|Three distinct patterns of histone H3Y41 phosphorylation mark active genes.|
|Jump to Full Text|
|PMID: 22999934 Owner: NLM Status: MEDLINE|
|The JAK2 tyrosine kinase is a critical mediator of cytokine-induced signaling. It plays a role in the nucleus, where it regulates transcription by phosphorylating histone H3 at tyrosine 41 (H3Y41ph). We used chromatin immunoprecipitation coupled to massively parallel DNA sequencing (ChIP-seq) to define the genome-wide pattern of H3Y41ph in human erythroid leukemia cells. Our results indicate that H3Y41ph is located at three distinct sites: (1) at a subset of active promoters, where it overlaps with H3K4me3, (2) at distal cis-regulatory elements, where it coincides with the binding of STAT5, and (3) throughout the transcribed regions of active, tissue-specific hematopoietic genes. Together, these data extend our understanding of this conserved and essential signaling pathway and provide insight into the mechanisms by which extracellular stimuli may lead to the coordinated regulation of transcription.|
|Mark A Dawson; Samuel D Foster; Andrew J Bannister; Samuel C Robson; Rebecca Hannah; Xiaonan Wang; Blerta Xhemalce; Andrew D Wood; Anthony R Green; Berthold Göttgens; Tony Kouzarides|
Related Documents :
|8362214 - Changes in gastric mucosal ornithine decarboxylase and tyrosine phosphorylation of prot...
23343194 - Wnt/β-catenin signalling: from plasma membrane to nucleus.
21459084 - Mammary branch initiation and extension are inhibited by separate pathways downstream o...
23049844 - Structural and functional characterization of the protein kinase mps1 in arabidopsis th...
22631074 - Determination of primary sequence specificity of arabidopsis map kinases, mpk3 and mpk6...
21187344 - Signalling to and from the secretory pathway.
|Type: Journal Article; Research Support, Non-U.S. Gov't Date: 2012-09-20|
|Title: Cell reports Volume: 2 ISSN: 2211-1247 ISO Abbreviation: Cell Rep Publication Date: 2012 Sep|
|Created Date: 2012-10-01 Completed Date: 2013-02-26 Revised Date: 2014-07-22|
Medline Journal Info:
|Nlm Unique ID: 101573691 Medline TA: Cell Rep Country: United States|
|Languages: eng Pagination: 470-7 Citation Subset: IM|
|Copyright © 2012 The Authors. Published by Elsevier Inc. All rights reserved.|
|APA/MLA Format Download EndNote Download BibTex|
Cell Line, Tumor
Histones / genetics, metabolism*
Janus Kinase 2 / genetics, metabolism*
Phosphorylation / physiology
Promoter Regions, Genetic / physiology*
STAT5 Transcription Factor / genetics, metabolism*
Signal Transduction / physiology*
Transcription, Genetic / physiology*
|079249//Wellcome Trust; 092096//Wellcome Trust; 12765//Cancer Research UK; G0800784//Medical Research Council; //Cancer Research UK; //Medical Research Council; //Wellcome Trust|
|0/Histones; 0/STAT5 Transcription Factor; EC 18.104.22.168/JAK2 protein, human; EC 22.214.171.124/Janus Kinase 2|
Journal ID (nlm-ta): Cell Rep
Journal ID (iso-abbrev): Cell Rep
Publisher: Cell Press
© 2012 The Authors
Received Day: 23 Month: 4 Year: 2012
Revision Received Day: 16 Month: 7 Year: 2012
Accepted Day: 16 Month: 8 Year: 2012
pmc-release publication date: Day: 27 Month: 9 Year: 2012
Print publication date: Day: 27 Month: 9 Year: 2012
Volume: 2 Issue: 3
First Page: 470 Last Page: 477
PubMed Id: 22999934
Publisher Id: CELREP140
|Three Distinct Patterns of Histone H3Y41 Phosphorylation Mark Active Genes|
|Mark A. Dawson1234|
|Samuel D. Foster24|
|Andrew J. Bannister14|
|Samuel C. Robson1|
|Andrew D. Wood2|
|Anthony R. Green23|
|Berthold Göttgens25∗||Email: firstname.lastname@example.org|
|Tony Kouzarides15∗∗||Email: email@example.com|
1Gurdon Institute and Department of Pathology, Tennis Court Road, Cambridge, CB2 1QN, UK
2Department of Haematology, Cambridge Institute for Medical Research and The Wellcome Trust and MRC Stem Cell Institute, University of Cambridge, Cambridge, CB2 0XY, UK
3Addenbrooke’s Hospital, University of Cambridge, Cambridge, CB2 0XY, UK
|∗Corresponding author firstname.lastname@example.org
∗∗Corresponding author email@example.com
4These authors contributed equally to this work
5These authors contributed equally to this work
The canonical JAK2-STAT5 pathway is one of the most widely studied cellular signaling cascades and is critical for normal hematopoiesis (Decker and Müller, 2012). It is also now established that hyperactivation of JAK-STAT signaling is a common event in several hematological neoplasms (Chen et al., 2012). These findings underline the importance of a comprehensive understanding of the function of this signaling pathway.
