H3K36 Histone Methyltransferase Setd2 Is Required ...

Article

H3K36 Histone Methyltransferase Setd2 Is Required for Murine Embryonic Stem Cell Differentiation toward Endoderm Graphical Abstract

Authors Yuanliang Zhang, Shugao Xie, ..., Qiuhua Huang, Saijuan Chen

Correspondence [email protected] (Q.H.), [email protected] (S.C.)

In Brief Chromatin regulation plays an important role in embryonic development. Zhang et al. now use Setd2 / murine embryonic stem cells to show that the Setd2 H3K36 histone methyltransferase regulates differentiation toward primitive endoderm. The authors connect Setd2 to Fgf signaling and find that Setd2 regulates endoderm differentiation by directly affecting transcriptional initiation of Fgfr3 through histone H3K36me3 modification in its distal promoter region.

Highlights Setd2 is critical for endodermal differentiation in mESCs The Erk pathway plays a role in endoderm defects caused by Setd2 deficiency Setd2 regulates Fgfr3 transcription initiation via H3K36me3 modification

Zhang et al., 2014, Cell Reports 8, 1989–2002 September 25, 2014 ª2014 The Authors http://dx.doi.org/10.1016/j.celrep.2014.08.031

Cell Reports

Article H3K36 Histone Methyltransferase Setd2 Is Required for Murine Embryonic Stem Cell Differentiation toward Endoderm Yuanliang Zhang,1,4 Shugao Xie,1,4 Yan Zhou,1,4,5 Yinyin Xie,1 Ping Liu,1 Mingming Sun,2 Huasheng Xiao,2 Ying Jin,3 Xiaojian Sun,1 Zhu Chen,1 Qiuhua Huang,1,* and Saijuan Chen1,* 1State Key Laboratory of Medical Genomics, Shanghai Institute of Hematology, Rui Jin Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai 200025, China 2National Engineering Center for Biochip at Shanghai, Shanghai 201203, China 3Shanghai Stem Cell Institute, Shanghai Jiao Tong University School of Medicine, Shanghai 200025, China 4Co-first author 5Present address: Medical Sciences Research Center, Renji Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai 200127, China *Correspondence: [email protected] (Q.H.), [email protected] (S.C.) http://dx.doi.org/10.1016/j.celrep.2014.08.031 This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).

SUMMARY

Setd2 is known as a histone-H3K36-specific methyltransferase. However, its role in physiological function remains unclear. In this study, we show that Setd2 mainly regulates differentiation of murine embryonic stem cells (mESCs) toward primitive endoderm. Furthermore, we show that downregulated endoderm-related genes in Setd2/ mESCs are associated with an aberrantly low level of Erk activity and that enforced expression of Fgfr3 can rescue the defective Erk pathway in Setd2/ mESCs. Interestingly, the transcriptional initiation of Fgfr3 is directly regulated through histone H3K36me3 modification in its distal promoter region by Setd2. These results indicate that Setd2 controls the primitive endoderm differentiation of mESCs by regulating the Fgfr3-Erk signaling. INTRODUCTION Embryonic stem cells (ESCs) are pluripotent cells with two key properties: self-renewal and differentiation into all cell types. Gene expression patterns can effectively distinguish the differences between lineage-committed cells and the pluripotent cells. It therefore stands to reason that proteins involved in the control of transcription in ESCs might play important roles in lineage commitment and differentiation of pluripotent cells (Reynolds et al., 2012). A complex regulatory network of transcription factors has been shown to play critical roles for the maintenance of ESCs lineage commitment or pluripotency (Chen et al., 2008). In addition, several studies have shown that epigenetic events can also regulate ESCs pluripotency. Knockdown or knockout of chromatin regulators such as Jarid2 and Wdr5 not only abolish self-renewal but also trigger multiple lineage differentiation of

ESCs (Ang et al., 2011; Bilodeau et al., 2009; Fazzio et al., 2008; Gaspar-Maia et al., 2009; Li et al., 2012; Loh et al., 2007; Peng et al., 2009; Shen et al., 2009; Xie et al., 2011; Yuan et al., 2009). Histone methylation is now recognized as one of the most important and complex epigenetic modification mechanisms, occurring mainly at gene promoter regions linked to both activated and repressed gene expression (Barski et al., 2007; Martin and Zhang, 2005). In murine embryonic stem cells (mESCs), nearly 80% of nucleosomes are trimethylated at K4 of histone H3 (H3K4me3) in promoter regions. These permissive chromatin regions endow ESCs with an open transcriptional state and provide them with more developmental plasticity (Efroni et al., 2008). Furthermore, ‘‘bivalent chromatin’’ containing both H3K4me3 and H3K27me3 modifications has also been found in ESCs and most of the genes with bivalent chromatin structure seem to be involved in cell fate determination (Bernstein et al., 2006). Disruption of H3K4me3 or H3K27me3 leads ESCs to inefficiently differentiate into a neural lineage (Jiang et al., 2011; Pasini et al., 2007). Although the enrichment pattern of H3K9me3 was observed at telemere, satellite, and long terminal repeats (Mikkelsen et al., 2007), two groups reported that more euchromatic genes contain H3K9me3-modified nucleosomes in mESCs (Bilodeau et al., 2009; Karimi et al., 2011) and disruption of H3K9me3 leads to reduction of stem cell capacity and induction of differentiation (Bilodeau et al., 2009). Enrichment pattern of H3K36me3 seems to be quite different from that of other histone methylations, because it is usually observed at body regions of active genes, especially at their 30 terminal regions (Mikkelsen et al., 2007). However, the function of H3K36me3 methylation in mESCs self-renewal and lineage-specific differentiation remains obscure. Set2 is the sole H3K36 histone methyltransferase in yeast (Strahl et al., 2002), and it can interact with elongating RNA polymerase II (RNA pol II). Mutation of Set2 results in an increased sensitivity to 6-azauracil, a useful tool for identifying proteins important for elongation. This suggests that Set2 is required

Cell Reports 8, 1989–2002, September 25, 2014 ª2014 The Authors 1989

Figure 1. Setd2 Is Required for mESCs Endoderm Differentiation (A) Morphology of WT and Setd2/ embryonic bodies (EBs) after being suspension cultured without LIF for 3 (left) and 6 days (right). The scale bar represents 10 mm. (B) Left: hematoxylin and eosin staining of d3 EBs after gelatin-attached culture for 3 days. Black dotted lines indicate EBs, and white dotted lines indicate the boundary of cells growing around EBs. Right: quantification of the areas of the cells growing around EBs. Data are presented as means ± SEM (n = 24–27); ***p < 0.001. The scale bar represents 5 mm. (C) Real-time PCR analysis of endoderm lineage markers expressed in d3, d6, and d9 WT and Setd2/ EBs. Relative expression levels were presented by 2^DDCt (knockout-WT), similarly hereinafter.

