshaorong@tongji.edu.cn (S.G.)

In Brief

Transcriptional regulation by pluripotent factors and epigenetic rearrangements both play critical roles in reprogramming. Chen et al. perform a study of Oct4 occupancy and histone modifications during stages of somatic cell reprogramming. Depicting the molecular roadmap of Oct4 binding and its interplay with histone modifications provides a rich resource for exploring molecular underpinnings of reprogramming.

Accession Numbers GSE67462 GSE67520

• Primed epigenetic changes are set for Oct4 binding

• Oct4 binds to enhancers in more intricate modes than to promoters

• Oct4 regulates gene expression with distinct functional themes at different stages

Highlights

• Oct4 binds to target sites in a hierarchical fashion during reprogramming

Chen et al., 2016, Cell Reports 14, 1540-1554 ciossMark February 16, 2016 ©2016 The Authors

http://dx.d0i.0rg/l 0.1016/j.celrep.2016.01.013

CelPress

Cell Reports

Resource

Hierarchical Oct4 Binding in Concert with Primed Epigenetic Rearrangements during Somatic Cell Reprogramming

Jun Chen,12,34Xiaolong Chen,3 4 Min Li,34Xiaoyu Liu,2 Yawei Gao,3 Xiaochen Kou,3 Yanhong Zhao,3 Weisheng Zheng,3 Xiaobai Zhang,3 Yi Huo,2 Chuan Chen,3 You Wu,3 Hong Wang,3 Cizhong Jiang,3 * and Shaorong Gao3 *

1College of Life Science, Beijing Normal University, Beijing 100875, China 2National Institute of Biological Sciences (NIBS), Beijing 102206, China

3Clinical and Translational Research Center of Shanghai First Maternity & Infant Hospital, Shanghai Key Laboratory of Signaling and Disease Research, Collaborative Innovation Center for Brain Science, School of Life Sciences and Technology, Tongji University, Shanghai 200092, China 4Co-first author

'Correspondence: czjiang@tongji.edu.cn (C.J.), gaoshaorong@tongji.edu.cn (S.G.) http://dx.doi.org/10.10167j.celrep.2016.01.013

This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

SUMMARY

The core pluripotency factor Oct4 plays key roles in somatic cell reprogramming through transcriptional control. Here, we profile Oct4 occupancy, epigenetic changes, and gene expression in reprogramming. We find that Oct4 binds in a hierarchical manner to target sites with primed epigenetic modifications. Oct4 binding is temporally continuous and seldom switches between bound and unbound. Oct4 occupancy in most of promoters is maintained throughout the entire reprogramming process. In contrast, somatic cell-specific enhancers are silenced in the early and intermediate stages, whereas stem cell-specific enhancers are activated in the late stage in parallel with cell fate transition. Both epigenetic remodeling and Oct4 binding contribute to the hyper-dynamic enhancer signature transitions. The hierarchical Oct4 bindings are associated with distinct functional themes at different stages. Collectively, our results provide a comprehensive molecular roadmap of Oct4 binding in concert with epigenetic rearrangements and rich resources for future reprogramming studies.

INTRODUCTION

Somatic cells can be reverted to a pluripotent stem cell state by forced expression of Yamanaka factors Oct4, Sox2, Klf4, and c-Myc (OSKM) (Takahashi et al., 2007; Takahashi and Yamanaka, 2006). Understanding the underlying mechanisms driving this process is crucial, not only for potential clinical applications requiring high-quality induced pluripotent stem cells (iPSCs), but also for answering fundamental questions regarding cell plasticity, cell identity, and cell fate determination. Recently, advances have been made in defining the cascade of transcrip-

tional and epigenetic events that result in cell fate transitions (Buganim et al., 2013; Hussein et al., 2014). During reprogramming, cells go through three distinct phases termed initiation, maturation, and stabilization (Samavarchi-Tehrani et al., 2010). The initiation phase is marked by a mesenchymal-to-epithelial transition (MET) and upregulation of proliferation genes (Li et al., 2010; Samavarchi-Tehrani et al., 2010). In the subsequent maturation phase, the somatic cell program is further repressed and early pluripotency markers are activated. The hierarchical activation of the core pluripotency circuitry is initiated by the stochastic activation of key pluripotency markers (Buganim et al., 2012; Polo et al., 2012). However, cells at this phase are not fully reprogrammed until a transgene repression-dependent transition to the stabilization phase has occurred (Golipour et al., 2012). Chromatin remodeling also plays a very important role in reprogramming, as it provides the epigenetic basis for transcriptional regulation. Extensive studies that profiled DNA methylation and key histone modifications (HMs) during reprogramming have described the role of epigenetic rearrangements in resetting somatic cells to an embryonic stem cell (ESC)-like state (Hussein et al., 2014; Koche et al., 2011; Lee et al., 2014; Mikkelsen et al., 2007; Polo et al., 2012).

In ESCs, pluripotency is established and maintained through highly interconnected protein-DNA and protein-protein networks. Genome-wide occupancy studies have revealed that pluripotency transcription factors (TFs) often co-occupy regulatory elements and coordinate to control gene expression through various TF combinations (Boyer et al., 2005; Chen et al., 2008; Kim et al., 2008; Loh et al., 2006; Sridharan et al., 2009). These factors form a transcriptional regulatory hierarchy to control the ESC-specific expression program, both by activating genes critical for the maintenance of pluripotency and by repressing lineage-specific genes (Jaenisch and Young, 2008; Orkin and Hochedlinger, 2011; Young, 2011). Furthermore, a recent study found that a catalog of super-enhancers (SEs), which are densely co-bound by the core pluripotency factors and Mediator, define ESC identity (Whyte et al., 2013).

Previous studies revealed distinct binding patterns for Yama-naka factors OSKM during different stages of reprogramming (Soufi et al., 2012; Sridharan et al., 2009). Despite these advances, previous studies focused solely on the initial stage of reprogramming or pre-iPSCs/iPSCs. The occupancy dynamics of the core TFs in the genome during somatic cell reprogramming remain unexplored. In addition to TFs, the core pluripo-tency factors interacted with diverse chromatin-remodeling and -modifying complexes to regulate chromatin organization and gene expression during reprogramming (Orkin and Hochedlinger, 2011; Papp and Plath, 2013). However, the interplay between epigenetic rearrangements and the binding of the core TFs remains elusive.

Figure 1. Mouse Secondary Reprogramming System

(A) Outline of design for secondary reprogramming and cell sample collections. Dox is removed after day 15. Details are in the Supplemental Experimental Procedures.

(B) Representative microscopy images show cell morphology changes during 2° induction. Scale bar, 200 mm.

(C) Stacked bar chart summarizing FACS analysis by surface markers and Oct4-GFP expression during 2° induction. Data are presented as the mean ± SEM (n R 3).

(D) Principal-component analysis (PCA) of gene expression profiles of cell populations at indicated reprogramming time points from this study and two other published studies is shown.

(E) Unsupervised hierarchical clustering of gene expression profiles well defines known phases of reprogramming.

(F) The number of differentially expressed (DE) genes identified between successive time points during secondary reprogramming. The enriched GO terms for DE genes are listed in Figure S1C.

To address the aforementioned gaps in knowledge, we generated maps of genome-wide exogenous/total Oct4 occupancy and four core HMs (H3K4me1/ H3K4me3/H3K27ac/H3K27me3) at nine defined time points spanning the initiation, maturation, and stabilization phases of mouse somatic cell reprogramming using a doxycycline (Dox)-inducible secondary reprogramming system. We aimed to elucidate the dynamic binding patterns of the core pluripotency factor Oct4 and the selected key HMs, and their interplay, during reprogramming. In combination with gene expression profiles, our results demonstrate that Oct4 binds to target sites with primed epigenetic rearrangements in a temporally hierarchical manner during somatic cell reprogramming, and the hierarchical Oct4 associated with distinct functional themes at

bindings are different stages.