Until recently, JAKs were largely ascribed the cytoplasmic function of phosphorylating STAT family members to promote STAT dimerization, nuclear translocation, and binding to cis-regulatory elements to regulate transcription (Decker and Müller, 2012). However, recent reports have highlighted that members of the JAK family also appear to have specific roles in the nuclei of various tissues (Dawson et al., 2009; Griffiths et al., 2011; Kamakura et al., 2004; Lefrancois-Martinez et al., 2011; Liu et al., 2011; Nilsson et al., 2006; Rui et al., 2010). Indeed, we have demonstrated that JAK2 functions as a histone tyrosine kinase, phosphorylating H3Y41 to regulate chromatin binding of HP1α in hematopoietic cells (Dawson et al., 2009). H3Y41ph represents the only characterized tyrosine phosphorylation in nonvariant mammalian histones. Here we report the genome-wide analysis of this modification by chromatin immunoprecipitation (ChIP) coupled to massively parallel DNA sequencing (ChIP-seq). These data provide insights into the mechanisms of transcription control exerted by this conserved and essential signaling pathway.
H3Y41ph is present at the transcriptional start site (TSS) of the LMO2 gene, overlapping the active H3K4me3 modification (Dawson et al., 2009), and is linked to transcriptional activation. To ascertain whether this relationship is maintained on a genomic scale, we generated a global map of the spatial distribution of H3Y41ph and H3K4me3 by performing ChIP-seq using chromatin from human erythroid leukemia (HEL) cells. When ranked in ascending order of H3Y41ph and H3K4me3 levels, global analysis of gene promoters demonstrates that these two histone modifications are coincident at many loci (Figure 1A, areas I and II).
A detailed analysis then examined the global relationship between H3K4me3 and H3Y41ph across TSSs. 28,996 TSSs were split into 10 sets (the first nine segments have 2,900 and the tenth has 2,896) based on levels of H3K4me3 (Figure 1B). These sets were ranked according to the levels of H3K4me3 at the TSS, as this has previously been shown to quantitatively reflect transcriptional activity; the most highly transcribed genes have the highest levels of H3K4me3 and vice versa (Barski et al., 2007; Min et al., 2011). Once ranked according to H3K4me3 levels, we then looked at the distribution and levels of H3Y41ph across the same promoter cohorts. This analysis revealed that, when present, H3Y41ph is often at the TSS of active genes, where it correlates closely with H3K4me3 (Figure 1C). Importantly, selective inhibition of JAK2 activity for 4 hours demonstrated a significant global reduction of H3Y41ph at the TSS of genes marked with this histone modification (Figure S1). Overall, these data indicate that, as predicted from our previous analysis of the LMO2 gene, H3Y41ph coincides with H3K4me3 surrounding the TSSs of a subset of transcriptionally active genes.
Analysis of individual gene loci revealed that H3Y41ph is also enriched at noncoding regions of several JAK-STAT target genes, such as CISH and PIM1 (Figures 2A and 2E). These sites have a high degree of sequence conservation and are also marked by H3K4me1, consistent with a potential function as distal cis-regulatory elements (Figure S2A). Indeed, some of these elements have already been defined as enhancer elements, such as those present at ID1 (Wood et al., 2009) and CISH (Nagy et al., 2009). Together these data raise the possibility that H3Y41ph also delineates a subset of conserved elements outside of genes, at least some of which are known functional STAT5 enhancers.
It is firmly established that JAK2 and STAT5 physically and functionally interact (Barahmand-Pour et al., 1998). Thus far, only a few direct target genes bound by STAT5, such as CISH, PIM1, ID1, and BCLXL, have been characterized in the erythroid lineage (Dumon et al., 1999; Matsumoto et al., 1997; Mui et al., 1996; Wood et al., 2009). As HEL cells are a model cell line for malignant erythropoiesis, we investigated the genome-wide distribution of STAT5 by ChIP-seq to see if it coincides with H3Y41ph. Bioinformatic analysis of our newly generated STAT5 ChIP-seq data set identified 676 high-confidence STAT5 binding sites (Table S1). Importantly, the few previously known functional STAT5 binding sites, such as those present at ID1 and BCLXL, were all identified as significant peaks (Figure S2B). Moreover, we now provide several hundred potential targets of STAT5 (Table S1), functional validation of which may uncover new insights into mechanisms of pathogenesis in JAK2-STAT5-driven neoplasms.