(legend continued on next page)

1990 Cell Reports 8, 1989–2002, September 25, 2014 ª2014 The Authors

for transcription elongation (Li et al., 2002). During RNA-pol-IImediated transcription elongation, H3K36me3 can be recognized by Rpd3s complex, causing suppression of intragenic transcription initiation through restraining histone exchange (Carrozza et al., 2005; Keogh et al., 2005; Venkatesh et al., 2012). This pathway can also regulate infrequently transcribed long genes for accurate transcription in yeast (Li et al., 2007). The human ortholog of Set2 is SETD2/HYPB, which was originally isolated from CD34+ hematopoietic stem/progenitor cells by our group (Mao et al., 1998). SETD2/HYPB also acts as an H3K36 methyltransferase in vitro and interacts with RNA pol II in human cells (Sun et al., 2005). Disruption of Setd2 in mice results in vascular remodeling defects and H3K36me3 downregulation (Hu et al., 2010). H3K36me3 tends to mark exons, especially alternative exons, rather than introns in mammalian genomes (de Almeida et al., 2011; Kolasinska-Zwierz et al., 2009). It has been suggested that Setd2 plays a role in alternative splicing (Luco et al., 2010). Interestingly, recent work revealed that human SETD2 is often mutated in clear cell renal cell carcinoma (cRCC) (Dalgliesh et al., 2010; Duns et al., 2010) and acute leukemia (Mar et al., 2014; Zhu et al., 2014), whereas some cancer cells lacking H3K36me3 display microsatellite instability and an elevated spontaneous mutation frequency (Li et al., 2013). Although some biochemical mechanisms of H3K36me3 and Setd2 have been revealed, transcriptional regulation through Setd2-catalyzed H3K36me3 modification in development and normal cellular function, such as ESCs self-renewal and lineage differentiation, is yet to be investigated. RESULTS Setd2 Deficiency Is Dispensable for Self-Renewal of Mouse ESCs We first set out to dissect whether Setd2 could play a role in maintaining the self-renewal and differentiation properties of ESCs. Several observations suggest that depletion of Setd2 does not affect mESCs self-renewal. First, Setd2/ mESCs kept aggregated morphology typical of undifferentiated mESCs with a microstructure similar to wild-type (WT) (Setd2+/+) mESCs, including cell chromatin and cell junction (Figure S1A). Second, staining for typical ESC markers including alkaline phosphatase (AP), Oct4, and Sox2 were similar between WT and Setd2/ mESCs (Figures S1B and S1C). Furthermore, the expression levels of core pluripotency factors (Oct4, Nanog, and Tbx3), as analyzed by real-time PCR, were not significantly affected by disruption of Setd2 (Figure S1D). These data indicate that Setd2 is dispensable for self-renewal of mESCs. We next examined whether the differentiation capacity of mESCs would be affected when Setd2 was deficient. mESCs were induced to differentiate into embryo bodies (EBs) in leukemia inhibitory factor (LIF)-free suspension culture. The morphology of day (d) 3 and d6 Setd2/ EBs showed no obvious differences from that of WT EBs (Figure 1A). We then

transferred single d3 suspension-cultured EBs on gelatin-coated dish. After 3 days of culture, the rate of growth and migration of Setd2/ cells was much lower than that of WT EBs (Figure 1B), similar to the phenotype of Grb2-deficient EBs (Cheng et al., 1998). These data demonstrate that Setd2 plays an important role in mESC differentiation. Setd2 Is Required for mESC Differentiation toward Endoderm Because Grb2 is required for ESC endoderm differentiation (Cheng et al., 1998), we speculated that Setd2 might also play a role in this process. We collected d3, d6, and d9 suspensioncultured mouse EBs and measured the expression level of lineage markers by real-time PCR. Major primitive and visceral endoderm lineage markers, including Gata6, Gata4, Sox7, Hnf1b, Dab2, Cubn, Sox17, and Afp, were dramatically downregulated in Setd2/ EBs (Figures 1C and S2A). Immunostaining of frozen section for Gata6 and Dab2 in d5 EBs revealed a lack of spontaneous differentiation of Setd2/ EBs to endoderm, in contrast to WT EBs (Figure 1D). Flow cytometry analysis showed a decreased expression of Gata6 in d5 Setd2/ EBs (Figure S2B). However, pluripotent, trophectoderm, ectoderm, and mesoderm markers did not show any significant change (Figures S2C–S2F). Given the obvious vessel-remodeling defects in Setd2/ mouse embryo and yolk sac (YC) revealed by our previous work (Hu et al., 2010), we further evaluated the expression level of vessel developmental markers. Neither endotheliocyte Pecam1 (Cd31) nor smooth muscle cell (SMA, Sm22a, and Desmin) markers showed significant difference (Figure S2G), suggesting that Setd2 was not essential for mesoderm-derived cell differentiation but could influence the function of some mesoderm-derived tissue cells, such as endotheliocytes. Following a protocol designed for neural differentiation of mESC, we observed that both WT and Setd2/ mESCs could differentiate into tubb3-positive neural cells and that the morphology of both cell lines did not show any difference during the entire process. However, compared with the WT mESCs, deficiency of Setd2 decreased the number of neuroectodermal cells significantly, indicating Setd2 might be required for cell growth, but not differentiation of neuroectoderm (Figure S2H). It was reported that retinoic acid (RA) could induce ESCs to differentiate into endoderm and further differentiation could be carried out in LIF-free adherent culture in the presence of RA (Smith et al., 2010). Indeed, as compared with WT, the induction of endoderm-associated genes (Gata6 and Cubn) was significantly decreased in Setd2/ mESCs (Figure 1E). In the standard cultured conditions with medium containing LIF and serum (Williams et al., 1988), mESCs expressed not only pluripotency-associated transcription factors but also low levels of lineage markers such as Fgf5 and Hex, because the heterogeneous population contains both functional ESCs and early lineage-like ESCs such as endoderm-like ESCs (Canham et al., 2010; Niwa et al., 2000). The endoderm-like ESCs play an

(D) Immunofluorescence analysis of Gata6 (top) and Dab2 (bottom) expressed in d6 WT and Setd2/ EBs. The scale bar represents 10 mm. (E) Real-time PCR analysis of endoderm-specific genes during RA-induced endoderm differentiation of WT and Setd2/ mESCs. (F) Real-time PCR detection of basic expression level of endoderm, mesoderm, ectoderm, and trophectoderm lineage genes in WT and Setd2/ mESCs. In (C), (E), and (F), means ± SD from at least triplicate measurements are shown. *p < 0.05; **p < 0.01; ***p < 0.001.


Figure 2. Enforced Expression of Human SETD2 Can Rescue Endoderm Defects in Setd2/ mESCs (A) RT-PCR detection of human SETD2 fragment (503–2,564 aa), 6set (SETD2 fragment with set domain deletion) expression (top) in SETD2-transfected Setd2/ mESCs. Gapdh (bottom) was used as an internal control for RT-PCR. (B) Immunoblotting analysis of histone H3K36me3 (top) and H3 (bottom) in SETD2-transfected Setd2/ and WT mESCs. (C and D) Real-time PCR analysis of endoderm-specific genes in SETD2-transfected Setd2/ mESCs and d5 EBs. Means ± SD from triplicate measurements are shown. **p < 0.01; ***p < 0.001. (E) Left: morphology of d3 SETD2-transfected Setd2/ EBs after gelatin-attached culture for 3 days. Right: quantification of the areas of the cells growing around EBs. Data are presented as means ± SEM (n = 21–28); ***p < 0.001. GFP vector was used as a control for transfection. The scale bar represents 10 mm.