RESULTS

Establishment and Verification of a Rapid and Stable 3FLAG-OSKM 2° Reprogramming System

To explore the fundamental role of the core pluripotency factor Oct4 in somatic cell reprogramming, we set up a modified tetO-inducible secondary reprogramming system (Brambrink et al., 2008; Kang et al., 2009; Stadtfeld et al., 2008; Figure 1A). We mainly modified our reprogramming system in two perspectives as follows: (1) in addition to Dox, the induction medium was

supplemented with vitamin C (Vc) to improve reprogramming efficiency (Esteban et al., 2010); and (2) exogenous Oct4 was labeled with a 3x FLAG tag to distinguish it from endogenous Oct4.

We observed prominent morphological changes in the first 3 days of reprogramming, and cells began to aggregate on day 5. Oct4-GFP+ cell colonies emerged on day 7 and increased overtime (Figure 1B). Fluorescence-activated cell sorting (FACS) analysis of progressive reprogramming cells revealed that our reprogramming system is rapid and efficient. The reprogramming kinetics of our system showed a transition from a Thy1 + to Thy1 —, SSEA1+, and Oct4-GFP+ states in turn (Figures 1С and S1A), which is consistent with the previously reported Collal-tetO-OKSM transgenic reprogramming system (Polo etal., 2012).

Gene Expression Profiling during Reprogramming Progress

Principal-component analysis (PCA) of gene expression profiles of bulk reprogramming intermediate cells showed a transcrip-tomic progression trajectory similar to that of sorted cells (Polo et al., 2012). In contrast, the bulk of reprogramming intermediate cells in another study (Mikkelsen et al., 2008) were clustered with Thy1 + cells (Figure 1D). Unsupervised clustering of gene expression data defined three phases of reprogramming, as has been reported previously (Samavarchi-Tehrani et al., 2010): days 1-7 as the first-phase initiation; days 11-15 as the second-phase maturation; and day 18 and iPSC as the third-phase stabilization (Figure 1E).

Consistent with a previous study (Polo etal., 2012), our results showed two waves of gene expression changes, at the beginning and the intermediate stages of reprogramming (Figure 1F). The first wave took place between days 0 and 1, whereas the second wave emerged between days 7 and 11, toward the end of reprogramming. The first wave may reflect the immediate response to exogenous OSKM expression, which included upregulation of ribosomal/RNA-processing, cell-cycle, and DNA-repair-pathway genes and downregulation of cell adhesion-related genes. This prepares the cells for enhanced cell proliferation and MET. The second wave was characterized by early activation of pluripotency genes and suppression of the lineage-specific program (Figure S1C). Note that the augmentation of differentially expressed genes between days 15 and 18 was likely due to the withdrawal of Dox after day 15, and it likely represented transcriptomic modulation toward a well-defined pluripotent state at the stabilization phase (Golipour et al., 2012; Tonge et al., 2014). Furthermore, we examined some representative genes from various categories and confirmed that their expression patterns (Figure S1B) were typical, as in the previous study (Polo et al., 2012). Interestingly, Fbxo15 is a later marker in our system whereas it is an early marker in other systems. This difference is likely the effect of different stoichiom-etry of exogenous factors between reprogramming systems. Taken together, we established a rapid and efficient 2° reprogramming system suitable for dissecting progressive molecular changes during reprogramming.

Next, we collected bulk cell populations at different reprogramming time points and performed chromatin immunoprecipi-

tation sequencing (ChIP-seq) of Oct4and multiple HMs. Notably, although our data suggest that our reprogramming system is relatively efficient and behaves like enriched cell populations, the heterogeneity of reprogramming cells may limit the detection of minority binding events, which occur only in small sets of cells but are important for reprogramming.

Turnover between Exogenous and Endogenous Oct4 Binding during Reprogramming

A landscape of exogenous and total (endogenous and exogenous) Oct4 binding in the genome during reprogramming was generated by ChIP-seq using FLAG and Oct4 antibodies, respectively. Both total and exogenous Oct4-binding signals formed a unique peak around transcription start sites (TSSs) on day 1, whereas Oct4 signals around TSSs were close to background on day 0 (Figure S2A). De novo motif discovery identified Oct4-binding motif in both datasets that resembled the canonical Oct4 motif in the JASPAR database (Bryne et al., 2008; Figure S2B). In addition, clustering of total Oct4-binding profiles at specified time points gave rise to the same three phases of reprogramming as the clustering by gene expression profiles (Figures S2C and 1E). Taken together, these data suggest that our Oct4-binding profiles represent its bona fide target sites and reflect the kinetics of reprogramming.

The expression of endogenous Oct4 is gradually activated during reprogramming. However, when the endogenous Oct4 is expressed and its weight in the total Oct4 remain elusive. To address this, we compared exogenous and total Oct4-binding data. First, we found exogenous and total Oct4-binding peaks largely overlapped, indicating that they basically bound to the same genomic loci as expected (Figure S2D). Subsequently, we used a 1-kb window to scan the genome and calculate the total and exogenous Oct4-binding signals at indicated time points during reprogramming. Intriguingly, pairwise correlation analysis of Oct4 occupancy showed that exogenous and total Oct4-binding profiles were highly correlated from day 1 until day 7, whereas correlations decreased from day 11 until the end of reprogramming. (Figure 2A). The decline of correlation from day 11 coincided with the activation of endogenous Oct4, as demonstrated by gene expression analysis at both the mRNA and protein levels (Figures 2B and 2C). Notably, following the removal of Dox, the correlation on day 18 showed a pronounced drop. Together, these data suggest that our ChIP data present well the timing of endogenous Oct4 activation and its increasing weight in the total Oct4 during reprogramming.

Characteristics of Different Clusters of Oct4-Binding Sites

To investigate patterns of Oct4 binding in the reprogramming process, we clustered total Oct4 peaks by K-means and obtained seven Oct4-binding clusters (Figure 2D). Oct4 extensively bound to the genome at the beginning of reprogramming (clusters 1, 4, 5, and 6). Oct4 occupancy in cluster 1 target sites precipitously dropped to the basal level after day 7, whereas Oct4 occupancy in cluster 5 and 6 sites gradually decreased at late stage (around day 18). In contrast, cluster 4 sites were constantly occupied by Oct4 during the entire reprogramming process. However, only a small portion of sites (clusters 2 and 3) were

Figure 2. Oct4-Binding Patterns along Reprogramming

(A) Correlation between exogenous and total Oct4 occupancy across the genome during reprogramming. The genome was scanned with a 1-kb window.

(B)The qRT-PCR analysis showing transcription levels of exogenous, total, and endogenous Oct4 during 2° induction. Gene expression level is normalized to the housekeeping control gene Hprt and compared to day 0. Data are presented as the mean ± SEM (n = 3).

(C) Western blotting showing protein levels of exogenous and total Oct4 during 2° reprogramming. Exogenous and total Oct4 levels were detected with FLAG and Oct4 antibodies, respectively. Protein levels were normalized to Tubulin.

(D) K-means (k = 7) clustering of total Oct4-binding peak occupancy at indicated reprogramming time points is shown.

(E) Frequency of each cluster of total Oct4 peaks within specified ranges of distance to TSS is shown.

(F) Distribution of total Oct4-binding sites with respect to the TSS at different reprogramming time points is shown.

(G) Enrichment analysis of TF motifs in each cluster of total Oct4 peaks. Gene expression profiles for selected TFs are provided in Figure S2G.

occupied by Oct4 at the intermediate or late stage of reprogramming, and these sites remained occupied until the end of reprogramming. Additionally, a small group of sites (cluster 7) were transiently occupied by Oct4 from day 7 to day 15. As expected, when we examined the exogenous Oct4-binding signals on these seven clusters of total Oct4-binding sites, it turned out the same patterns (Figure S2E).

