Author response: PHF13 is a molecular reader and transcriptional co-regulator of H3K4me2/3

Article Figures and data Abstract eLife digest Introduction Results Discussion Materials and methods Data availability References Decision letter Author response Article and author information Metrics Abstract PHF13 is a chromatin affiliated protein with a functional role in differentiation, cell division, DNA damage response and higher chromatin order. To gain insight into PHF13's ability to modulate these processes, we elucidate the mechanisms targeting PHF13 to chromatin, its genome wide localization and its molecular chromatin context. Size exclusion chromatography, mass spectrometry, X-ray crystallography and ChIP sequencing demonstrate that PHF13 binds chromatin in a multivalent fashion via direct interactions with H3K4me2/3 and DNA, and indirectly via interactions with PRC2 and RNA PolII. Furthermore, PHF13 depletion disrupted the interactions between PRC2, RNA PolII S5P, H3K4me3 and H3K27me3 and resulted in the up and down regulation of genes functionally enriched in transcriptional regulation, DNA binding, cell cycle, differentiation and chromatin organization. Together our findings argue that PHF13 is an H3K4me2/3 molecular reader and transcriptional co-regulator, affording it the ability to impact different chromatin processes. https://doi.org/10.7554/eLife.10607.001 eLife digest In human and other eukaryotic cells, DNA is packaged around proteins called histones to form a structure known as chromatin. Chemical tags added to the histones alter how the DNA is packaged and the activity of the genes encoded by that DNA. For example, many active genes are packaged around histone H3 proteins that have "Lysine 4 tri-methyl" tags attached to them. Another protein that is associated with chromatin is called PHF13 and it has several roles, including repairing damaged DNA. However, it was not known whether PHF13 binds to chromatin via the chemical tags, or in another way. Ho-Ryun, Xu, Fuchs et al. used several biochemical techniques in mouse and human cells to explore how PHF13 specifically interacts with chromatin. These experiments showed that PHF13 binds specifically to DNA and to two types of methyl tags (lysine 4-tri-methyl or lysine 4-di-methyl). These chemical tags are predominantly found at active promoters as well as at a small subset of less active promoters known as bivalent promoters. PHF13 interacted with other proteins on the chromatin that are known to either drive or repress gene activity and it's depletion affected the activity of many genes. Whether PHF13 increased or decreased gene activity depended on whether it was bound to active or bivalent promoters. The active promoters targeted by PHF13 had higher numbers of the tri-methyl tags whereas the di-methyl tags were more common on the bivalent promoters. These findings provide preliminary evidence that a protein binding to different methyl tags in the same place on histone H3 can have opposite effects on gene activity. Ho-Ryun, Xu, Fuchs et al. now intend to find out more about the other proteins that interact with PHF13 on chromatin. https://doi.org/10.7554/eLife.10607.002 Introduction PHF13, also known as Survival time associated PHD finger in Ovarian Cancer 1 (SPOC1), is a chromatin affiliated protein that is conserved from zebra fish to humans. PHF13 has been shown to modulate various processes including development (Bördlein et al., 2011), DNA damage (Frohns et al., 2014; Mund et al., 2012), cell cycle (Kinkley et al., 2009), antiviral host cell response (Schreiner et al., 2013) and higher order chromatin structure (Kinkley et al., 2009; Mund et al., 2012), underlining its biological importance. Furthermore, PHF13 expression and chromatin localization are temporally regulated (Kinkley et al., 2009) and its misregulation correlates with malignant phenotypes (Mohrmann et al., 2005) and defective differentiation (Bördlein et al., 2011). Together these observations argue that altered PHF13 expression has consequential outcomes and further underscores the necessity of understanding the molecular interplay and contexts governing PHF13 function. To date many chromatin binding domains recognizing specific histone posttranslational modifications, have been identified. Included in these are PHD domains which are recruited predominantly to methylated lysine residues (He et al., 2013; Li