The quality of our STAT5 ChIP-seq data set was confirmed by an unbiased de novo motif discovery, which identified the previously reported TTCYNRGAA STAT consensus binding site as the only significantly enriched motif (Figure 2B). Furthermore, we observed an enrichment of STAT5 binding proximal to gene promoters, although the vast majority of STAT5 binding events were seen in nonpromoter regions (Figure 2C). These findings are analogous to those reported from recent genome-wide surveys of STAT5 binding in mouse liver (Zhang et al., 2012). Our ChIP-seq data indicate that both the H3Y41ph modification and STAT5 binding often occur at the same genomic regions (Figure 2E; Table S1). Indeed, when the spatial distribution of the H3Y41ph peaks are determined relative to the STAT5 peaks, we find that the location of H3Y41ph is directly coincident with the STAT5 binding event (Figures 2E and S2A).
Although we observe a significant coincidence of H3Y41ph with STAT5, our data clearly demonstrate that the majority of H3Y41ph is independent of STAT5 (Tables S1 and S2). Nevertheless, the coincidence of H3Y41ph with STAT5 suggests that the interaction of JAK2/STAT5 may extend beyond the cytoplasm and culminate at the chromatin of a key subset of genes. We therefore asked if JAK2 signaling coordinates both the posttranslational modification of histones and the concomitant binding of STAT5 to its target DNA. For this, we performed both an H3Y41ph and STAT5 ChIP-seq analysis with chromatin from HEL cells that had been treated with the selective JAK2 inhibitor, TG101209, for 4 hours. Global analysis demonstrated that H3Y41ph and STAT5 binding events were significantly decreased after JAK2 inhibition (Figures 2D and 2E). These findings are consistent with our previous results derived from western blot analyses, where we see a global reduction in H3Y41ph following exposure of HEL cells to a 4 hr treatment with TG101209 (Dawson et al., 2009). The dual regulation of H3Y41ph and STAT5 binding by JAK2 signaling is specifically demonstrated at the PIM1 locus in Figure 2E. Several loci identified in the ChIP-seq data were independently validated by ChIP-PCR at various JAK2-STAT5 target genes in HEL cells (Figure S2C). Moreover, these results were validated using AT9283 (Dawson et al., 2010), a second chemically distinct JAK2 inhibitor (Figure S2D). Finally, we demonstrate the broader application of these data by observing similar findings in SET-2 cells, a megakaryoblastic cell line containing JAK2V617F (Figure S2E). Taken together, these data are consistent with JAK2 being a major histone kinase responsible for H3Y41ph and indicate that signaling via JAK2 concomitantly regulates both H3Y41ph and STAT5 binding at certain JAK2/STAT5 target genes.
We next asked whether cytokine induction via JAK2 dynamically regulates the appearance of H3Y41ph and the binding of STAT5 on cytokine-regulated genes. TF1 cells (cytokine-dependent human erythroid-leukemia cells diploid for wild-type [WT] JAK2) were stimulated for 1 hr with IL3, a cytokine that signals primarily via JAK2 (Decker and Müller, 2012). The expression of two well-characterized cytokine inducible JAK2-STAT5 target genes, PIM1 (Mui et al., 1996) and CISH (Matsumoto et al., 1997), were markedly increased by IL3, whereas the transcription of the β2M housekeeping gene was unchanged (data not shown). Importantly, ChIP analyses from the same cell populations clearly demonstrated that the increased gene expression of PIM1 and CISH was linked to an increase of both H3Y41ph and STAT5 at the respective PIM1 and CISH cis-regulatory elements (Figure 2F). Notably, levels of H3Y41 and STAT5 were unaltered at the promoter of the β2M housekeeping gene (Figure 2F). These data provide further evidence for a communication between JAK2 and STAT5, which extends to the chromatin interface in order to regulate transcription of these genes.