To better understand the biological function of Oct4-binding events, we examined whether there was bias in the genomic locations of each cluster of Oct4-binding sites. Intriguingly, cluster 4, 5, and 6 Oct4-binding sites were significantly enriched in promoter regions, whereas cluster 1 sites were

depleted in promoter regions. Other clusters of sites were relatively randomly located in the genome (Figure S2F). Approximately 50% of cluster 4 sites and ~25% of cluster 5 and 6 sites were located within 5 kb of the TSSs, while >80% of remaining clusters of sites were located more than 5 kb from the TSSs. Of note, >70% of cluster 1 sites were located beyond 50 kb of TSSs (Figure 2E). Moreover, there was a tendency for Oct4 to bind to upstream regions distal from the TSS in the early stage (day 1 to day 7), whereas its binding skewed toward regions more proximal to the TSS at the late stage (day 11 to iPSC) (Figure 2F), which was in agreement with the previous finding (Soufi et al., 2012).

Next, we searched for enriched TF-binding motifs within each cluster of Oct4-binding sites using the motif search tool Cistrome SeqPos (Liu et al., 2011). As expected, motifs of the TFs (Oct4, Sox2, Nanog, and Smad), which form the core pluripotency module, were enriched in all Oct4-binding sites (Figure 2G). Motif of another pluripotency TF Klf4 was depleted only in cluster 1 sites. In contrast, cluster 2 and 3 sites showed significant motif enrichment for ESC-specific TFs, including Esrrb, Tcfap2c, Zfp423, and Nr5a2. Many of these TFs were connected to the maintenance of pluripotency and self-renewal. Cluster 4 sites mainly included motifs of TFs (including members of the Elf family, Yy1, and Nfya) that have been implicated in transcription regulation through binding at proximal promoters. Notably, no c-Myc motif was enriched in any of Oct4-binding site clusters, and this is in line with c-Myc being outside of the core pluripotency network (Orkin and Hochedlinger, 2011). Additionally, we found that some of these TFs had an expression profile matching the patterns of Oct4 occupancy (Figures 2G and S2G).

Taken together, our results suggest a spatiotemporally regulated Oct4-binding pattern during reprogramming. Moreover, Oct4 binding may be orchestrated by different sets of TFs at different stages to execute distinct functions.

Epigenetic Landscape Presets a Context for Oct4 Binding

To address what resulted in the Oct4-binding patterns discussed in previous sections, we next explored the chromatin structure and DNA methylation level of Oct4-binding sites. FAIRE is a technique that identifies open genomic regions (Giresi et al., 2007). FAIRE sequencing (FAIRE-seq) reads of mouse embryonic fibroblast (MEF) (Wapinski et al., 2013) were enriched in the target sites that Oct4 bound to at the beginning of reprogramming (clusters 1, 4, 5, and 6), but they were depleted in clusters 2 and 3 to which Oct4 did not bind until the late stage (Figure 3A). Conversely, FAIRE-seq reads of mouse ESC line R1 (Buecker et al., 2014) were depleted in clusters 1, 5, and 6 but enriched in clusters 2, 3, and 4. This indicates that Oct4 prefers binding to locally open genomic regions. We further used MEF MNase-seq reads (Teif et al., 2012) to calculate nucleosome occupancy in Oct4-binding sites on day 0. All the sites were classified to ten groups by a 10% interval of ascendingly sorted nucleosome occupancy in MEFs. Intriguingly, Oct4 occupancy on day 1 was negatively correlated to nucleosome occupancy in MEFs (Figure 3B). This suggests that the openness status in genomic regions presets a context for and facilitates the initial Oct4 binding during reprogramming.

HM is an important factor that regulates chromatin structure. To explore whether there is any relationship between HMs and Oct4 binding, we calculated the signal of the key HMs on all Oct4-binding sites and conducted correlation analysis between each HM and Oct4 binding. We observed a positive correlation between Oct4 occupancy and the active markers H3K4me3, with a correlation coefficient median of 0.36, and H3K27ac, with a correlation coefficient median of 0.46, but no correlation between Oct4 binding and H3K27me3 (correlation coefficient median of —0.19) (Figures 3C and S3A).

We further obtained normalized DNA methylation data at MEFs and iPSCs (Lee et al., 2014) to examine DNA methylation

status on all Oct4-binding sites. We observed a significant change in average DNA methylation level in cluster 1, 2, 3, and 7 sites between MEFs and iPSCs (Figure 3D), which was opposite to Oct4 occupancy change in reprogramming. We further defined Oct4 sites with at least 20% change in DNA methylation level between MEFs and iPSCs. DNA methylation may be important for these sites. Among these sites, 83% of cluster 2 sites and 74% of cluster 3 sites exhibited decreases in DNA methylation level in iPSCs compared to MEFs. Also, 78% of cluster 1 sites and 71% of cluster 7 sites exhibited increases in DNA methylation level in iPSCs compared to MEFs (Figure S3B). We next examined the DNA methylation level dynamic changes at the selected Oct4-binding sites in clusters 2 and 3 at different time points of reprogramming and found that DNA methylation levels were negatively correlated with Oct4 occupancy (Figures 3E, S3C, and S3D). This suggests that DNA methylation levels are closely related to Oct4 binding in these target sites.

Dynamics of Oct4 Binding and Chromatin Remodeling in Promoter Regions

Dynamic chromatin remodeling through H3K4me3 and H3K27me3 changes during reprogramming has been investigated extensively (Hussein etal., 2014; Polo etal., 2012). Overall, our results of H3K4me3/H3K27me3 dynamics in promoter regions were consistent with previous findings that extensive chromatin remodeling occurred in promoter regions during reprogramming and that HM changes were closely coupled with gene expression changes (Figures S4A and S4B).

Next, we questioned how Oct4 binding in promoter regions regulates gene expression in coordination with chromatin remodeling. Auto-correlation analysis of Oct4 occupancy in promoter regions across time points divided the reprogramming process into three phases identical to those defined by expression profiles (Figures 1E and 4A). This suggests that Oct4-bind-ing dynamics in promoters reflects the molecular kinetics of reprogramming. We further focused on three groups of genes on the basis of Oct4-binding dynamics in their promoters. First, a majority of genes displayed relatively unchanged Oct4 binding during reprogramming, and the promoters of these genes maintained a stable and high H3K4me3 level and lacked the H3K27me3 mark (Figure 4B). Moreover, the expression levels of these genes were much higher than the average of all genes (Figure 4C). Gene Ontology (GO) analysis revealed that these genes were enriched for DNA repair, cell-cycle, and RNA-pro-cessing pathways, indicating that they maintained basic cellular functions as housekeeping genes (Figure 4B). Second, Oct4 occupancy decreased in the promoters of a group of genes during reprogramming in parallel with a decrease in H3K4me3 and a slight increase in H3K27me3. As a result, expression levels of these genes decreased (Figure 4D). GO analysis of this group of genes identified ion transport, cell adhesion, regulation of transcription, etc. Third, another group of genes gained Oct4 binding in their promoters during reprogramming. Accordingly, the H3K4me3 level increased and the H3K27me3 level decreased slightly. As a consequence, gene expressions increased (Figure 4E). GO analysis of this group of genes identified pluripotency, stem cell maintenance, etc. We further showed that activation of early, intermediate, and late

Figure 3. Relationship between Oct4 Binding and Chromatin Structure

(A) The accessibility of each cluster of Oct4-binding sites and their flanking regions before (day 0) and at the end stage of reprogramming as measured by the FAIRE-seq read count of MEF and ESC, respectively. The reads per kilobase of transcript per million mapped reads (RPKM) was scaled to 0-1 for comparison.