et al., 2006; Mansfield et al., 2011; Peña et al., 2006; Shi et al., 2006; Wysocka et al., 2006; Xie et al., 2012) with a few exceptions (Ali et al., 2012; Hu et al., 2009; Lan et al., 2007; Lange et al., 2008; Mansfield et al., 2011; Org et al., 2008; Tsai et al., 2010). The ability of different PHD domain containing proteins to recognize the same modified residue argues that additional factors may regulate their recruitment. Aside from differences in binding affinities, the sensitivity to adjacent sequence modifications also contributes to their specificity (Ali et al., 2013; Fiedler et al., 2008; Gatchalian et al., 2013; Iberg et al., 2008; Ramon-Maiques et al., 2007; Vermeulen et al., 2007; Yuan et al., 2012) indicating that many PHD domains read the local combinatorial chromatin environment. While the molecular readers containing PHD domains interact selectively with chromatin, they do so albeit with relatively weak affinities (Musselman and Kutateladze, 2011). To achieve a more stable chromatin association, many chromatin readers contact chromatin in a multivalent fashion, via multiple chromatin binding modules (Adams-Cioaba et al., 2012; Lange et al., 2008; Liu et al., 2013; Patel et al., 2013; Rothbart et al., 2013; Ruthenburg et al., 2011) or in complex with other chromatin readers, which cooperatively read multivalent chromatin signatures (Ballare et al., 2012; Nayak et al., 2011). The majority of characterized PHD domain proteins are known to be affiliated with chromatin modulating complexes (Morra et al., 2012; Todd and Picketts, 2012; Wysocka et al., 2006) and co-regulate important chromatin processes, including epigenetic programming (Lan et al., 2007; Wen et al., 2010), transcription (Fortschegger and Shiekhattar, 2011), DNA repair (Li et al., 2013; Mund et al., 2012), differentiation (Bördlein et al., 2011; Gatchalian et al., 2013), cell cycle (Kinkley et al., 2009; Lim et al., 2013) and higher chromatin order (Papait et al., 2008). Underscoring their importance in the co-regulation of chromatin function, the dysregulation or aberrant fusion of several PHD domain containing proteins has been shown to lead to genomic instability and cancer (Wang et al., 2009). PHF13 contains a single C-terminal PHD domain, which we demonstrate biochemically and structurally, is a molecular reader of H3K4me2/3. Additionally, we show that PHF13 directly interacts with DNA via a centrally located domain, indicating that it can form multivalent chromatin interactions. These interactions were confirmed by peptide binding assays, gel shift assays, ChIP sequencing and x-ray crystallography allowing us to additionally map these interactions and identify key cis-acting molecular determinants that affect PHF13's affinity for chromatin. Furthemore, we utilized mass spectrometry, size exclusion chromatography and co-immunoprecipitation experiments to identify Polycomb repressive complex 2 (PRC2) and RNA polymerase II (RNA PolII) complexes as novel PHF13 chromatin interaction partners. Consistently, PHF13 ChIP sequencing targets co-occurred with CpG rich DNA, H3K4me2/3, PRC2 and the hypophosphorylated, serine 5 and serine 7 phosphorylated forms of RNA PolII in murine embryonic stem cells (mESCs). PHF13 depletion in mESCs resulted in the reduced binding of SUZ12 and RNA PolII S5P to H3K4me3 and H3K27me3 and altered gene expression of a fraction of PHF13 bound genes. Genes that were up regulated upon PHF13 knockdown were enriched in H3K4me2/3, Polycomb and RNA PolII S5P, while down regulated genes were enriched in H3K4me2/3 and RNA PolII S2P and S5P. Finally, PHF13 target genes were enriched in the functional categories of transcription regulation, cell cycle, chromosome organization and differentiation, consistent with earlier publications describing a role of PHF13 in these processes (Bördlein et al., 2011; Kinkley et al., 2009; Mohrmann et al., 2005; Mund et al., 2012; Schreiner et al., 2013). Together, these findings argue that PHF13 is a transcriptional co-regulator and a novel H3K4me2/3 molecular reader. Results PHF13 interacts with nucleosomes and DNA We have previously demonstrated via differential nuclear fractionation experiments that PHF13 is predominantly affiliated with the chromatin fraction of nuclear lysate, implicating a role in chromatin function (Kinkley et al., 2009). Therefore, to gain clearer insight into whether PHF13 contacts chromatin