important role in determining the fate of subsequent endoderm differentiation (Canham et al., 2010). Having established the importance of Setd2 in regulating endoderm differentiation, we next asked whether the defects in Setd2/ mESCs occurred after the initiation of differentiation or due to the influence on the basal gene expression in mESCs. As compared with WT mESCs, the expression level of many endoderm-related genes, but not

other lineage markers, were downregulated in Setd2/ mESCs (Figure 1F). To further verify this phenotype, a human SETD2 fragment (503–2,564 amino acid [aa]), which contains the major domain of SETD2 gene, was transfected into Setd2/ mESCs (Figure 2A). As expected, increased level of H3K36me3 was observed in SETD2-transfected Setd2/ mESCs, and the expression levels of Gata6 and Sox7, but not T and Fgf5, were


augmented as well (Figures 2B and 2C). Moreover, the abnormal expression of endoderm lineage marker, as well as the growth and migration defects in Setd2/ EB cells could be rescued by enforced SETD2 expression. In order to verify that Setd2-dependent H3K36me3 is required for the expression of endodermrelated genes and the phenotype of EBs, set-domain-deleted SETD2 (6set) was transfected into Setd2/ mESCs. As compared with WT SETD2, 6set showed much less-effective rescue for the phenotype of Setd2/ mESCs, although some slight migration increase was observed, which might be due to the overexpression of other domains (Figures 2D and 2E). These data indicate that Setd2-dependent H3K36me3 principally regulates endoderm-specific basal gene expression in mESCs and induces subsequent differentiation of endoderm. Setd2 Is Required for Endoderm Development In Vivo Although Setd2 could specifically regulate endoderm development in vitro, it was still not clear whether it could regulate endoderm development in vivo. To address this, we subcutaneously injected mESCs into nonobese diabetic (NOD)-severe combined immunodeficiency (SCID) mice as a standard assay to test the potential of teratoma formation. Both WT and Setd2/ cells could form tumors in 4 weeks (Figure 3A), but the tumors in Setd2/ mESCs injected mice were much smaller than those with WT ones (Figures 3B and 3C). Histology analysis showed that both WT and Setd2/ mESCs were differentiated into ectoderm-derived nervous-like tissue and mesoderm-derived muscle/adipocyte tissues, respectively (Figure 3D, left and middle). All WT teratomas (4/4) formed normal intestinal epithelium tissue derived from endoderm, whereas only two out of four Setd2/ teratomas gave rise to much smaller intestinal epithelium tissue (Figure 3D, right). Morphological analysis of embryonic day 10 (E10) WT and Setd2/ YCs, mainly consisting of visceral endoderm and mesoderm, showed a disorganized phenotype and loss of classic columnar epithelial morphology of visceral endoderm in Setd2/ YCs (Figure 3E). Functional annotation analysis of differentially expressed genes between WT and Setd2/ YCs revealed that upregulated genes were highly enriched for those associated with ‘‘cytoskeletal protein binding’’ (Figure 3F). Real-time PCR validated the differential expression patterns of a panel of genes encoding these proteins, including both upand downregulated genes (Figure 3G). F-actin staining revealed no apical F-actin in YC visceral endoderm, whereas residual F-actin was maintained in some tight junctions (Figure 3H). Taken together, these data demonstrate that Setd2 is crucial for the maintenance of normal visceral endoderm morphology in vivo and might also influence its epithelial cell polarity. Setd2 Regulates Endoderm Development through the Fgfs-Fgfrs-Mek-Erk Signaling Pathway In order to understand how Setd2 regulates endoderm differentiation, RNA sequencing (RNA-seq) experiments were performed to compare potential differences in gene expression levels between WT and Setd2/ mESCs. Several interesting observations were made. First, although most highly transcribed genes showed no differences at transcriptional expression levels, a number of infrequently transcribed genes exhibited distinct expression patterns between WT and Setd2/ mESCs

(Figures S3A and S3B), a phenomena similar to what was described in yeast (Li et al., 2007). Second, in accordance with the recent report of SETD2 in human cell line (Carvalho et al., 2013), deregulation of intragenic transcriptional initiation was observed in 692 genes as compared with WT mESCs according to Carvalho’s criteria (Table S1). Third, although distinct alternative splicing pattern was found in 157 genes in Setd2/ mESCs (Figure S4; Table S2), consistent with the previous report (Luco et al., 2010), scrutiny of these genes did not reveal specific clues to the endoderm development defect. Last, but not least, by gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses (Figures S3C and S3D), we noticed that among 2,516 genes appeared to be deregulated in Setd2/ mESCs (Figure 4A; Table S3) and a subset of downregulated genes are related to the mitogen-activated protein kinase (MAPK) signaling pathway (Figures 4B and S3E). It should be pointed out that many MAPK-pathway-linked genes such as Grb2 and Stk40 have been associated with endoderm differentiation (Hamazaki et al., 2006; Li et al., 2010), suggesting that this pathway might also be involved in the regulation of endoderm development mediated by Setd2. Because there are many different MAPK kinases that can be activated by different signals (Figure 4C, left), we performed western blots to determine the differences in expression levels of the MAPK kinases between WT and Setd2/ mESCs. As shown in Figure 4C (right panel), phosphorylated (p)-Erk was downregulated, whereas other kinases appeared to be normal. It has been shown that the activation of Erk mainly depends on fibroblast growth factors (Fgfs) in mESCs (Kunath et al., 2007). Based on this, we cultured WT and Setd2/ mESCs in starvation condition for 12 hr, followed by Fgf2 and Fgf4 stimulation. The Erk signaling pathway was significantly reduced in Setd2/ mESCs, as compared with that in WT mESCs (Figure 4D). Subsequently, these two cell lines were treated, respectively, with inhibitors against Fgfr (PD073074), Mek (PD0325901), Jnk (SP600125), and phosphatidylinositol 3-kinase (LY294002) for 48 hr. It was found that PD073074 and PD0325901 specifically inhibited Gata6 expression (Figure 4E) and induced Nanog expression in both cell lines (Figure S5A). No significant changes of the expression level of T and Fgf5 were observed (Figures S5B and S5C). In order to confirm that the Erk pathway is specifically required for endoderm differentiation in our ESC line, we used a small hairpin RNA (shRNA) construct against Erk1 gene to knockdown its expression (Figure S5D). Effective suppression of endoderm-related gene expression was observed in the Erk knockdown mESCs (Figure 4F). Sodium orthovanadate (Na3VO4), a tyrosine phosphatase inhibitor, has been shown to be able to induce WT mESCs to differentiate into endoderm through the Erk pathway. Indeed, after treatment with 20 mM Na3VO4 for 24 hr, the level of p-Erk in Setd2/ mESCs was induced to a level similar to that in WT ESCs (Figure S5E). Moreover, Gata6 could be significantly induced in both cell lines (Figure 4G). On the other hand, in previous reports, no such induction effects existed in Grb2/ mESCs and c-Myc-overexpressed mESCs (Hamazaki et al., 2006; Smith et al., 2010), which is in agreement with the concept that Erk is located downstream to Grb2 in signal transduction. It has been known that the Erk pathway can also be activated by