(B) The boxplot shows the Oct4-binding signal in Oct4-binding peaks grouped by nucleosome occupancy in MEFs.

(legend continued on next page)

pluripotency genes was consistent with sequential Oct4 occupancy in their promoter regions (Figures 4F, S4C, and S4D). Overall, Oct4 binding in promoter regions was positively correlated with H3K4me3 but negatively correlated with H3K27me3. Oct4 dynamic binding and chromatin remodeling in promoter regions are well coordinated to regulate gene expression.

Hyperdynamic Enhancers in Reprogramming

In addition to promoter regions, many Oct4 sites are located in regions distal to TSSs. To explore Oct4-binding dynamics in these regions, we identified 47,878-72,737 putative enhancers at each time point of reprogramming based on the histone mark H3K4me1. We further identified 5,710-20,968 active enhancers at each time point based on the H3K27ac mark as previously described (Creyghton etal., 2010; FigureS5A). This result was consistent with the estimate in the previous genome-wide studies that there may be ~1 million enhancers in all mammalian genome, whereas only tens of thousands were active in one given cell type (Bernstein et al., 2012; Heintzman et al., 2009; Thurman et al., 2012). To explore state changes in enhancer and the dynamics of Oct4 binding in enhancers in the reprogramming process, we next merged enhancers and defined 158,946 enhancers during reprogramming (see Supplemental Experimental Procedures). Thus, at any given reprogramming time point, an individual enhancer was defined as having one of the following three states: active (H3K4me1+ and H3K27ac+), poised (H3K4me1+ only), or off (H3K4me1-). Surprisingly, 99.2% of enhancers underwent state change during reprogramming; 24.1% of them went through all three states during reprogramming, while 63.9% changed states between poised and off (Figure 5A). Additional examination of enhancer state changes between successive time points revealed that direct changes between the active and off states were relatively rare. Instead, the transition to the poised state was often involved in a state change between the active and off states (Figures 5A and S5B).

Emerging evidence has shown that epigenomic enhancer signatures define cell identity and state, and that epigenetic patterning of enhancers occurs prior to cell fate decisions during development (Calo and Wysocka, 2013; Giresi et al., 2007; Liu et al., 2011; Zhang et al., 2008). Therefore, we explored transitions in active enhancers during reprogramming. Many MEF-active enhancers rapidly became non-active in the first 7 days, while only a small portion remained active (Figure 5B). Of the en-hancersthat were active in MEFs but non-active in iPSCs, 40.7% and 27.7% first changed to de novo non-active on days 1 and 3, respectively (Figure 5C). Moreover, the enhancers rarely changed back to the active state during reprogramming once they became de novo non-active (Figure 5D). In contrast, the majority of iPSC-active enhancers did not became active until day 15 (Figure 5B). Of the enhancers that were active in iPSCs but non-active in MEFs, 11.7% and 58.4% became de novo active on day 1 and in iPSCs, respectively (Figure 5E). Similarly,

most of the enhancers remained active once they became de novo active. Their states did not switch back and forth (Figure 5F). Our results demonstrate for the first time that the erasure of somatic cell memory occurs first and that the stem cell identity is established subsequently, at the enhancer level, during reprogramming.

Next, we used the Genomic Regions Enrichment of Annotations Tool (GREAT) to analyze the functional significance of enhancers that display distinct epigenomic dynamics during reprogramming. GO term analysis of the constantly poised enhancers identified enrichment for embryonic morphogenesis, pattern specification, and regionalization, suggesting that these enhancers are associated with early embryonic development and cell differentiation (Figure S5C). We propose that the primed yet inactivated state might be beneficial to permit a quick response to differentiation cues. GO analysis of the constantly active enhancers identified enrichment for fundamental biological processes, such as actin filament organization, RNA splicing, RNA processing, and cytoskeleton organization (Figure S5D). Surprisingly, a cluster of enhancers that were intermediately activated during reprogramming showed GO term enrichment for histone acetylation, peptidyl-lysine acetylation, and HM (Figure 5G). Consistent with the reported beneficial effect conferred by the inhibition of histone deacetylases (HDACi) on the transition from somatic cells or F-class cells to iPSCs during reprogramming (Plath and Lowry, 2011; Tonge et al., 2014), our results restated the importance of histone acet-ylation during reprogramming. As expected, enhancers that were activated late played roles mainly in the maintenance of pluripotency (Figure 5G).

Dynamics of Oct4 Binding in Enhancer Regions during Reprogramming

Oct4 bound to 38.3% (n = 60,864) of total enhancers during reprogramming. We grouped these enhancers into five clusters (A-E) based on the Oct4-binding signal using K-means algorithm (k = 5) (Figure 6A). The signals for key HMs (H3K27ac, H3K4me3, and H3K27me3) and RNA Pol II binding in each enhancer also were calculated. There was prominent positive correlation between Oct4 binding and H3K4me1, H3K27ac, H3K4me3, and RNA Pol II signals, except for H3K27me3 (Figure S6A). A careful examination of this correlation revealed that the presence of H3K4me1 in enhancers preceded Oct4 binding, implying that the H3K4me1 mark presets a context for Oct4 binding in enhancers. However, the acquisition of H3K27ac on these enhancers was often accompanied by or lagged behind Oct4 binding. Hence, our data suggest that enhancer activation occurs in a stepwise manner, first priming with the H3K4me1 mark, followed by TF binding, and subsequently the deposition of H3K27ac (Figure 6A). To help us understand the dynamics of Oct4 binding in enhancers, we tracked the source of Oct4 binding on these enhancers. Cluster A enhancers mainly contained cluster 4 Oct4 target sites, cluster C enhancers primarily included cluster 1

(C) Heatmaps of H3K4me3, H3K27ac, and H3K27me3 in each cluster of total Oct4-binding peaks. Peaks are shown In the same order.

(D) DNA methylation level in each cluster of Oct4-binding sites before (MEF) and at the end stage (iPSC) of reprogramming is shown (**p < 0.01).

(E) Oct4 ChIP-seq binding profiles at a chosen genomic locus (left), and DNA methylation levels at the Oct4 peak at the indicated time points (right). More loci examples are provided in Figure S3D.

Figure 4. Crosstalk between Oct4 Binding and HMs in Promoter Regions

(A) Pairwise correlations of Oct4 signal in promoter regions at the indicated reprogramming time points are shown.

(B) Promoters with constant Oct4-binding signal immediately following reprogramming exhibited constantly high H3K4me3 levels and lacked H3K27me3 (left). GO terms for the corresponding genes were enriched for housekeeping functions (right).

(C) Transcription levels of the genes in (B) were much higher than the average transcription level of all genes (red line) during reprogramming.

(D) Heatmap shows changes in the occupancy in "Oct4 loss'' promoters of Oct4/H3K4me3/H3K27me3 and in the corresponding gene expression levels.

(E) The gradual gain of Oct4 signal in promoters was positively correlated with an increase in H3K4me3 occupancy.

(F) Oct4 signal in Nodal locus, an early pluripotency gene. The corresponding expression change of Nodal is shown in Figure S1B. More loci examples are provided in Figures S4C and S4D.

Figure 5. Dynamics of Enhancer Status along Reprogramming

(A) Enhancers switch between three states (A, active; P, poised; and O, off) during reprogramming. The pie chart summarizes the proportion of each enhancer category.

(legend continued on next page)

Oct4 target sites, and Oct4 sites in cluster E enhancers largely consisted of cluster 2 and 3 Oct4 target sites (Figures S6B and 2D).