directly, we explored the ability of GST-PHF13 and different GST-PHF13 deletion mutants to interact with recombinant mono-nucleosomes (Figure 1A–C). Mono-nucleosomes were generated using recombinant histone octamers and either a 200 bp DNA fragment (Figure 1B) or a 151 bp DNA fragment (Figure 1C) to recapitulate mono-nucleosomes with or without linker DNA. GST-PHF13 was found to very efficiently shift reconstituted mono-nucleosomes with DNA overhangs, similar to ISWI (positive control) and in contrast to GST (Figure 1B). We also noted that the free DNA in the reaction was notably absent in GST-PHF13 lanes in comparison to ISWI and GST, suggesting that PHF13 may also interact with DNA. To further map PHF13's interaction with mono-nucleosomes we analyzed the ability of different PHF13 fragments to shift mono-nucleosomes devoid of DNA overhangs (Figure 1C). Surprisingly, the PHD domain of PHF13 was not found to interact with the recombinant mono-nucleosomes and the interaction was mapped to the middle region of PHF13 (101–200; Figure 1C). These observations indicate that linker DNA is not necessary for PHF13 to affiliate with the nucleosomes and that the middle 100 aa of PHF13 is capable of forming a direct contact with recombinant mono-nucleosomes. Again we noted the lack of free DNA in the PHF13 101–200 lanes suggesting that this region may interact with free and nucleosomal complexed DNA. To test this idea and if a direct interaction with DNA exists, we performed DNA electrophoretic mobility shift assays (EMSA) using full-length GST-PHF13 and GST-PHF13 deletion fragments (Figure 1D–E). Since PHF13 lacks a predicted DNA binding domain, no DNA sequence specificity could be inferred. Therefore the EMSA's were performed using two random and unrelated DNA fragments, a 248 bp DNA (Figure 1D) or a 40 bp DNA fragment (Figure 1E). Increasing amounts of GST-PHF13, GST-ΔPHD and ACF1 (positive control) strongly retarded the electrophoretic mobility of the DNA, in contrast to GST alone (Figure 1D), indicating that PHF13 can directly interact with DNA independent of its PHD domain. Consistent with these and our previous observations, we mapped the DNA binding region to the middle 100aa (101–200) of PHF13, whereas the N- and C-terminal 100 amino acids of PHF13 did not shift the DNA (Figure 1E). Together, these findings support that PHF13 can directly contact chromatin via a direct interaction with DNA. Figure 1 Download asset Open asset PHF13 binds to DNA and recombinant nucleosomes. (A) Schematic of the putative domain structure of PHF13. (B) Mononucleosome EMSA using recombinant mononucleosomes reconstituted on a 200 bp DNA fragment (20 nM) and increasing concentrations of GST (150 nM, 1200 nM), GST-ISWI (28 nM, 226 nM) and GST-PHF13 (70 nM, 140 nM, 280 nM, 560 nM). (C) Mononucleosome EMSA using recombinant mononucleosomes reconstituted on a 151 bp fragment (20 nM) and increasing concentrations (80, 160, 320 nM) of GST-1-100, GST-101-200, GST-201-300, GST-PHD and PWWP (positive control). (D) EMSA: 248 fM of P32 radioactively labeled 248 bp DNA with increasing concentrations of GST (37.5, 150, 300 nM) GST-PHF13 (17, 34, 68, 102 and 135 nM) and GST-ΔPHD (18.5, 37, 74, 111, 148 nM) and ACF1 (10.5, 21, 31.5 and 42 nM). ACF1 served as a positive control. (E) EMSA: 10 nM Cy5 labeled 40 bp DNA with increasing concentrations (40, 80, 160 nM) of GST, GST-1-100, GST-101-200, GST-201-300, GST-PHD and GST-PWWP (positive control). The input DNA and mononucleosomes are indicated in B–E. https://doi.org/10.7554/eLife.10607.003 PHF13's PHD domain is a specific H3K4me2/me3 reader The inability of PHF13's PHD domain to interact with recombinant mono-nucleosomes which are devoid of histone post-translational modifications, suggested that it may selectively interact with modified histone tails. Consistently, previous mass spectrometry and bioinformatic predictions indicated that PHF13 may interact with H3K4me3 (Nikolov et al., 2011; Ruthenburg et al., 2007; Slama and Geman, 2011). To address this possibility, we evaluated the ability of full-length GST-PHF13 and different deletion fragments to directly interact with either non-modified or differentially methylated histone peptides or with H3K4me3 from nuclease digested chromatin lysate (Figure 2). As postulated both GST-PHF13 and the GST-PHF13 fragment containing the PHD domain (201–300) co-precipitated with the H3K4me1, H3K4me2 