To address the possibility that STAT5 plays a direct role in recruiting JAK2 to phosphorylate H3Y41 at a subset of genomic loci, we initially performed small interfering RNA (siRNA)-mediated knockdown of STAT5 in HEL cells and monitored both transcriptional activity and H3Y41ph/STAT5 binding at known JAK2/STAT5 target genes. However, despite a 90% knockdown of STAT5, transcription and H3Y41ph levels were only modestly decreased at some (e.g., CISH) but not all (e.g., PIM1) JAK2/STAT5 target genes (Figures S1F–S1H). However, since trace amounts of STAT5 were sufficient for essentially normal development and function of the original STAT5 knockout mice (Teglund et al., 1998), the absence of a strong knock down phenotype observed here most likely reflects the fact that the residual 10% of STAT5 is sufficient for near-full activity. An alternative nonmutually exclusive explanation is that other STAT family members may compensate for the reduced levels of STAT5, as previously reported (Lim and Cao, 2006).
Due to the limitations of the siRNA approach, we chose to investigate the role of STAT5 and JAK2 in mediating H3Y41ph using an alternative strategy. Our method employed a chromosomally integrated Gal4 reporter assay in 293T cells. This approach allowed us to specifically address the issue of whether STAT5 and JAK2 are both required for H3Y41ph at a synthetically engineered chromatinized locus (Figure 2G). In this assay, STAT5 is exogenously expressed in-frame with a Gal4 DNA binding domain (DBD). By using ChIP assays to monitor H3Y41ph levels at the chromatinized Gal4 reporter, we clearly demonstrate that both STAT5 (targeted to chromatin via the Gal4 DBD) and JAK2 are necessary to phosphorylate H3Y41 at this site (Figures 2G, S1I, and S1J). Importantly, GAL4-STAT5 by itself does not lead to an increase in H3Y41ph, as 293T cells contain little, if any JAK2 (Figure 2G). In this assay, we tested human WT STAT5A and STAT5B as well as constitutively active murine STAT5A and STAT5B and observed similar results (Figures 2G, S1G, and S1H). These mechanistic studies are consistent with our genome-wide correlation and together suggest that STAT5 may recruit JAK2 to certain genomic loci to phosphorylate H3Y41.
Interrogation of the genomic distribution of H3Y41ph revealed a third pattern unique to this histone modification, i.e., “blanketing” of H3Y41ph over the entire coding region of a gene (Figures 3A–3C and S3A). Notably, H3Y41ph invariably also spans the TSS at these genes (Figure 3B). In order to identify genes most prominently blanketed by H3Y41ph, we analyzed the read density and distribution of H3Y41ph over the entire coding region of all annotated genes. This shows that H3Y41ph heavily blankets only a small fraction of genes (Figures 3A and S3A), many of which are key hematopoietic genes. Indeed, 71% of the identified blanketed genes are enriched in hematopoietic gene sets, and this group includes several genes, such as GATA2, EGFL7, and TAL1 (Figure 3B), that have previously been implicated in both normal hematopoiesis and the pathogenesis of acute myeloid leukemia (Table S1). To control for copy number alterations, which could lead to overrepresentation in sequencing data, giving a spurious impression that regions of the genome were “blanketed,” we performed an immunoglobulin G (IgG) control from the same chromatin preparation as the H3Y41ph results. When the IgG ChIP-seq data are compared with published genome-wide SNP6 array data in HEL cells, we clearly observe that the “background” of sequencing signal varies, as expected with the copy number (Figure S3B). In contrast, when we look at a locus that is blanketed by H3Y41ph, the IgG trace over the blanketed region does not mirror the pattern of H3Y41ph (Figure S3B). Similar results were also observed for the STAT5 ChIP-seq data (data not shown). Furthermore, a similar “blanketing” analysis for another active histone modification, H3K4me3, does not recapitulate these findings (Table S1).
Interestingly, the H3Y41ph-blanketed genes are not the most highly expressed genes in HEL cells (Figure S3C). Thus, the relatively high level of H3Y41ph spread throughout these genes does not seem to be a consequence of a high rate of transcription. Finally, the vast majority of heavily H3Y41ph-blanketed genes that are responsive to JAK2 inhibition do not contain STAT5 binding events (Table S2). These data serve to demonstrate that, while STAT5 and potentially other STATs may target JAK2 to certain genomic loci, this phenomenon is not ubiquitous, and the majority of genes appear to be marked by H3Y41ph in a STAT-independent manner. Nevertheless, it is clear that JAK2 inhibition is sufficient to substantially reduce H3Y41ph and gene expression, even at these STAT-independent loci (Figure 3C; Table S2).