We further performed functional annotation of these Oct4-bound enhancers (Figure S6C). GO terms enriched in cluster A enhancers included macromolecule methylation, actin filament organization, histone methylation, and epigenetic regulation of gene expression. Cluster B and C enhancers were enriched for a variety of cell differentiation or organ development GO terms. Notably, cluster D enhancers were annotated to kinase activity regulation and multiple signaling pathways, including MAPK, TGFb, and EGF. Because these signaling pathways have been implicated in multiple mesodermal or endodermal lineage induction processes, we speculate that Oct4 binding on these enhancers is associated with its role as a lineage specifier (Thomson et al., 2011; Wang et al., 2012). In addition, cluster E enhancers that acquired Oct4 signal at the late stage of reprogramming were enriched for stem cell-related pathways (Figures 6A and S6C).

We further profiled the histone mark H3K27ac and Oct4 binding in a set of super-enhancers (SEs) that are specific to ESCs (Whyte et al., 2013). The results showed that both the H3K27ac mark and Oct4 binding were gradually established in a few SEs at the early stage (before day 7) of reprogramming, that they rapidly occupied many SEs on day 11, and that they occupied almost all SEs in iPSCs (Figures 6B and 6C). Together with late Oct4 binding at cluster E enhancers, these data highlight the function of Oct4 in the establishment and maintenance of pluripotency at the late stage of reprogramming.

Hierarchical Oct4 Binding Facilitates the Establishment of the Pluripotency Network during Reprogramming

Pluripotency was established through highly interconnected plu-ripotency regulatory networks consisting of core and other TFs (Orkin and Hochedlinger, 2011; Young, 2011). To address how Oct4 binding in enhancers regulates the activation of pluripo-tency networks, we collected member genes of the pluripotency networks from published works (Apostolou and Hochedlinger, 2013; Buganim et al., 2013; Hussein et al., 2014; Muller et al., 2008; Polo et al., 2012). We defined these genes as pluripotency network-related genes. Based on their expression profiles during reprogramming, they were clustered into four groups as follows: constantly active, early activated, intermediately activated, and late activated. Notably, the timing of Oct4 binding in their enhancers matched the timing of their transcriptional activation, indicating a possible causative link (Figure 6D). This suggests that Oct4 activates pluripotency networks in a hierarchical manner through sequential binding.

One gene can be regulated by multiple enhancers as well as the promoter. To study how Oct4 binding in these regulatory elements orchestrates the activation of downstream genes, we used Nanog and Oct4 (also known as Pou5f1) as examples to examine Oct4-binding status and key HM changes during reprogramming. Oct4 bound to these enhancers at different time points. Moreover, Oct4 binding to the enhancers often preceded its binding to the promoter. Oct4-binding status in the promoter was tightly coupled with transcriptional activation. (Figures 6E and S6E). Our data also imply that an additional layer of regulation might exist between Oct4 binding to distal elements and gene activation, such as enhancer-promoter looping, as has been reported (Apostolou et al., 2013). Consistent with our results mentioned above, epigenetic remodeling concorded with Oct4 binding. For example, H3K4me1 was deposited in the classical Nanog enhancer (E3, 4.5 kb upstream of TSS) in MEFs. Oct4 bound to E3 as early as on day 1, and the H3K27ac signal subsequently increased. In addition, the H3K27me3 signal decreased and the H3K4me3 signal increased as Oct4 bound to the Nanog promoter from day 7 (Figure S6D). Taken together, our data demonstrate that hierarchical Oct4 binding to the regulatory elements is in accord with epigenetic rearrangements to activate pluripotency gene expression during reprogramming.

DISCUSSION

Oct4 plays a central role in establishing and maintaining pluripo-tency in ESCs and during somatic cell reprogramming. Here we provide the first comprehensive roadmap of Oct4 binding throughout the reprogramming process. Our data reveal that Oct4 dynamically targets regulatory elements with distinct functional themes at different stages of reprogramming (Figure 7). Oct4 binds to a group of target sites from day 1 and maintains a constant and high level of Oct4 occupancy throughout reprogramming. These Oct4-binding sites reside predominantly within the promoters of housekeeping genes. There is another group of Oct4 target sites that are located mainly in regions distal to TSSs and that have functions in development and differentiation. Oct4 occupies these sites during the early and intermediate stages of reprogramming. Oct4 signals on these sites decrease dramatically in parallel with the repression of somatic or lineage-specific transcription programs. Oct4 begins binding to a set of target sites at the intermediate or late stage of reprogramming and maintains its stable occupancy. These sites are enriched in both promoters and enhancers of pluripotency genes. Oct4 occupancy in these regulatory regions plays an important role in the acquisition and maintenance of pluripotency during reprogramming (Figure 7). Moreover, our data indicate that

(B) Gradual loss of somatic cell-specific active enhancers and establishment of pluripotent stem cell-specific active enhancers during reprogramming. The number of active enhancers overlapping with MEF- and iPSC-specific active enhancers at the indicated reprogramming time points are shown in black and red, respectively.

(C) The percentage of enhancers that are deactivated de novo at each indicated time point during reprogramming is shown.

(D) Heatmap shows state switches between active and non-active of the de novo deactivated enhancers at each time point.

(E) The percentage of enhancers that are activated de novo at each indicated time point during reprogramming is shown.

(F) Heatmap shows state switches between active and non-active of the de novo activated enhancers at each time point.

(G) Top heatmaps display H3K27ac signal in two clusters of enhancers: intermediately and late activated. Bottom bar plots list biological process GO terms enriched in the corresponding enhancers.

See also Figure S5.

(legend on next page)

Figure 7. Model Depicting Dynamic Oct4 Binding in Concert with Epigenetic Changes during Reprogramming

(Left) Hierarchical patterns of Oct4 binding in enhancers and promoters with transcriptional control of distinct functional themes. (Right) Primed epigenetic rearrangements (e.g., nucleosome occupancy, HMs, and DNA methylation) give rise to a permissive state for Oct4 binding that in turn facilitates local chromatin remodeling and, consequently, activates the expression of target genes. If there is a looping between enhancers and promoters remains unknown.

sequential Oct4 binding plays a critical role in the hierarchical activation of pluripotency networks. The timing of initial Oct4 binding to promoters and enhancers is well matched to the sequential activation of different pluripotency genes (Figures 4F, 6D, S4C, and S4D).

We used integrative analysis of Oct4-binding data, in addition to epigenomic and transcriptomic data across the reprogramming process, to systemically evaluate the interplay between Oct4 binding and epigenetic remodeling and to explore how their combined action controls downstream gene transcription. We found that epigenetic rearrangements usually predispose a locus to a permissive state for Oct4 binding. For instance, Oct4 initially preferred to bind in open chromatin in MEFs, while condensed chromatin, DNA methylation, or the deposition of H3K27me3 on a promoter impeded Oct4 targeting. Prior to Oct4 binding, epigenetic rearrangements, including DNA demethylation and enhancer priming by H3K4me1 deposition, enabled a permissive environment. Subsequent Oct4 occupancy further activated the local chromatin, as indicated by an increase in the active histone marks H3K4me3 and H3K27ac. Overall, Oct4 bound to enhancers and promoters of genes and activated gene transcription in coordination with active epigenetic rearrangements (Figure 7).