and H3K4me3 peptides but not with the un-modified H3K4me0 peptide (Figure 2A), explaining the inability of the PHD domain to interact with and shift the recombinant mono-nucleosomes. In contrast, GST-only and the GST-PHF13 fragments lacking the PHD domain (1–100 and 101–200) were incapable of co-precipitating with the biotinylated histone peptides. These data strongly support that PHF13's PHD domain interacts with methylated histone H3K4 and suggest that it has a stronger binding preference for H3K4me2 and H3K4me3, in comparison to H3K4me1 (Figure 2A). Figure 2 Download asset Open asset PHF13's PHD domain specifically interacts with H3K4me2/3 (A) Biotinylated histone peptide pull down.Equivalent amounts (1 µg) of GST, GST-PHF13 and GST-PHF13 deletion fragments 1–100, 100–200 and 200–300 aa were incubated with 1 µg of differentially modified biotinylated histone peptides and streptavidin beads. Co-precipitation of GST-proteins was analyzed using a GST specific antibody. (B) Fluorescence Polarization Assay. The dissociation constant of the PHF13-PHD only protein with differentially methylated H3 and H4 peptides. (C) GST pull down of H3K4me3 from nuclease digested chromatin lysates using GST-alone, GST-PHF13 and the indicated GST-PHF13 deletion and point mutant proteins. Precipitation of H3K4me3 was analyzed with a specific antibody. Amount of GST proteins were controlled by Coomassie staining. https://doi.org/10.7554/eLife.10607.004 To further evaluate the specificity of PHF13's PHD domain for H3K4me2/3 in relation to other methyl lysine residues and to calculate approximate binding constants we employed fluorescence polarization analysis using the PHD domain only of PHF13 and differentially modified histone H3 peptides (Figure 2B). PHF13's PHD domain bound strongest to H3K4me3 (Kd = 88.5 ± 20.6 µM), with a slightly reduced affinity to H3K4me2 (Kd = 122 ± 29.4 µM) and with a ~4-fold reduced affinity to H3K4me1 (Kd = 332.4 ± 82.6 µM). In contrast to H3K4, no detectable binding was observed for other tri-methylated histone peptides demonstrating a strong preference and specificity of the PHF13 PHD domain for H3K4me2/3 (Figure 2B). Finally, to explore whether PHF13's PHD domain can interact with H3K4me3 in a native chromatin context, we analyzed the ability of the different recombinant PHF13 proteins to precipitate H3K4me3 from nuclease digested Hela chromatin lysates (Figure 2C). These experiments showed that PHF13 and the deletion fragment containing its PHD domain (201–300) were capable and sufficient to precipitate H3K4me3, an interaction that was lost by deletion of the PHD domain, specific point mutations in the PHD domain (M246A and W255A) that are predicted to disrupt the PHD domain structure, and in fragments not containing the PHD domain (1–100 and 101–200). Together these findings demonstrate that PHF13's PHD domain is a specific H3K4me2/3 molecular reader and together with its DNA binding ability that PHF13 can interact with chromatin in a direct and multivalent manner. 3-D structural crystallography of PHF13's PHD domain and H3K4me3 To gain structural insight into this specific recognition of histone H3K4me3 by the PHD domain of PHF13, we determined the crystal structures of both the apo-PHF13 PHD domain (aa 226–280) and the PHF13 PHD-H3K4me3 peptide complex (Figure 3A-C). The diffraction data and refinement statistics for these structures can be found in Figure 3—source data 1. Similar to other PHD domains (Adams-Cioaba and Min, 2009; Musselman and Kutateladze, 2009), the PHD domain of PHF13 folds into a canonical Cys4-His-Cys3 (or C4HC3) motif, that coordinates two zinc ions with a well-conserved globular domain (Figure 3A). In the complex structure of PHF13 PHD-H3K4me3, we can see that the histone H3K4me3 peptide is bound against the central double-strand anti-parallel β-sheet of the PHF13 PHD domain and completes a three-stranded β-sheet (Figure 3A). The first four residues of the H3K4me3 peptide are embedded in an exposed binding groove on the surface of the PHF13 PHD domain (Figure 3B–C). Similar to other methyl lysine binding proteins (Adams-Cioaba and Min, 2009), the trimethyl lysine 4 (K4me3) is bound in an aromatic cage formed by residues F241, M246 and W255 of PHF13 (Figure 3B–C). In addition, the K4me3 residue also forms two main chain hydrogen bonds with M246 (Figure 3C). The first residue alanine (Ala1) of the H3K4me3 