Together these data underline the fact that blanketing is neither a simple attribute of transcriptional activity nor a common feature of other active histone modifications. This raises the possibility that H3Y41ph may “blanket” genes whose expression is necessary to propagate or maintain a particular cellular state. Consistent with this notion is the fact that unbiased interrogation of the H3Y41ph-blanketed genes in HEL cells for common knock-out mouse phenotypes using the Genomic Regions Enrichment of Annotations Tool (GREAT) (McLean et al., 2010) revealed that these genes are critical for myelopoiesis, especially erythropoiesis and megakaryopoiesis (Figure 3E). To further test this hypothesis, we analyzed a published data set of H3Y41ph in a lymphoid cell line, K1106 (Rui et al., 2010). Interestingly, the genes blanketed by H3Y41ph in this data set were identified by the GREAT database to be germane to B-lymphopoiesis (Figure 3E).
Finally, when we intersected the two data sets, we found that the heavily blanketed gene sets in the two cell types show very little overlap (Figure 3D). These data suggest that H3Y41ph may uniquely delineate genes that define a particular cell lineage. Consistent with this concept is the fact that another independent gene set enrichment analysis of the most highly blanketed genes in HEL cells (using the Database for Annotation, Visualization and Integrated Discovery: DAVID) (Huang da et al., 2009) shows that these genes are primarily expressed in myeloid cells (Figure S3D), and in contrast, the genes most blanketed by H3Y41ph in K1106 cells are those primarily expressed in lymphoid tissues (Figure S3D).
Two decades of research have firmly established that a key function of cytoplasmic JAK2 is to mediate cytokine-induced intracellular signaling. The data presented here challenge the view that JAK2 is functionally restricted to the cytoplasm in this signaling pathway. They indicate that JAK2 can phosphorylate chromatin at H3Y41 and that this modification takes place on a number of genes previously linked to the JAK/STAT pathway. Unfortunately our attempts to localize JAK2 at chromatin with ChIP assays have been unsuccessful. This most likely reflects the transient nature of JAK2′s interaction with chromatin, which is in contrast to the MAPK family of enzymes that have a consensus DNA binding motif and have been successfully localized at chromatin (Hu et al., 2009). Nevertheless, we have demonstrated in various cell types that H3Y41ph is a reliable marker for JAK2′s activity at chromatin (Dawson et al., 2009). The data reported here show that H3Y41ph is coincident with STAT5 at several loci and that both these events are dynamically modulated at the regulatory elements of cytokine-activated genes. Together, these data establish a previously unrecognized nuclear component of the JAK/STAT signaling pathway.
Coincidence of H3Y41ph and STAT5 (Table S1) is seen only at a subset of the total number of genes marked by H3Y41ph or STAT5. Since STAT5 is activated by multiple pathways (Lim and Cao, 2006), genes bound by STAT5 alone may represent interplay with alternative kinases, modifying distinct sites on histones. Alternatively, STAT function at certain sites may not require the cooperation of a kinase, but may depend, for example, upon the recruitment of specific coactivator/repressor complexes. Similarly, the genes marked by H3Y41ph alone may represent genes regulated by other STAT family members or by a STAT-independent pathway (Dawson et al., 2009). In fact, when we cross-referenced our data with the published STAT1 and STAT2 ChIP-seq data sets performed in K562 cells (another erythroid leukemia cell line), we found that a further 10% of the H3Y41ph sites not bound by STAT5 are bound by STAT1/2 (data not shown). A further possible explanation for H3Y41ph peaks that do not overlap STAT5 binding sites is that H3Y41ph is laid down by at least one other kinase (Dawson et al., 2009; Griffiths et al., 2011), which may function independently of STATs.
The data presented here have implications beyond the JAK/STAT pathway; global analysis of H3Y41ph suggests that this modification behaves differently to other well-characterized histone modifications, such as lysine acetylation and methylation, which are often linked to active genes, but their global levels are not regulated by any single signaling pathway. In contrast, H3Y41ph marks a specific set of genes stimulated by a specific signaling pathway and blankets a set of key lineage-specific hematopoietic genes. Moreover, the data presented here provide further evidence that signal-transducing kinases can extend their activity to chromatin. Since many of these pathways have been the focus of intense drug discovery programs, the identification of a nuclear component of kinase cascades increases the scope for future therapeutic intervention.
HEL, TF-1, and SET-2 cells were grown at 37°C and 5% CO2 in RPMI-1640 (Sigma-Aldrich) supplemented with 10%–20% fetal calf serum, 1% penicillin/streptomycin, and 10 ng/ml IL3 (TF-1 only).
Chromatin immunoprecipitation was performed as previously described (Dawson et al., 2009). Immunoprecipitated DNA was either amplified for sequencing or analyzed on an ABI7900 RT-PCR machine using TaqMan PCR mastermix according to the manufacturer’s instructions (see Extended Experimental Procedures for primers/probes).