Figure 3. Setd2 Regulates Visceral Endoderm Development In Vivo (A) Teratoma development in mice. WT (left) and Setd2/ (right) mESCs (5 3 106) were subcutaneous inoculated into NOD-SCID mice on both side. Dotted circles indicate the areas occupied by tumors. (B) Morphology of teratomas derived from WT (top) and Setd2/ (bottom) mESCs. (C) Weights of teratomas obtained after injection of WT (left; n = 9) and Setd2/ (right; n = 8) mESCs. Data are presented as means ± SEM; *p < 0.05. (D) Histology of WT (top) and Setd2/ (bottom) teratoma. Left: nervous tissue (ectoderm); middle: muscle and adipocyte tissue (mesoderm); right: intestinal epithelium tissue (endoderm). The scale bar represents 50 mm. Black dotted circles indicate intestinal epithelium vessel. (E) Light microscopy (left) and electron microscopy images (right) of E10 WT (top) and Setd2/ (bottom) yolk sac. Arrows and asterisks indicate visceral endoderm of yolk sac. The scale bar represents 1 mm. (F) DAVID functional annotation analysis of upregulated (left) and downregulated (right) genes between E10 WT and Setd2/ yolk sac. (G) Real-time PCR validation of differentially expressed cytoskeletal binding genes between E10 WT and Setd2/ yolk sac. Means ± SD from at least triplicate measurements are shown. *p < 0.05; **p < 0.01; ***p < 0.001. (H) F-actin immunofluorescence staining of E10 WT (left) and Setd2/ (right) yolk sac. The scale bar represents 10 mm.


Figure 4. Setd2 Regulates mESC Endoderm Differentiation through MAPK Signaling Pathway (A) 2D plots of total RNA-seq genes (17,827) in WT and Setd2/ mESCs. Up- and downregulated genes with fold change R2.0 are highlighted. (B) KEGG pathway analysis of downregulated genes in Setd2/ mESCs compared with WT. (C) Left: a schematic diagram of the MAPK signaling pathway, showing the major regulators and their inhibitors written in circles and squares, respectively. Right: immunoblotting analysis of the MAPK signaling pathway in WT and Setd2/ mESCs with antibodies indicated on the right. (D) Immunoblotting analysis of Erk phosphorylation in WT and Setd2/ mESCs that were serum starved and treated with 20 ng/ml Fgf2 and Fgf4 for 10 and 20 min. (E) Real-time PCR analysis of Gata6 expression in WT and Setd2/ mESCs treated with different inhibitors indicated on the right. (F) Real-time PCR analysis of endoderm-lineage-related genes in WT mESCs treated with mouse Erk1 shRNA. Scrambled shRNA (NC) was used as a negative control. (G) Real-time PCR analysis of Gata6 expression (right) in WT and Setd2/ mESCs treated with DMSO and Na3VO4. (H) Real-time PCR analysis of endoderm-lineage-related gene expression in Setd2/ mESCs transfected with N-RasG12D. All qPCR data are presented as means ± SD of triplicate measurements. *p < 0.05; **p < 0.01; ***p < 0.001.


N-RasG12D (Cuiffo and Ren, 2010). As expected, the expression level of p-Erk increased dramatically (Figure S5F), accompanied by upregulation of Gata6, Sox7, and Dab2 in N-RasG12Dtransfected Setd2/ ESCs (Figure 4H). All these results suggest that blockage of the Erk signaling pathway by Setd2 deficiency should be the major cause for endoderm defects. Fgfr3 Transcriptional Initiation Is Impaired in Setd2/ mESCs Having shown that the Mek-Erk pathway is required for the endoderm development, whereas Fgf2 and Fgf4 fail to activate Erk in Setd2/ ESCs, we set to address the mechanism by which Setd2 regulates this pathway. Comprehensive analysis of RNA-seq data and real-time PCR assay revealed that Fgfr3 and Fgfr4 were downregulated in Setd2/ ESCs (Figure 5A). The specific expression pattern of Fgfr3 during endoderm differentiation (Chen et al., 2000) raised the possibility that aberrant expression of Fgfr3 could play a role in the endoderm defects caused by Setd2 deficiency. Indeed, enforced expression of human FGFR3 led to upregulation of endoderm-related genes in Setd2/ mESCs (Figure 5B), suggesting that Setd2 regulates endoderm development through Fgfr3. We then moved to address how Setd2 modulates the expression of Fgfr3. It has been reported that Set2 could regulate transcriptional elongation in yeast (Li et al., 2002). We thus analyzed the intron-containing nascent transcripts to determine the potential defects of transcription elongation in Setd2/ mESCs by real-time PCR with primers covering different intronic regions of Fgfr3 gene. The expression of almost all of the examined intronic regions, from N-terminal to C-terminal region of Fgfr3, was declined at the similar rate in Setd2/ mESCs (Figure 5C), indicating that there should be no major elongation defects to this gene. Further investigation showed that all isoforms of mature Fgfr3 transcripts were downregulated (Figure 5D), suggesting that transcriptional initiation defects might exist in Fgfr3. An RNA pol II chromatin immunoprecipitation (ChIP) assay was performed to test this hypothesis. As expected, the binding affinity of RNA pol II was reduced significantly at 0.2 kb upstream from the transcriptional start site and intron1 (200 bp downstream from the transcriptional start site) of Fgfr3 in Setd2/ mESCs (Figure 5E). To address if Fgfr3 transcription could be specifically regulated by Setd2, we transfected human SETD2 fragment (503–2,564 aa) into Setd2/ mESCs and examined the Fgfr3 expression. Indeed, SETD2 upregulated the mRNA level of Fgfr3 (Figure 5F). We then cloned the 2 kb upstream promoter sequence of Fgfr3 into a PGL3-basic luciferase vector and transfected it into WT and Setd2/ mESCs. Dual-luciferase reporter assays showed that the activity of this region was much lower in mutant mESCs (Figure 5G). And these transcription defects could be rescued by enforced expression of SETD2 (Figure 5H). These data indicate that the transcriptional initiation of Fgfr3 is regulated by Setd2. Setd2-Mediated H3K36me3 Modification in Distal Promoter Region Regulates Fgfr3 Gene Transcription We next analyzed how Setd2 influenced the transcriptional initiation of Fgfr3. In view of the roles of H3K4 and H3K27 trimethylation in the regulation of gene transcriptional activation and inhibition, ChIP-PCR assays were performed to clarify whether