Enhancers played important roles in the control of cell identity (Bulger and Groudine, 2011; Spitz and Furlong, 2012; Xie and Ren, 2013). However, the study of enhancers in a continuous and dynamic system is difficult because of the diversity and complexity of enhancer states that exist in different cell states. We have, for the first time, analyzed the hyperdynamic enhancer signature transition during reprogramming. Intrigu-ingly, somatic cell-specific enhancers were silenced in the early stage of reprogramming, followed by activation of stem cell-specific enhancers in the late stage, in parallel with the cell fate transition. Previous studies have developed an alter-

native lineage conversion strategy in which cell identity transition is first through a plastic intermediate state, induced by brief exposure to reprogramming factors, and followed by differentiation. Several different cell types, including murine cardiac and neuronal cells and human angioblast-like progenitor cells, have been derived successfully using this approach (Efe and Ding, 2011; Efe et al., 2011; Kim et al., 2011; Kurian et al., 2013). In relation to enhancer signatures, our observations support the existence of this plastic intermediate state in the intermediate reprogramming cells around day 7 and 11, which are the least like MEFs and iPSCs and have the smallest number of active enhancers (Figures 5B and S5A). Furthermore, our data confirm the importance of histone acet-ylation during reprogramming via a new insight that the activation of acetylation-related enhancers precedes that of the pluripotency-related counterparts (Figure 5G). Our results thus support the previously reported beneficial effect of HDACi on reprogramming and F-class-to-iPSC transition (Plath and Lowry, 2011; Tonge et al., 2014).

In summary, our study not only provides new insights into the roadmap of Oct4 binding and its instructive roles in reprogramming but also provides a rich data resource for further study of mechanisms involved in somatic cell reprogramming. Comprehensive profiling is still required for the other three factors: Sox2, Klf4, and c-Myc. Moreover, a careful examination of the precise functional consequences of different TF-binding dynamics is currently lacking due to technical limitations. It will be critical to study genomic architecture dynamics through circular chromatin conformation capture (4C)-based methods in order to fill the gap between TF binding in regulatory elements and transcription activation. Finally, the binding events detected in our study by ChIP-seq may reflect only a general trend in a bulk population, and we may have missed some key stochastic

Figure 6. Hierarchical Oct4 Binding in Enhancers during Reprogramming

(A) Heatmaps of H3K4me1, Oct4 occupancy, H3K27ac, RNA Pol II, H3K4me3, and H3K27me3 in Oct4-bound enhancers that were clustered using K-means clustering (k = 5) by Oct4 signal. Enhancers are displayed in the same order.

(B) H3K27ac signal in the ESC-specific super-enhancers (SEs) at indicated reprogramming time points is shown.

(C) Oct4 signal in the SEs at indicated reprogramming time points is shown.

(D) Heatmaps showing transcription levels of four groups of pluripotency genes during reprogramming (left). Heatmaps display Oct4 signal in the corresponding enhancers (right). The genes and the enhancers are in the same order.

(E) Schematic diagrams showing the transcriptional control of Oct4 binding in enhancers and promoters at Nanog and Oct4 (also known as Pou5f1) loci. The detailed ChIP-seq profiles are shown in Figures S6D and S6E.

binding events that occur in the early stage. New technologies and approaches will be required to study these processes at the single-cell level in the future.

EXPERIMENTAL PROCEDURES Animals

Mouse procedures were carried out according to the guidelines for the use of laboratory animals of the National Institute of Biological Sciences (NIBS).

TetO-FUW-3FLAG-Oct4 Lentivirus Vector Construction and 1° Reprogramming

The Oct4 CDS was cloned from the TetO-FUW-Oct4 plasmid (Addgene plasmid 20323) and inserted into PBS-N-strep-4FLAG backbone (a gift from Yuelin Zhang, NIBS); the 3FLAG-Oct4 fragment was amplified by PCR and cloned back into the TetO-FUW backbone. The 1° reprogramming was performed as previously described, using inducible lentivirus vectors (Brambrink et al., 2008; Kang et al., 2009; Stadtfeld et al., 2008). When GFP-positive colonies appeared ~2 weeks after Dox and Vc induction, ten iPS colonies were picked and propagated after Dox withdrawal.

Tetraploid Complementation Assay

The tetraploid complementation assay was performed as previously described (Kang et al., 2009). Two-cell stage B6D2F1 embryos were electrofused and cultured to blastocyst stage in vitro. Approximately 10-15 iPS cells were injected into the cavities of tetraploid blastocysts. The complemented blastocysts were then transplanted into pseudo-pregnant ICR recipients. MEFs (20) were derived from embryonic day (E)13.5 embryos and genotyped for cell origin.

Flow Cytometry

Reprogramming cells were harvested and incubated with antibodies against Thy1.2 (PE/Cy7, BioLegend) and SSEA1 (AlexaFluor647, BioLegend). Flow cytometry analyses for GFP, Thy1, and SSEA1 were performed as previously described (Gao et al., 2013).

Reprogramming cells were collected at different time points, crosslinked, and sonicated. Sonicated chromatin was incubated with pretreated antibody-coupled Protein G Dynabeads (Invitrogen) at 4°C overnight. Antibodies used in ChIP assays and other experiments are listed in Table S1. ChIP DNA was reverse-crosslinked, eluted, and purified by phenol-chloroform-isoamyl alcohol extraction, followed by ethanol precipitation. ChIP DNA was then quantified using a Qubit dsDNA HS assay kit (Life Technologies). We performed two or three independent IP reactions for each sample, and ChIP DNA was pooled for library preparation. For sequencing, 10-100 ng DNA was used for library preparation following the New England Biolabs library preparation protocol.

ChIP-Seq

ChIP products were subjected to high-throughput sequencing with HiSeq 2000/HiSeq 2500 sequencer using a 50-bp SE protocol.

DNA Methylation Analysis by Bisulfite Sequencing

Genomic DNA was extracted using a TIANamp Genomic DNA Kit (TIANGEN) and bisulfite converted with an EpiTect Bisulfite Kit (QIAGEN). Methylation PCR primers were designed by MethPrimer and data analysis was performed using QUMA (Kumaki et al., 2008; Li and Dahiya, 2002).

RNA Profiling

Gene expression profiling was performed using Affymetrix Mouse Gene 1.0 ST arrays. Microarray datasets were preprocessed and quantile normalized with the robust multi-array average (RMA) method.

Analysis of Global Oct4 Occupancy

After Oct4 ChIP-seq reads were mapped to mouse genome, we calculated FLAG and total Oct4 occupancy by scanning the genome with a 1-kb sliding

windowfor correlation analysis at distinct time points. We called Oct4-binding peaks at each time point that were further merged as the final set of Oct4-bind-ing sites in this study. FAIRE-seq and MNase-seq reads mapped in each site were counted to measure the accessibility of Oct4 sites. The signals of HMs in Oct4 sites were calculated in a similar manner. We further correlated the HM signals with Oct4 occupancy in the sites.

Oct4 Binding and HMs in Promoters and Enhancers

Gene promoters were defined as —1.5 kb to +0.5 kb of gene TSSs. Read counts of HMs and Oct4 in promoters defined their signal intensity that was used to analyze their dynamic changes in reprogramming. H3K4me1 peaks were called in a way similar to Oct4 peaks and defined enhancers excluding those overlapped with promoters. All enhancers of distinct time points were merged as the final set of enhancers in this study. Signals of Oct4, HMs, and RNA Pol II in an enhancer were calculated to define its activation state, analyze correlation and dynamic changes, and so forth.

Hierarchical Activation Analysis of Pluripotency Genes through Oct4 Binding in Enhancers

Pluripotency genes were downloaded through literature mining. We clustered the genes based on their expression profiles, and we further correlated Oct4 occupancy in their enhancers and their expression levels in reprogramming. See the Supplemental Experimental Procedures for more detailed methods.

ACCESSION NUMBERS

The accession numbers for the microarray transcriptomic data and the ChIP-seq reported in this paper are GEO: GSE67462 and GSE67520, respectively.

SUPPLEMENTAL INFORMATION

Supplemental Information includes Supplemental Experimental Procedures, six figures, and three tables and can be found with this article online at http://dx.doi.org/10.1016/j.celrep.2016.01.013.