peptide is anchored in a small and secluded pocket created by I247, V268, P269, E270 and F272, and the backbone amine group of Ala1 forms a hydrogen bond with the backbone carbonyl oxygen of P269 and E270 in PHF13 (Figure 3C). The limited length between the aromatic cage and the Ala1 binding pocket and the restricted nature of the Ala1 binding pocket determines that the PHF13 PHD domain selectively recognizes only methylated histone H3K4. Consistently, PHF13 PHD domain was not co-precipitated by biotinylated H3K4me0 (Figure 2A), nor did it interact with recombinant mononucleosomes (Figure 1B) or H3K4me0 fluorescent peptides (Figure 2B). The Arg2 residue in the H3K4me3 peptide is bound by two backbone hydrogen bonds with E248 (Figure 3B–C). Mutation of any of the residues affiliated with the H3K4me3 or H3R2 binding pockets (T234A, F241A, M246A, W255A and E248A) abrogated binding of PHF13's PHD domain to H3K4me3 (Figure 2C and 3D). Figure 3 with 1 supplement see all Download asset Open asset Crystal structure of PHF13's PHD domain. (A–B) The crystal structure of apo-PHF13 PHD domain (A) and PHF13 PHD finger in complex with H3K4me3 (B). (C) Electrostatic surface potential of PHF13's PHD finger in complex with H3K4me3. Dashed lines represent intermolecular hydrogen bonds. (D) Fluorescence Polarization Assay - The dissociation constant of PHF13 and PHF13 PHD point mutants with H3K4me3. (E) Isothermal titration Calorimetry (ITC) - Binding affinity of PHF13's PHD domain for H3R2K4me3 (left panel) or H3R2me2aK4me3 (right panel). https://doi.org/10.7554/eLife.10607.005 Figure 3—source data 1 Data collection and refinement statistics – Detailed specifications of the data obtained from the crystallization of Apo-PHF13-PHD and the PHF13-PHD-H3K4me3 complex. https://doi.org/10.7554/eLife.10607.006 Download elife-10607-fig3-data1-v4.docx Figure 3—source data 2 PHF13 PHD domain binding to a differentially methylated histone peptide tail chip. (A) Binding of His-PHF13-PHD-only to a histone peptide array was detected with an anti-His antibody. Boxes denote the positive controls (12 x histidine) and an interaction with H3K4me3 when the neighboring R2 is not di-methylated. (B) Peptide key of the spotted differentially modified histone peptides on the histone peptide array. The length, position and sequence of the spotted peptides are annotated in the table. https://doi.org/10.7554/eLife.10607.007 Download elife-10607-fig3-data2-v4.docx The surface charge representation shows that Arg2 sits in a shallow negatively charged pocket, reminiscent of the Arg2 binding in the ING2 and BPTF2 PHD domain-H3K4me3 complex structures, but in contrast to the RAG2-H3K4me3 complex structure, in which Arg2 is bound in a relatively hydrophobic pocket (Figure 3C and Figure 3—figure supplement 1). This indicates that similar to ING2 and BPTF, that Arg2 methylation should diminish PHF13 binding to H3K4me3, whereas it has been shown to enhance RAG2 binding (Iberg et al., 2008; Yuan et al., 2012). Therefore to address this prediction and test the influence of Arg2 methylation on PHF13's PHD domain H3K4me3 binding ability we performed isothermal titration calorimetry (ITC; Figure 3E). Our binding results show that the binding of PHF13's PHD domain to H3K4me3 was diminished by the simultaneous asymmetric dimethylation of Arg2 (H3R2me2aK4me3). Consistently, the binding of PHF13's PHD domain to H3K4me3 on a differentially modified histone peptide chip was also significantly reduced in the presence of dimethyl Arg2 (Figure 3—source data 2). Together, these results confirm that the PHD domain of PHF13 is an H3K4me3 reader and demonstrate the structural relationship between them. Furthermore, they identify key residues within PHF13's PHD domain involved in mediating this molecular interaction as well as inhibiting modifications of neighboring histone residues that opposes their interaction. PHF13 interacts with PRC2 and RNA polymerase II To gain additional information about PHF13 chromatin interactions, we immunoprecipitated PHF13 from the nuclease digested chromatin fraction of mESCs and analyzed the co-immunoprecipitating proteins in comparison to an IgG control by mass spectrometry (Figure 4A and Figure 4—source data 1). Independent PHF13 IPs were digested by trypsin or LysC (Figure 4—figure supplement 1) prior to MS analysis and the co-precipitating proteins were combined (Figure 4A