Messenger RNA (mRNA) was prepared using the QIAGEN RNeasy kit according to the manufacturer’s instructions. Complementary DNA (cDNA) was prepared using Superscript III reverse transcriptase (Invitrogen) and analyzed on an ABI7900 RT-PCR machine using power SYBR green PCR mastermix according to the manufacturer’s instructions (see Extended Experimental Procedures for primers).
TG101209, TargeGen Inc. (San Diego, CA, USA) and AT9283, Astex Pharmaceuticals, were used at 1 μM and 300 nM, respectively, for 4 hr.
BED files were created from uniquely mapped reads extended to 200 bp and used to create genome-wide density profiles as described (Wilson et al., 2009). For genome-wide feature analysis, regions of 10 kb width were extracted and the mean density plotted normalized relative to the total number of sequence reads in the corresponding data sets. Heat maps were plotted using the same data and ranked based on the ratio of signal at the central 2 kb divided by the average signal across the 10 kb region. To examine global relationships between H3K4me3 and H3Y41ph, TSSs were binned into 10 cohorts of descending H3K4me3 levels. For quantile normalization, TSSs were ranked relative to the quantity of H3K4me3 and H3Y41ph sequence tags within ±500 bp of the TSS and displayed using the kde2d R function for kernel density plots (Venables and Ripley, 2002).
Refseq genes were split into 10 equally sized regions and two measures were calculated for each: the proportion of bases showing significant normalized H3Y41ph sequence tag score >10 (“overlap”) and sum of normalized H3Y41ph read counts per base (“density”). Both measures were highly correlated (Spearman correlation = 1.00, p < 2 × 10−16). Genes within 50 kb of the centromere or less than 2.5 kb were removed. Of the 19,067 unique genes analyzed, 1,009 showed H3Y41ph coverage. A histogram of the median H3Y41ph overlap of each gene identified a subset of genes with high levels of protracted gene coverage. Genes were classified as “blanketed” or “highly blanketed” for H3Y41ph if the overlap was greater than 20% (324 genes) or 50% (45 genes), respectively. Three control gene sets were created of a similar size to the blanketed gene list: active genes (based on H3K4me3 read density at the promoter), active genes with no H3Y41ph, and genes with H3Y41ph in the promoter but not across the gene body. Functional gene set analysis was conducted using the DAVID tool (Huang da et al., 2009).
Peak calling for the STAT5 and H3Y41ph data sets was performed using Findpeaks 3.1 (Fejes et al., 2008) and MACs (Zhang et al., 2008). In addition, STAT5 peaks were called using PeakSeq (Rozowsky et al., 2009). Peaks called by all programs were retained and filtered further based on Findpeaks 3.1 peaks in the control IgG ChIP-seq data set, distance of more than 100 kb of any RefSeq or University of California Santa Cruz (UCSC) annotated gene, and more than 70% of repetitive sequence. De novo motif discovery was performed using Multiple Expectation Maximization for Motif Elicitation (MEME) with the central 100 bp of STAT5 peaks (Bailey et al., 2009) with default settings.
Overlap between STAT5- and H3Y41ph-enriched regions was determined based on peaks that overlapped by at least one base pair and also determined for STAT1/STAT2 peaks in K562 cells from the ENCODE project (Zhang et al., 2007). Statistical significance of overlaps was calculated by performing 1,000 random permutations and a nonparametric Wilcoxon rank-sum test under the NULL hypothesis that STAT and H3Y41ph enrichments are independent of one another.
Extended Experimental Procedures
Cell Culture and siRNA/Plasmid Transfection
HEL cells were grown in RPMI-1640 medium (Sigma-Aldrich) supplemented with 10% fetal calf serum and 1% penicillin/streptomycin. TF-1 cells were grown in RPMI-1640 medium (Sigma-Aldrich) supplemented with 10% fetal calf serum and 1% penicillin/streptomycin and 10 ng/ml of IL3. SET-2 cells were grown in RPMI-1640 medium (Sigma-Aldrich) supplemented with 20% fetal calf serum and 1% penicillin/streptomycin. Cells were incubated at 37°C and 5% CO2. Cytokine stimulation was performed in TF-1 cells following 18 hr of serum starvation. Interleukin-3 (10 ng/ml) was used in isolation to stimulate the cells for up to two hours. Short interfering RNA (siRNA) duplexes targeting human STAT5A (#S100048426; #S100048433) and STAT5B (#S100100401; #S100100408) from QIAGEN and non-targeting siRNA duplexes from Dharmacon (non-targeting siRNA pool #D-001206-13-20) were used according to the manufacturers instructions. The siRNA duplexes were transfected by electroporation, according to the Nucleofection protocol of the manufacturer’s program #X-005 (for Amaxa® Nucleofector® Kit V, Amaxa Inc. Scientific Support, Walkersville, MD, USA). The cells were transfected and incubated for 48 hr. 293T cells with an integrated GAL4 reporter were grown in DMEM medium (Sigma-Aldrich) supplemented with 10% fetal calf serum, 2 ug/ml puromycin and 1% penicillin/ streptomycin. Plasmids in which STAT5 was cloned in-frame with a Gal4 DNA binding domain (DBD) were transfected with and without a separate plasmid containing JAK2 V617F using Lifofectamine 2000 (Invitrogen) according to the manufacturer’s instructions. Gal4 fusions of human wild-type STAT5A and STAT5B as well as murine constitutively active STAT5A∗ and STAT5B∗ were used. The total amount of transfected DNA was kept constant by co-transfecting the appropriate empty vector as necessary.