AUTHOR CONTRIBUTIONS

C.J. and S.G. conceived and designed the study. J.C. performed most of the experiments. X.L., Y.G., and Y.H. participated in mice-related experiments. X.K. and Y.Z. performed the tetraploid complementation assay. C.C., Y.W., and H.W. participated in bisulfite sequencing assay. X.C. and M.L. did most of the bioinformatic analysis. W.Z. and X.Z. assisted in data analysis. J.C., C.J., and S.G. wrote the paper.

ACKNOWLEDGMENTS

We thank Professor Jeong-Sun Seo (Seoul National University) for kindly providing DNA methylation data of MEFs and iPSCs and Professor B. Franklin Pugh for his comments (Pennsylvania State University). This work was supported by the National Natural Science Foundation of China (91319306, 31325019, 31430056, 91519309, 91019017, and 31271373), the Ministry of Science and Technology (grants 2012CBA01300, 2011CB965104, and 2012AA020405), the Aurora Talent Project of Shanghai (10SG24), and the program for Eastern Scholar of Shanghai.

Received: August 18, 2015 Revised: November 17, 2015 Accepted: December 30, 2015 Published: January 28, 2016

REFERENCES

Apostolou, E., and Hochedlinger, K. (2013). Chromatin dynamics during cellular reprogramming. Nature 502, 462-471.

Apostolou, E., Ferrari, F., Walsh, R.M., Bar-Nur, O., Stadtfeld, M., Cheloufi, S., Stuart, H.T., Polo, J.M., Ohsumi, T.K., Borowsky, M.L., et al. (2013).

Genome-wide chromatin interactions of the Nanog locus in pluripotency, differentiation, and reprogramming. Cell Stem Cell 12, 699-712. Bernstein, B.E., Birney, E., Dunham, I., Green, E.D., Gunter, C., Snyder, M., Epstein, C.B., Frietze, S., Harrow, J., and Kaul, R.; ENCODE Project Consortium (2012). An integrated encyclopedia of DNA elements in the human genome. Nature 489, 57-74.

Boyer, L.A., Lee, T.I., Cole, M.F., Johnstone, S.E., Levine, S.S., Zucker, J.P., Guenther, M.G., Kumar, R.M., Murray, H.L., Jenner, R.G., et al. (2005). Core transcriptional regulatory circuitry in human embryonic stem cells. Cell 122, 947-956.

Brambrink, T., Foreman, R., Welstead, G.G., Lengner, C.J., Wernig, M., Suh, H., and Jaenisch, R. (2008). Sequential expression of pluripotency markers during direct reprogramming of mouse somatic cells. Cell Stem Cell 2, 151-159.

Bryne, J.C., Valen, E., Tang, M.H., Marstrand, T., Winther, O., da Piedade, I., Krogh, A., Lenhard, B., and Sandelin, A. (2008). JASPAR, the open access database of transcription factor-binding profiles: new content and tools in the 2008 update. Nucleic Acids Res. 36, D102-D106.

Buecker, C., Srinivasan, R., Wu, Z., Calo, E., Acampora, D., Faial, T., Simeone, A., Tan, M., Swigut, T., and Wysocka, J. (2014). Reorganization of enhancer patterns in transition from naive to primed pluripotency. Cell Stem Cell 14, 838-853.

Buganim, Y., Faddah, D.A., Cheng, A.W., Itskovich, E., Markoulaki, S., Ganz, K., Klemm, S.L., van Oudenaarden, A., and Jaenisch, R. (2012). Single-cell expression analyses during cellular reprogramming reveal an early stochastic and a late hierarchic phase. Cell 150, 1209-1222.

Buganim, Y., Faddah, D.A., and Jaenisch, R. (2013). Mechanisms and models of somatic cell reprogramming. Nat. Rev. Genet. 14, 427-439. Bulger, M., and Groudine, M. (2011). Functional and mechanistic diversity of distal transcription enhancers. Cell 144, 327-339.

Calo, E., and Wysocka, J. (2013). Modification of enhancer chromatin: what, how, and why? Mol. Cell 49, 825-837.

Chen, X.,Xu, H., Yuan, P., Fang, F., Huss, M., Vega, V.B., Wong, E., Orlov, Y.L., Zhang, W., Jiang, J., et al. (2008). Integration of external signaling pathways with the core transcriptional network in embryonic stem cells. Cell 133, 1106-1117.

Creyghton, M.P., Cheng, A.W., Welstead, G.G., Kooistra, T., Carey, B.W., Steine, E.J., Hanna, J., Lodato, M.A., Frampton, G.M., Sharp, P.A., et al. (2010). Histone H3K27ac separates active from poised enhancers and predicts developmental state. Proc. Natl. Acad. Sci. USA 107, 21931-21936. Efe, J.A., and Ding, S. (2011). Reprogramming, transdifferentiation and the shifting landscape of cellular identity. Cell Cycle 10, 1886-1887. Efe, J.A., Hilcove, S., Kim, J., Zhou, H., Ouyang, K., Wang, G., Chen, J., and Ding, S. (2011). Conversion of mouse fibroblasts into cardiomyocytes using a direct reprogramming strategy. Nat. Cell Biol. 13, 215-222. Esteban, M.A., Wang, T., Qin, B., Yang, J., Qin, D., Cai, J., Li, W., Weng, Z., Chen, J., Ni, S., et al. (2010). Vitamin C enhances the generation of mouse and human induced pluripotent stem cells. Cell Stem Cell 6, 71-79. Gao, Y., Chen, J., Li, K., Wu, T., Huang, B., Liu, W., Kou, X., Zhang, Y., Huang, H., Jiang, Y., et al. (2013). Replacement of Oct4 byTet1 during iPSC induction reveals an important role of DNA methylation and hydroxymethylation in reprogramming. Cell Stem Cell 12, 453-469.

Giresi, P.G., Kim, J., McDaniell, R.M., Iyer, V.R., and Lieb, J.D. (2007). FAIRE (Formaldehyde-Assisted Isolation of Regulatory Elements) isolates active regulatory elements from human chromatin. Genome Res. 17, 877-885. Golipour, A., David, L., Liu, Y., Jayakumaran, G., Hirsch, C.L., Trcka, D., and Wrana, J.L. (2012). A late transition in somatic cell reprogramming requires regulators distinct from the pluripotency network. Cell Stem Cell 11, 769-782. Heintzman, N.D., Hon, G.C., Hawkins, R.D., Kheradpour, P., Stark, A., Harp, L.F., Ye, Z., Lee, L.K., Stuart, R.K., Ching, C.W., et al. (2009). Histone modifications at human enhancers reflect global cell-type-specific gene expression. Nature 459, 108-112.

Hussein, S.M., Puri, M.C., Tonge, P.D., Benevento, M., Corso, A.J., Clancy, J.L., Mosbergen, R., Li, M., Lee, D.S., Cloonan, N., et al. (2014). Genome-wide characterization of the routes to pluripotency. Nature 516, 198-206. Jaenisch, R., and Young, R. (2008). Stem cells, the molecular circuitry of pluripotency and nuclear reprogramming. Cell 132, 567-582. Kang, L., Wang, J., Zhang, Y., Kou, Z., and Gao, S. (2009). iPS cells can support full-term development of tetraploid blastocyst-complemented embryos. Cell Stem Cell 5, 135-138.

Kim, J., Chu, J., Shen, X., Wang, J., and Orkin, S.H. (2008). An extended transcriptional network for pluripotency of embryonic stem cells. Cell 132, 10491061.