H3K4 and H3K27 trimethylation were involved in transcriptional initiation of Fgfr3 in Setd2/ mESCs. In fact, no obvious differences appeared when the results in Setd2/ mESCs were compared with those of WT mESCs (Figure S6A). In addition, Setd2 was not associated with the DNA methylation pattern of Fgfr3 CpG islands (Figure S6B). Ferrari et al. (2014) recently showed that knocking down of Setd2 could facilitate the conversion from H3K27me1 to H3K27me2, which may be associated with gene transcriptional regulation, affecting the expression level of Fgfr3 and other endoderm-related genes. To test this notion, we performed H3K27me1 and H3K27me2 ChIP assays on WT and Setd2/ mESCs. The ratios of these two modifications on endoderm-related genes did not exhibit significant differences between WT and Setd2/ mESCs (Figure S6C), suggesting that the conversion from H3K27me1 to H3K27me2 is not involved in regulation of Fgfr3 expression by Setd2. To address the possibility of other epigenetic abnormalities in Setd2/ mESCs, histone modification states of WT and Setd2/ mESCs cell lines were determined by western blot, and the results showed that H3K36me3, but not other histone modification, was dramatically downregulated in Setd2/ mESCs (Figure S6D). The distribution of H3K36me3 on the whole genome was thus analyzed based on published ChIP-seq data (Mikkelsen et al., 2007). About 18% of H3K36me3 peaks were found to distribute on upstream enhancer region (Figure 6A), suggesting that H3K36me3 could serve as a landmark for an enhanced regulation of transcriptional machinery. Our ChIPquantitative PCR (qPCR) results demonstrated that H3K36me3 mainly distributed in the coding region of Fgfr3 in WT mESCs, whereas the modification level in Fgfr3 proximal promoters (0.2 kb) and N- terminal region (intron1 and intron2) was very low, similar to that in Setd2/ mESCs. However, the H3K36me3 modification in its distal promoter region (0.5 kb, 1.5 kb, and 2 kb) was much more enriched and displayed significant differences between WT and Setd2/ mESCs (Figure 6B). Subsequently, we performed a ChIP assay of ectopically expressed FLAG-tagged SETD2 with anti-FLAG in Setd2/ mESCs and observed a direct binding of SETD2 on Fgfr3 locus (Figure S6E). These data indicate that Setd2 could regulate Fgfr3 expression from the distal promoter/enhancer. Furthermore, Arc and Cd97, two other genes downregulated by Setd2 disruption, showed similar H3K36me3 distribution pattern in mESCs (Figure 6C). Arc has also been known to be required for endoderm development (Liu et al., 2000). In order to investigate whether Setd2-dependent H3K36me3 could directly regulate the Fgfr3 transcription, we tethered human SETD2 on Fgfr3 promoter region by using Gal4-upstream activating sequence system. Luciferase reporter assay results showed that SETD2 fragment (503–2,564 aa) or Set region (1,418–1,714 aa) and low-charge region (LCR) (2,198–2,564 aa) of SETD2 could effectively activate Fgfr3 promoter in 293T cells and Setd2/ mESCs. To determine which domain plays a crucial role in Fgfr3 transcription, we deleted the set or LCR region in SETD2 fragment, respectively. As compared with the SETD2 fragment, 6set activates the Fgfr3 promoter with a significantly less efficiency. However, 6LCR displayed only a slightly lower transcriptional activation effect than the SETD2 fragment did (Figure 6D). Taken


Figure 5. Defect of Fgfr3 Transcriptional Initiation in Setd2/ mESCs (A) Real-time PCR analysis of Fgfr3 and Fgfr4 expression level in WT and Setd2/ mESCs. (B) Left: real-time PCR analysis of the expression level of endoderm-related genes in GFP- and FGFR3-overexpressed Setd2/ mESCs. Right: western blot analysis of the protein expression level of FGFR3 in GFP- and FGFR3-overexpressed Setd2/ mESCs. (C) Real-time PCR analysis of the expression level of Fgfr3 heterogeneous nuclear RNA by using indicated primer sets. Gapdh served as a control. (D) Real-time PCR analysis of different Fgfr3-splicing isoforms in WT and Setd2/ mESCs. (E) ChIP-qPCR analysis of RNA polymerase II binding to the Fgfr3 locus. Primer sets are indicated in the upper diagram. Groups without antibody were served as negative control for ChIP assay. (F) Real-time PCR analysis of Fgfr3 expression in Setd2/ mESCs transfected with human SETD2. GFP vector was used as a control for transfection. (G) Luciferase assay to analyze Fgfr3 promoter activity in WT and Setd2/ mESCs. (H) Coexpression with Setd2 in Setd2/ mESCs increases Fgfr3 promoter activity, as indicated by luciferase reporter assay. Averages ± SD are from triplicate measurements. *p < 0.05; **p < 0.01; ***p < 0.001.


Figure 6. H3K36me3 on Distal Promoter Is Essential for Fgfr3 Gene Transcription (A) Distribution of H3K36me3-enriched regions relative to mouse genome. Top: schematic for seven counting categories. Bottom: pie chart for percentage distribution of H3K36m3 tags in each category. Published H3K36me3 ChIP-seq data are from Mikkelsen et al. (2007).

(legend continued on next page)


together, these data show that Setd2-dependent H3K36me3 could directly regulate Fgfr3 transcriptional initiation in mESCs. DISCUSSION Recent studies of histone modifications in ESCs have focused on the function of H3K4me3, H3K27me3, H3K9me2/H3K9me3, and histone acetylation in self-renewal and lineage progenitor cell differentiation (Ang et al., 2011; Li et al., 2012; Loh et al., 2007; Peng et al., 2009). Unlike these modifications that mostly occur at the promoter region, H3K36me3 is known to mainly distribute in gene body regions in mESCs and other mammalian cells (Mikkelsen et al., 2007). For this reason, researches on H3K36me3 have been largely concentrated on these regions. It has also been shown that H3K36me3 can regulate transcriptional elongation through interacting with RNA pol II (Li et al., 2002), whereas H3K36me3 can suppress histone exchange to maintain the accuracy of transcription by RNA pol II in body regions after elongation (Venkatesh et al., 2012). Moreover, H3K36me3 has also been recognized as a marker for alternative splicing (Luco et al., 2010). However, no evidence of direct link between H3K36me3 and transcriptional initiation has yet been revealed. The impact of H3K36me3 on the actual cellular functions such as ESC self-renewal or lineage differentiation, as well as the underlying molecular mechanism, was largely unknown. In this study, we found that knockout of Setd2 could specifically reduce the basal expression level of many primitive endoderm-related genes in mESCs (Figure 1F), thereby blocking subsequent differentiation of EBs endoderm lineage (Figures 1C, S2A, and S2B), but had no effects on mESCs self-renewal and other lineage differentiation (Figures S1 and S2C–S2H). Further studies showed that the Erk pathway was significantly impaired in Setd2/ mESCs (Figures 4C and 4D). Erk inhibitor or activator could suppress or stimulate the expression of endoderm-related genes in both Setd2/ and WT mESCs (Figures 4E–4H). It has been reported that this pathway can trigger transition of ESCs from self-renewal to endoderm lineage commitment (Kunath et al., 2007). In mESCs, Fgf receptor tyrosine kinase can activate the Erk pathway and specify the primitive endoderm development in vivo (Frankenberg et al., 2011), whereas inhibiting this pathway allows the maintenance of ESC self-renewal ability (Ying et al., 2008). Our RNA-seq data combined with real-time PCR revealed that Fgfr3, which should be