and Figure 4—source data 1). Similar complexes were retrieved regardless of whether the results were (Figure 4A and Figure 4—source data 1) or (Figure 4—figure supplement 1). For all interactions only a single peptide was detected the mass spectrometry is to their (Figure 4—figure supplement The mass spectrometry findings confirmed previous observations that PHF13 interacts with DNA damage response and et al., 2012) and that PHF13 interacts with an RNA polymerase complex as well as with several of PRC2, and (Figure of the associated functional affiliated with PHF13 indicated a role in RNA processes, gene RNA and binding (Figure 4—source data 2). To these observations we performed co-immunoprecipitation in mESCs which confirmed that PHF13, SUZ12 and RNA PolII S5P interact with other and that they all with H3K4me3 and H3K27me3 (Figure To provide additional evidence that these proteins in a common chromatin molecular complex we performed size exclusion chromatography from nuclease digested chromatin fraction of mESCs to for PHF13 with PRC2, RNA H3K4me3 and H3K27me3 (Figure PHF13 was in different a very molecular fraction it with RNA PolII S5P and H3K4me3 and a molecular fraction it with RNA PolII S5P and PRC2 and H3K4me3 and H3K27me3 and in molecular it only with RNA PolII (Figure Together, these findings the that PHF13 with different RNA PolII containing PRC2 and histones and that do To address this we the very and and whether PHF13 could RNA PRC2, H3K4me3 and H3K27me3 from these different (Figure PHF13 interacted with RNA PolII S2P and in all and as it interacted with RNA PolII S5P, H3K4me3 and H3K27me3 only in the molecular and Furthermore, PRC2 predominantly co-precipitated with RNA PHF13, H3K4me3 and H3K27me3 in Together, these findings that PHF13 interacts with RNA PolII in different complex and support the idea that PHF13 may a role in gene Figure 4 with 4 see all Download asset Open asset PHF13 interacts with RNA polymerase II and PRC2 (A) functional protein of PHF13 chromatin interactions as by mass interactions were and an interaction with PRC2 and RNA polymerase II complex. (B) of co-immunoprecipitating interactions from control PHF13, PolII S2P and PolII S5P from nuclease digested chromatin fraction of (C) of fraction numbers from a size exclusion chromatography using nuclease digested chromatin fraction of The of RNA PolII with PRC2, H3K4me3, H3K27me3 and PHF13 in is by (D) of proteins from an IgG and PHF13 of the chromatography and for SUZ12 and H3K27me3 are to the that they represent a in relation to the Figure 4—source data 1 spectrometry of PHF13 proteins. the proteins by mass spectrometry obtained from PHF13 from nuclease digested chromatin lysates from that were either trypsin digested or LysC is the of peptides as well as the or Download Figure 4—source data 2 functional of mass spectrometry proteins. are the biological processes and molecular and their determined using by the functional protein Download PHF13 with H3K4me2/3, PRC2, RNA and CpG Our results strongly suggest that PHF13 binds nucleosomes by H3K4me2/3 via its C-terminal PHD domain – an interaction that can be further by PHF13's ability to DNA and chromatin affiliated PRC2 and RNA PolII. we that PHF13 should with H3K4me2/3, PRC2 and RNA PolII and DNA sequence in and genome To test this idea we PHF13 bound by chromatin by sequencing and their with or bound by H3K4me1, H3K4me3, and the different and phosphorylated forms of RNA PolII (Figure obtained from for mESCs and We PHF13 bound and confirmed a few of by ChIP (Figure supplement PHF13 bound strongly with H3K4me3 and H3K4me2 whereas this was decreased with H3K4me1 Figure with the biochemical and structural these observations support that PHF13 is a H3K4me2/3 reader. Figure 5 with 2 see all Download asset Open asset shows a genome wide with methylated H3K4, CpG PRC2 and (A) for the of (B) for the around PHF13 are two of that are the of a of all the from PHF13, the and the CpG the the for the two is we the of PHF13 bound with and H3K27me3 repressive modifications (Figure we observed no with a that is to H3K4me3 in mESCs at genomic et al., 2009). In contrast we did find a with H3K27me3 repressive Similar to H3K27me3 we observed a but of PHF13 with and SUZ12 whereas of and of SUZ12 with PHF13. These findings demonstrate that PHF13 also at a subset of Polycomb and is

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.