Chromatin Immunoprecipitation Assay and Real-Time PCR Analysis
Chromatin immunoprecipitation was performed as previously described (Dawson et al., 2009). Immunoprecipitated DNA was either amplified for massive parallel sequencing or analyzed on an ABI 7900 real-time PCR machine, using TaqMan® PCR mastermix according to the manufacturer’s instructions. The following primers and probes were used in the analysis:
Gene Expression Analysis
mRNA was prepared from cell extracts using the QIAGEN RNeasy kit according to the manufacturer’s instructions. cDNA was then prepared using Superscript III reverse transcriptase (Invitrogen) and analyzed on an ABI 7900 real-time PCR machine, using power SYBR®green PCR mastermix according to the manufacturer’s instructions. The following primer pairs were then used in the analysis:
TG101209, TargeGen Inc. (San Diego, CA, USA), is commercially available and was used at 1 μM for four hours in vivo as previously described. AT9283, Astex Pharmaceuticals, was used at 300nM for four hours.
Antibodies and Immunoblotting
The antibodies used in the manuscript were anti-H3Y41ph (ab26127, Abcam); anti-STAT5 (C-17:sc-835 Santa Cruz) - this antibody detects both STAT5A and STAT5B; anti-H3 (ab1791, Abcam); anti-GAPDH (ab9483, Abcam); Rabbit polyclonal IgG (ab27472, Abcam); anti-GAL4 (ab1396, Abcam). Whole cell lysates were prepared, resolved by SDS-PAGE, transferred to nitrocellulose and probed with relevant antibodies. Membranes were then analyzed using the Odyssey Imaging System (Li-COR Biosciences) according to the manufacturer’s instructions.
Sample Preparation for ChIP-seq
ChIP libraries used for massive parallel sequencing were prepared using the Illumina ChIP-seq kit with the following adjustments. Following chromatin immunoprecipitation and DNA purification, 10-50ng of DNA was used to perform end repair of the ChIP DNA. The following reaction mix was used, DNA (10-50ng resuspended in 30 μl), DNase and RNase free ddH2O 10 μl, T4 DNA ligase buffer with 10 mM ATP 5 μl, 10 mM dNTP mix 2 μl, T4 DNA polymerase 1 μl, Klenow DNA polymerase 1 μl and T4 PNK 1 μl. The sample was incubated for 30 min at 20°C and then purified using the Zymo DNA Clean & Concentrator-5™ kit with the DNA eluted in 11 μl of ddH2O. Following end repair adenine bases were added to the 3′ end of the DNA fragments using DNA from above 11 μl, ddH2O 33 μl, Klenow buffer 5 μl, dATP 10 μl and Klenow (3′ to 5′ exonuclease deficient) 1 μl. The sample was incubated for 30 min at 37°C and the DNA was purified using the Zymo kit. Commercially supplied adapters (Illumina; the sequence is commercially protected) were then ligated to the DNA fragments using DNA ligase buffer 14 μl, adaptor oligonucleotide mix 2 μl, DNA ligase 4 μl and DNA 10 μl. The reaction was incubated for 15 min at room temperature and DNA purified using the Zymo kit. Following the ligation of adapters the DNA was amplified by PCR. The reaction conditions included DNA 10 μl, ddH2O 26 μl, 5X Phusion buffer 10 μl, 10 mM dNTP mix 1.5 μl, Phusion polymerase 0.5 μl and 1 μl of both forward and reverse PCR primers the sequence of which is commercially protected. The reaction conditions are as follows: 98°C, 30 s; 98°C, 10 s; 65°C, 30 s; 72°C, 30 s; back to 2 (18 cycles); 72°C for 5 min.