specifically expressed in endoderm cells during differentiation (Chen et al., 2000), was downregulated in Setd2/ mESCs (Figure 5A). Enforced expression of this gene could effectively rescue the expression of primitive endoderm-related genes in Setd2/ mESCs (Figure 5B). We also found that, different from the role of H3K36me3 in transcriptional elongation, intragenic transcriptional initiation, and alternative splicing, the Fgfr3 transcriptional initiation was defected in Setd2/ mESCs (Figures 5C–5G). Although the noncoding region of exon1 of Fgfr3 exhibited alternative 50 splicing site defect (A5SS; Figure S4D), it hardly constituted a convincing mechanism to explain the low expression level of Fgfr3 observed in Setd2/ mESCs. Histone H3K4me3, H3K27me3, and DNA methylation are the major mechanisms for regulation of genetranscriptional initiation. Nevertheless, no differences in these regulatory mechanisms were detectable in Fgfr3 promotor region between WT and Setd2/ mESCs (Figures S6A and S6B). When the histone modification state was checked, only H3K36me3 was found to be downregulated in Setd2/ mESCs (Figure S6D). The results suggest that H3K36me3 might directly regulate the initiation of gene expression in mESCs. Overall speaking, despite the main distribution of H3K36me3 in body regions of actively transcribed genes, many H3K36me3 tags were found in distal promoter or upstream enhancer regions, whereas only scarce distribution of these tags existed in proximal promoter (Figure 6A). It is worth noting that this distal promoter distribution pattern of H3K36me3 was observed in Fgfr3 (Figure 6B). Moreover, artificial targeting of Set domain or SETD2 fragment (503–2,564 aa) to Fgfr3 promoter region led to its transcriptional activation (Figure 6D). Notably, some other genes downregulated by Setd2 deficiency, such as Arc and Cd97, also displayed this distribution pattern of H3K36me3 (Figure 6C). Hence, we put forward a model that Setd2 influences mESCs endoderm differentiation through regulating transcriptional initiation of Fgfr3, as shown in Figure 6E. Recently, several groups have revealed that human SETD2 was frequently mutated or downregulated in some types of cancer through whole-genome sequencing, such as cRCC (Cancer Genome Atlas Research Network, 2013; Dalgliesh et al., 2010; Duns et al., 2010) and acute leukemia (Mar et al., 2014; Zhu et al., 2014). In one report, H3K36me3 could interact with hMSH6 PWWP domain of hMutSa, which is important for DNA mismatch repair and maintains genome stability during replication, and RNAi knockdown of SETD2 could elevate spontaneous

(B) ChIP-qPCR analysis of H3K36me3 enrichment on Fgfr3 locus in WT and Setd2/ ESCs. Primer sets are indicated on the top. Groups without antibody were served as negative control for ChIP assay. (C) Real-time PCR analysis of gene expression (left) and ChIP-qPCR analysis of H3K36me3 enrichment (right) of two other genes, Arc and Cd97. Groups without antibody were served as negative control for ChIP assay. (D) Luciferase reporter assay of the role of Setd2 on Fgfr3 promoter activation. a: schematic representation of Gal4-tagged Setd2 constructs. b and c: induction of Gal4-Fgfr3 promoter-driven luciferase gene transcription by Gal4-tagged Setd2 in HEK293T (b) and Setd2/ mESCs (c). Averages ± SD are from triplicate measurements. *p < 0.05; **p < 0.01; ***p < 0.001. (E) Schematic representing of the role of Setd2 in endoderm differentiation of mESC. Left: in normal mESCs, the upstream enhancer region of Fgfr3 can be combined by unknown regulatory factors, which may recruit Setd2, and then Setd2 catalyzes H3K36me3 of this region. This marker can be recognized by some cofactors or readers and then help to form transcription initiation complex, which can regulate Fgfr3 expression. Fgfs stimulate Mek-Erk signaling cascade through Fgfr3 and keep endoderm-related genes on a basal expression level in mESCs, which is the prerequisite for normal endoderm differentiation after LIF withdrawal. Right: in Setd2/ mESCs, Fgfr3 loses H3K36me3 in the upstream enhancer region and its expression is decreased, which blocks the Mek-Erk signaling cascade and specifically influences basal expression level of the endoderm-related genes in mESCs, thus impairing subsequent endoderm differentiation after LIF withdrawal.


mutation frequency and microsatellite instability in tumor cell line (Li et al., 2013). These findings may provide an explanation for the mechanism of tumor formation. However, deregulation of gene expression in these cancers can hardly be ascribed to spontaneous mutation caused by SETD2 disruption. Our finding that H3K36me3 distal promoter distribution plays a role in endoderm differentiation may help to understand the role of SETD2 in tumorigenesis. EXPERIMENTAL PROCEDURES Detailed procedures are described in the Supplemental Experimental Procedures. ESC Culture and Embryonic Body Formation WT and Setd2/ ESCs were cultured on feeder cells in standard ES medium (Dulbecco’s modified Eagle’s medium supplemented with 15% fetal bovine serum [GIBCO], 1 3 nonessential amino acids, 2 mM L-glutamine, 100 U/ml penicillin, 100 mg/ml streptomycin, 0.1 mM b-mercaptoethanol, and 1,000 U/ml LIF [ESGRO; Millipore]). Trypsinized ESCs were transferred on gelatin-coated cell culture dishes for half an hour to remove feeder cells, and supernatant ESCs were collected, plated onto Petri dishes, and cultured in suspension without LIF for 3, 6, and 9 days. For attached cultures, day 3 suspension-cultured EBs were transferred on gelatin-coated cover glasses for 3 days and then stained with wrights and Giemsas. AP activity was monitored using the commercial Alkaline Phosphatase Detection kit (Vector Laboratories; SK-5400) according to the manufacturer’s instructions. Teratoma Formation 5 3 106 WT and Setd2/ ESCs were injected subcutaneously into the same NOD-SCID mice at different side. Four weeks later, mice were euthanized and tumors were removed and fixed in formalin for several days and then subsequently imbedded in paraffin, sectioned, and stained with hematoxylin and eosin for histological analysis. Mice were used according to animal care standards, and all protocols were approved by the Committee of Animal Use for Research at Shanghai Jiao Tong University School of Medicine (China). mRNA-Seq and qRT-PCR For RNA-seq, about 20 ng of Poly (A) RNA was purified from total RNA and then converted to double-stranded cDNA, and cDNA samples were sequenced using standard Solexa protocols. The sequencing reads were mapped to mouse genome (mm9) using TopHat (version 1.0.13). Avadis NGS (version 1.3) was used to calculate reads per kb per million reads (RPKM) values and find genes or different spliced genes. Differential genes were called at 2-fold changes using RPKM. GO and KEGG analyses were performed with DAVID (http://david. abcc.ncifcrf.gov/). For qRT-PCR, total RNA was reverse transcripted with the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems). Realtime PCR was performed with the SYBR Premix Ex Taq (TAKARA). The relative gene-expression levels were presented by 2^DDCt, and Gapdh was used as internal control. Primers will be provided upon request. 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.2014.08.031.