Kim, J., Efe, J.A., Zhu, S., Talantova, M., Yuan, X., Wang, S., Lipton, S.A., Zhang, K., and Ding, S. (2011). Direct reprogramming of mouse fibroblasts to neural progenitors. Proc. Natl. Acad. Sci. USA 108, 7838-7843. Koche, R.P., Smith, Z.D., Adli, M., Gu, H., Ku, M., Gnirke, A., Bernstein, B.E., and Meissner, A. (2011). Reprogramming factor expression initiates widespread targeted chromatin remodeling. Cell Stem Cell 8, 96-105. Kumaki, Y., Oda, M., and Okano, M. (2008). QUMA: quantification tool for methylation analysis. Nucleic Acids Res. 36, W170-W175. Kurian, L., Sancho-Martinez, I., Nivet, E., Aguirre, A., Moon, K., Pendaries, C., Volle-Challier, C., Bono, F., Herbert, J.M., Pulecio, J., et al. (2013). Conversion of human fibroblasts to angioblast-like progenitor cells. Nat. Methods 10, 77-83.

Lee, D.S., Shin, J.Y., Tonge, P.D., Puri, M.C., Lee, S., Park, H., Lee, W.C., Hussein, S.M., Bleazard, T., Yun, J.Y., et al. (2014). An epigenomic roadmap to induced pluripotency reveals DNA methylation as a reprogramming modulator. Nat. Commun. 5, 5619.

Li, L.C., and Dahiya, R. (2002). MethPrimer: designing primers for methylation PCRs. Bioinformatics 18, 1427-1431.

Li, R., Liang, J., Ni, S., Zhou, T., Qing, X., Li, H., He, W., Chen, J., Li, F., Zhuang, Q., et al. (2010). Amesenchymal-to-epithelial transition initiates and is required for the nuclear reprogramming of mouse fibroblasts. Cell Stem Cell 7, 51-63. Liu, T., Ortiz, J.A., Taing, L., Meyer, C.A., Lee, B., Zhang, Y., Shin, H., Wong, S.S., Ma, J., Lei, Y., et al. (2011). Cistrome: an integrative platform for transcriptional regulation studies. Genome Biol. 12, R83. Loh, Y.H., Wu, Q., Chew, J.L., Vega, V.B., Zhang, W., Chen, X., Bourque, G., George, J., Leong, B., Liu, J., et al. (2006). The Oct4 and Nanog transcription network regulates pluripotency in mouse embryonic stem cells. Nat. Genet. 38, 431-440.

Mikkelsen, T.S., Ku, M., Jaffe, D.B., Issac, B., Lieberman, E., Giannoukos, G., Alvarez, P., Brockman, W., Kim, T.-K., Koche, R.P., et al. (2007). Genome-wide maps of chromatin state in pluripotent and lineage-committed cells. Nature 448, 553-560.

Mikkelsen, T.S., Hanna, J., Zhang, X., Ku, M., Wernig, M., Schorderet, P., Bernstein, B.E., Jaenisch, R., Lander, E.S., and Meissner, A. (2008). Dissecting direct reprogramming through integrative genomic analysis. Nature 454, 49-55.

Müller, F.J., Laurent, L.C., Kostka, D., Ulitsky, I., Williams, R., Lu, C., Park, I.H., Rao, M.S., Shamir, R., Schwartz, P.H., et al. (2008). Regulatory networks define phenotypic classes of human stem cell lines. Nature 455, 401-405. Orkin, S.H., and Hochedlinger, K. (2011). Chromatin connections to pluripotency and cellular reprogramming. Cell 145, 835-850. Papp, B., and Plath, K. (2013). Epigenetics of reprogramming to induced pluripotency. Cell 152, 1324-1343.

Plath, K., and Lowry, W.E. (2011). Progress in understanding reprogramming to the induced pluripotent state. Nat. Rev. Genet. 12, 253-265. Polo, J.M., Anderssen, E., Walsh, R.M., Schwarz, B.A., Nefzger, C.M., Lim, S.M., Borkent, M., Apostolou, E., Alaei, S., Cloutier, J., et al. (2012). A molecular roadmap of reprogramming somatic cells into iPS cells. Cell 151, 16171632.

Samavarchi-Tehrani, P., Golipour, A., David, L., Sung, H.-K., Beyer, T.A., Datti, A., Woltjen, K., Nagy, A., and Wrana, J.L. (2010). Functional genomics reveals

a BMP-driven mesenchymal-to-epithelial transition in the initiation of somatic cell reprogramming. Cell Stem Cell 7, 64-77.

Soufi, A., Donahue, G., and Zaret, K.S. (2012). Facilitators and impediments of the pluripotency reprogramming factors' initial engagement with the genome. Cell 151, 994-1004.

Spitz, F., and Furlong, E.E. (2012). Transcription factors: from enhancer binding to developmental control. Nat. Rev. Genet. 13, 613-626. Sridharan, R., Tchieu, J., Mason, M.J., Yachechko, R., Kuoy, E., Horvath, S., Zhou, Q., and Plath, K. (2009). Role of the murine reprogramming factors in the induction of pluripotency. Cell 136, 364-377.

Stadtfeld, M., Maherali, N., Breault, D.T., and Hochedlinger, K. (2008). Defining molecular cornerstones during fibroblast to iPS cell reprogramming in mouse. Cell Stem Cell 2, 230-240.

Takahashi, K., and Yamanaka, S. (2006). Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell 126, 663-676.

Takahashi, K., Tanabe, K., Ohnuki, M., Narita, M., Ichisaka, T., Tomoda, K., and Yamanaka, S. (2007). Induction of pluripotent stem cells from adult human fibroblasts by defined factors. Cell 131, 861-872.

Teif, V.B., Vainshtein, Y., Caudron-Herger, M., Mallm, J.P., Marth, C., Höfer,T., and Rippe, K. (2012). Genome-wide nucleosome positioning during embryonic stem cell development. Nat. Struct. Mol. Biol. 19, 1185-1192. Thomson, M., Liu, S.J., Zou, L.N., Smith, Z., Meissner, A., and Ramanathan, S. (2011). Pluripotency factors in embryonic stem cells regulate differentiation into germ layers. Cell 145, 875-889.

Thurman, R.E., Rynes, E., Humbert, R., Vlerstra, J., Maurano, M.T., Haugen, E., Sheffield, N.C., Stergachls, A.B., Wang, H., Vernot, B., et al. (2012). The accessible chromatin landscape of the human genome. Nature 489, 75-82.

Tonge, P.D., Corso, A.J., Monettl, C., Hussein, S.M., Purl, M.C., Michael, I.P., Li, M., Lee, D.S., Mar, J.C., Cloonan, N., et al. (2014). Divergent reprogramming routes lead to alternative stem-cell states. Nature 516, 192-197.

Wang, Z., Oron, E., Nelson, B., Razis, S., and Ivanova, N. (2012). Distinct lineage specification roles for NANOG, OCT4, and SOX2 in human embryonic stem cells. Cell Stem Cell 10, 440-454.

Wapinski, O.L., Vierbuchen, T., Qu, K., Lee, Q.Y., Chanda, S., Fuentes, D.R., Giresi, P.G., Ng, Y.H., Marro, S., Neff, N.F., et al. (2013). Hierarchical mechanisms for direct reprogramming of fibroblasts to neurons. Cell 155, 621-635.

Whyte, W.A., Orlando, D.A., Hnisz, D., Abraham, B.J., Lin, C.Y., Kagey, M.H., Rahl, P.B., Lee, T.I., and Young, R.A. (2013). Master transcription factors and mediator establish super-enhancers at key cell identity genes. Cell 153, 307-319.

Xie, W., and Ren, B. (2013). Developmental biology. Enhancing pluripotency and lineage specification. Science 341, 245-247.

Young, R.A. (2011). Control of the embryonic stem cell state. Cell 144, 940-954.

Zhang, Y., Liu, T., Meyer, C.A., Eeckhoute, J., Johnson, D.S., Bernstein, B.E., Nusbaum, C., Myers, R.M., Brown, M., Li, W., and Liu, X.S. (2008). Modelbased analysis of ChIP-Seq (MACS). Genome Biol. 9, R137.