ACKNOWLEDGMENTS We specially thank Dr. Zhaohui Gu for helping us to analyze cryptic intragenic transcription initiation. We appreciate our colleagues at the Shanghai Institute of Hematology for constructive discussion. This work was supported in part by the Chinese National Key Basic Research Project (973, 2009CB825604, and 2013CB966801), National Natural Science Foundation of China (30971622), the Special Foundation for Doctor Discipline of Shanghai (11R21415500), the Postdoctor Foundation of China (2013M530199), and the Samuel Waxman Cancer Research Foundation. Received: February 4, 2014 Revised: May 6, 2014 Accepted: August 4, 2014 Published: September 18, 2014 REFERENCES Ang, Y.S., Tsai, S.Y., Lee, D.F., Monk, J., Su, J., Ratnakumar, K., Ding, J., Ge, Y., Darr, H., Chang, B., et al. (2011). Wdr5 mediates self-renewal and reprogramming via the embryonic stem cell core transcriptional network. Cell 145, 183–197. Barski, A., Cuddapah, S., Cui, K., Roh, T.Y., Schones, D.E., Wang, Z., Wei, G., Chepelev, I., and Zhao, K. (2007). High-resolution profiling of histone methylations in the human genome. Cell 129, 823–837. Bernstein, B.E., Mikkelsen, T.S., Xie, X., Kamal, M., Huebert, D.J., Cuff, J., Fry, B., Meissner, A., Wernig, M., Plath, K., et al. (2006). A bivalent chromatin structure marks key developmental genes in embryonic stem cells. Cell 125, 315–326. Bilodeau, S., Kagey, M.H., Frampton, G.M., Rahl, P.B., and Young, R.A. (2009). SetDB1 contributes to repression of genes encoding developmental regulators and maintenance of ES cell state. Genes Dev. 23, 2484–2489. Canham, M.A., Sharov, A.A., Ko, M.S., and Brickman, J.M. (2010). Functional heterogeneity of embryonic stem cells revealed through translational amplification of an early endodermal transcript. PLoS Biol. 8, e1000379. Carrozza, M.J., Li, B., Florens, L., Suganuma, T., Swanson, S.K., Lee, K.K., Shia, W.J., Anderson, S., Yates, J., Washburn, M.P., and Workman, J.L. (2005). Histone H3 methylation by Set2 directs deacetylation of coding regions by Rpd3S to suppress spurious intragenic transcription. Cell 123, 581–592. Carvalho, S., Raposo, A.C., Martins, F.B., Grosso, A.R., Sridhara, S.C., Rino, J., Carmo-Fonseca, M., and de Almeida, S.F. (2013). Histone methyltransferase SETD2 coordinates FACT recruitment with nucleosome dynamics during transcription. Nucleic Acids Res. 41, 2881–2893. Chen, Y., Li, X., Eswarakumar, V.P., Seger, R., and Lonai, P. (2000). Fibroblast growth factor (FGF) signaling through PI 3-kinase and Akt/PKB is required for embryoid body differentiation. Oncogene 19, 3750–3756. 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. Cheng, A.M., Saxton, T.M., Sakai, R., Kulkarni, S., Mbamalu, G., Vogel, W., Tortorice, C.G., Cardiff, R.D., Cross, J.C., Muller, W.J., and Pawson, T. (1998). Mammalian Grb2 regulates multiple steps in embryonic development and malignant transformation. Cell 95, 793–803. Cancer Genome Atlas Research Network (2013). Comprehensive molecular characterization of clear cell renal cell carcinoma. Nature 499, 43–49.

AUTHOR CONTRIBUTIONS

Cuiffo, B., and Ren, R. (2010). Palmitoylation of oncogenic NRAS is essential for leukemogenesis. Blood 115, 3598–3605.

Y. Zhang, S.X., and Y. Zhou performed and analyzed the experiments; M.S. and H.X. performed the bioinformatics analyses; Y.J. established Setd2 mESC line; Y.X., P.L., and X.S. discussed the results; and Y. Zhang, Z.C., Q.H., and S.C. designed the overall research and wrote the manuscript.

Dalgliesh, G.L., Furge, K., Greenman, C., Chen, L., Bignell, G., Butler, A., Davies, H., Edkins, S., Hardy, C., Latimer, C., et al. (2010). Systematic sequencing of renal carcinoma reveals inactivation of histone modifying genes. Nature 463, 360–363.


de Almeida, S.F., Grosso, A.R., Koch, F., Fenouil, R., Carvalho, S., Andrade, J., Levezinho, H., Gut, M., Eick, D., Gut, I., et al. (2011). Splicing enhances recruitment of methyltransferase HYPB/Setd2 and methylation of histone H3 Lys36. Nat. Struct. Mol. Biol. 18, 977–983. Duns, G., van den Berg, E., van Duivenbode, I., Osinga, J., Hollema, H., Hofstra, R.M., and Kok, K. (2010). Histone methyltransferase gene SETD2 is a novel tumor suppressor gene in clear cell renal cell carcinoma. Cancer Res. 70, 4287–4291. Efroni, S., Duttagupta, R., Cheng, J., Dehghani, H., Hoeppner, D.J., Dash, C., Bazett-Jones, D.P., Le Grice, S., McKay, R.D., Buetow, K.H., et al. (2008). Global transcription in pluripotent embryonic stem cells. Cell Stem Cell 2, 437–447.

Li, X., Li, L., Pandey, R., Byun, J.S., Gardner, K., Qin, Z., and Dou, Y. (2012). The histone acetyltransferase MOF is a key regulator of the embryonic stem cell core transcriptional network. Cell Stem Cell 11, 163–178. Li, F., Mao, G., Tong, D., Huang, J., Gu, L., Yang, W., and Li, G.-M. (2013). The histone mark H3K36me3 regulates human DNA mismatch repair through its interaction with MutSa. Cell 153, 590–600. Liu, D., Bei, D., Parmar, H., and Matus, A. (2000). Activity-regulated, cytoskeleton-associated protein (Arc) is essential for visceral endoderm organization during early embryogenesis. Mech. Dev. 92, 207–215. Loh, Y.H., Zhang, W., Chen, X., George, J., and Ng, H.H. (2007). Jmjd1a and Jmjd2c histone H3 Lys 9 demethylases regulate self-renewal in embryonic stem cells. Genes Dev. 21, 2545–2557.

Fazzio, T.G., Huff, J.T., and Panning, B. (2008). An RNAi screen of chromatin proteins identifies Tip60-p400 as a regulator of embryonic stem cell identity. Cell 134, 162–174.

Luco, R.F., Pan, Q., Tominaga, K., Blencowe, B.J., Pereira-Smith, O.M., and Misteli, T. (2010). Regulation of alternative splicing by histone modifications. Science 327, 996–1000.

Ferrari, K.J., Scelfo, A., Jammula, S., Cuomo, A., Barozzi, I., Stu¨tzer, A., Fischle, W., Bonaldi, T., and Pasini, D. (2014). Polycomb-dependent H3K27me1 and H3K27me2 regulate active transcription and enhancer fidelity. Mol. Cell 53, 49–62.

Mao, M., Fu, G., Wu, J.-S., Zhang, Q.-H., Zhou, J., Kan, L.-X., Huang, Q.-H., He, K.-L., Gu, B.-W., Han, Z.-G., et al. (1998). Identification of genes expressed in human CD34(+) hematopoietic stem/progenitor cells by expressed sequence tags and efficient full-length cDNA cloning. Proc. Natl. Acad. Sci. USA 95, 8175–8180.

Frankenberg, S., Gerbe, F., Bessonnard, S., Belville, C., Pouchin, P., Bardot, O., and Chazaud, C. (2011). Primitive endoderm differentiates via a three-step mechanism involving Nanog and RTK signaling. Dev. Cell 21, 1005–1013. Gaspar-Maia, A., Alajem, A., Polesso, F., Sridharan, R., Mason, M.J., Heidersbach, A., Ramalho-Santos, J., McManus, M.T., Plath, K., Meshorer, E., and Ramalho-Santos, M. (2009). Chd1 regulates open chromatin and pluripotency of embryonic stem cells. Nature 460, 863–868. Hamazaki, T., Kehoe, S.M., Nakano, T., and Terada, N. (2006). The Grb2/Mek pathway represses Nanog in murine embryonic stem cells. Mol. Cell. Biol. 26, 7539–7549. Hu, M., Sun, X.J., Zhang, Y.L., Kuang, Y., Hu, C.Q., Wu, W.L., Shen, S.H., Du, T.T., Li, H., He, F., et al. (2010). Histone H3 lysine 36 methyltransferase Hypb/ Setd2 is required for embryonic vascular remodeling. Proc. Natl. Acad. Sci. USA 107, 2956–2961.

Mar, B.G., Bullinger, L.B., McLean, K.M., Grauman, P.V., Harris, M.H., Stevenson, K., Neuberg, D.S., Sinha, A.U., Sallan, S.E., Silverman, L.B., et al. (2014). Mutations in epigenetic regulators including SETD2 are gained during relapse in paediatric acute lymphoblastic leukaemia. Nat Commun 5, 3469. Martin, C., and Zhang, Y. (2005). The diverse functions of histone lysine methylation. Nat. Rev. Mol. Cell Biol. 6, 838–849. 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. Niwa, H., Miyazaki, J., and Smith, A.G. (2000). Quantitative expression of Oct-3/4 defines differentiation, dedifferentiation or self-renewal of ES cells. Nat. Genet. 24, 372–376.

Jiang, H., Shukla, A., Wang, X., Chen, W.Y., Bernstein, B.E., and Roeder, R.G. (2011). Role for Dpy-30 in ES cell-fate specification by regulation of H3K4 methylation within bivalent domains. Cell 144, 513–525.

Pasini, D., Bracken, A.P., Hansen, J.B., Capillo, M., and Helin, K. (2007). The polycomb group protein Suz12 is required for embryonic stem cell differentiation. Mol. Cell. Biol. 27, 3769–3779.

Karimi, M.M., Goyal, P., Maksakova, I.A., Bilenky, M., Leung, D., Tang, J.X., Shinkai, Y., Mager, D.L., Jones, S., Hirst, M., and Lorincz, M.C. (2011). DNA methylation and SETDB1/H3K9me3 regulate predominantly distinct sets of genes, retroelements, and chimeric transcripts in mESCs. Cell Stem Cell 8, 676–687.

Peng, J.C., Valouev, A., Swigut, T., Zhang, J., Zhao, Y., Sidow, A., and Wysocka, J. (2009). Jarid2/Jumonji coordinates control of PRC2 enzymatic activity and target gene occupancy in pluripotent cells. Cell 139, 1290–1302.

Keogh, M.C., Kurdistani, S.K., Morris, S.A., Ahn, S.H., Podolny, V., Collins, S.R., Schuldiner, M., Chin, K., Punna, T., Thompson, N.J., et al. (2005). Cotranscriptional set2 methylation of histone H3 lysine 36 recruits a repressive Rpd3 complex. Cell 123, 593–605. Kolasinska-Zwierz, P., Down, T., Latorre, I., Liu, T., Liu, X.S., and Ahringer, J. (2009). Differential chromatin marking of introns and expressed exons by H3K36me3. Nat. Genet. 41, 376–381. Kunath, T., Saba-El-Leil, M.K., Almousailleakh, M., Wray, J., Meloche, S., and Smith, A. (2007). FGF stimulation of the Erk1/2 signalling cascade triggers transition of pluripotent embryonic stem cells from self-renewal to lineage commitment. Development 134, 2895–2902. Li, J., Moazed, D., and Gygi, S.P. (2002). Association of the histone methyltransferase Set2 with RNA polymerase II plays a role in transcription elongation. J. Biol. Chem. 277, 49383–49388. Li, B., Gogol, M., Carey, M., Pattenden, S.G., Seidel, C., and Workman, J.L. (2007). Infrequently transcribed long genes depend on the Set2/Rpd3S pathway for accurate transcription. Genes Dev. 21, 1422–1430. Li, L., Sun, L., Gao, F., Jiang, J., Yang, Y., Li, C., Gu, J., Wei, Z., Yang, A., Lu, R., et al. (2010). Stk40 links the pluripotency factor Oct4 to the Erk/MAPK pathway and controls extraembryonic endoderm differentiation. Proc. Natl. Acad. Sci. USA 107, 1402–1407.

Reynolds, N., Latos, P., Hynes-Allen, A., Loos, R., Leaford, D., O’Shaughnessy, A., Mosaku, O., Signolet, J., Brennecke, P., Kalkan, T., et al. (2012). NuRD suppresses pluripotency gene expression to promote transcriptional heterogeneity and lineage commitment. Cell Stem Cell 10, 583–594. Shen, X., Kim, W., Fujiwara, Y., Simon, M.D., Liu, Y., Mysliwiec, M.R., Yuan, G.C., Lee, Y., and Orkin, S.H. (2009). Jumonji modulates polycomb activity and self-renewal versus differentiation of stem cells. Cell 139, 1303–1314. Smith, K.N., Singh, A.M., and Dalton, S. (2010). Myc represses primitive endoderm differentiation in pluripotent stem cells. Cell Stem Cell 7, 343–354. Strahl, B.D., Grant, P.A., Briggs, S.D., Sun, Z.W., Bone, J.R., Caldwell, J.A., Mollah, S., Cook, R.G., Shabanowitz, J., Hunt, D.F., and Allis, C.D. (2002). Set2 is a nucleosomal histone H3-selective methyltransferase that mediates transcriptional repression. Mol. Cell. Biol. 22, 1298–1306. Sun, X.J., Wei, J., Wu, X.Y., Hu, M., Wang, L., Wang, H.H., Zhang, Q.H., Chen, S.J., Huang, Q.H., and Chen, Z. (2005). Identification and characterization of a novel human histone H3 lysine 36-specific methyltransferase. J. Biol. Chem. 280, 35261–35271. Venkatesh, S., Smolle, M., Li, H., Gogol, M.M., Saint, M., Kumar, S., Natarajan, K., and Workman, J.L. (2012). Set2 methylation of histone H3 lysine 36 suppresses histone exchange on transcribed genes. Nature 489, 452–455. Williams, R.L., Hilton, D.J., Pease, S., Willson, T.A., Stewart, C.L., Gearing, D.P., Wagner, E.F., Metcalf, D., Nicola, N.A., and Gough, N.M. (1988). Myeloid


leukaemia inhibitory factor maintains the developmental potential of embryonic stem cells. Nature 336, 684–687. Xie, L., Pelz, C., Wang, W., Bashar, A., Varlamova, O., Shadle, S., and Impey, S. (2011). KDM5B regulates embryonic stem cell self-renewal and represses cryptic intragenic transcription. EMBO J. 30, 1473–1484. Ying, Q.L., Wray, J., Nichols, J., Batlle-Morera, L., Doble, B., Woodgett, J., Cohen, P., and Smith, A. (2008). The ground state of embryonic stem cell self-renewal. Nature 453, 519–523.

Yuan, P., Han, J., Guo, G., Orlov, Y.L., Huss, M., Loh, Y.H., Yaw, L.P., Robson, P., Lim, B., and Ng, H.H. (2009). Eset partners with Oct4 to restrict extraembryonic trophoblast lineage potential in embryonic stem cells. Genes Dev. 23, 2507–2520. Zhu, X., He, F., Zeng, H., Ling, S., Chen, A., Wang, Y., Yan, X., Wei, W., Pang, Y., Cheng, H., et al. (2014). Identification of functional cooperative mutations of SETD2 in human acute leukemia. Nat. Genet. 46, 287–293.