Following PCR amplification, the DNA was purified using the MinElute PCR purification kit (QIAGEN). The clean DNA was then size fractionated in a 1.5% agarose gel. Each sample was run in isolation in the gel in a dedicated clean agarose tank in 1 X TAE buffer. The gel was subsequently stained with SybrGreen in a clean staining tank for 1 hr. Following this, 200 – 400 bp fragments were excised from the DNA smear present on the gel, and purified using the MinElute QIAGEN Gel-Purification Kit. The completed library was then tested for both size selection and quantity of DNA using a 2100 Agilent Bioanalyser. The amplified DNA was used for cluster generation on the Illumina genome analyzer as previously described (Barski et al., 2007).
Density Profiling and Genome-Wide Feature Analysis
Sequence data were processed and only reads that could be uniquely mapped to the genome were retained. A BED file was created where each sequence tag was extended from its start coordinate to be a length of 200bp, the mean length of a ChIP’d DNA fragment. This BED file was then used to create a genome-wide density profile as previously described (Wilson et al., 2009). This density profile was then used for i) uploading into the UCSC genome browser for visual analysis of the data and ii) for genome-wide feature analysis and iii) individual heatmap visualization. For ii, regions of 10kb width were extracted from the density profile and the mean density of the population was plotted. The data were normalized relative to the total number of sequence reads in the corresponding data sets. Heat maps were plotted using the same extracted data, and peak regions were ranked as follows: read density was calculated from −5kb to +5kb either side of the peak center, Next, the ratio of signal at the central 2kb divided by the average signal across the 10 kb region was determined, and all peaks were ranked from the highest to the lowest ratio. This order was then used to display the heatmaps of H3Y41ph +/− inhibitor. The same procedure was used to generate the analogous heatmaps for the STAT5 peaks.
Mean Enrichment Pattern Across Transcriptional Start Sites
To examine the precise global relationship between H3K4me3 and H3Y41ph at transcriptional start sites (TSS) a mean enrichment pattern for these two histone modifications across populations of TSS were plotted. We ranked TSSs based on their H3K4me3 levels and profiled the distribution of H3K4me3 for all promoters, binned into 10 cohorts of descending H3K4me3 leves. We then determined the spatial distribution of H3Y41ph by surveying its location across the same 10 promoter cohorts.
Identification of Blanketed Genes
We defined “blanketed” genes as those loci showing high coverage of H3Y41ph across the entire gene. Transcription start and end coordinates for 34,328 genes (21,520 unique) were taken from the RefSeq track on the UCSC browser. To identify those genes where H3Y41ph coverage was significant across the entire gene rather than restricted to a single area such as the promoter region, genes were split into 10 equally sized sections. For each of these sections, two measurements of H3Y41ph coverage were calculated and the median over the 10 sections was taken for each gene. The first measurement was the proportion of bases within the gene section that also showed a significant level of signal for H3Y41ph (normalized sequence tag score > 10). The second measurement was the density of H3Y41ph reads per base, calculated by summing the normalized read tag scores for each base in the gene section and dividing by the number of bases. These measures were highly correlated (Spearman correlation = 0.9998723, p < 2 × 10−16), but the gene overlap provided a more intuitive measure of H3Y41ph coverage.
Genes showing H3K4me3 at the promoter regions were identified by calculating the read density per base as described above, but this time looking at a single region within a 20 kb window around the transcription start site.
Functional Annotation of Blanketed Genes
We removed all genes within 50kb of a centromere, as previous studies have identified these regions as containing artificially high numbers of reads using the Illumina sequencing platform. Genes less than 2.5 kb were removed to avoid the spurious identification of smaller blanketed genes. Of the 19,067 unique genes analyzed, 1,009 showed some level of H3Y41ph coverage. A histogram of the median H3Y41ph overlap of each gene identified a subset of genes with high levels of protracted gene coverage (Figure S3A). We used a threshold of 20% to identify genes showing potentially blanketed levels of H3Y41ph, and a threshold of 50% to identify those genes with clearly blanketed levels of H3Y41ph, reasoning that greater than half coverage of the gene by significant levels of H3Y41ph when spread over the entire gene body indicated a clearly blanketed profile (Figure S3A). Using these thresholds we identified 324 potentially blanketed genes and 45 clearly blanketed genes.
Genes were also ranked on the read density of H3K4me3 at the promoter, and 5 gene sets of an identical size to that of the blanketed set were selected from the top of this list to use as controls. We also selected a similar list, but with no signs of Y41ph over the gene region. Our final set of control genes was selected from a ranked list of genes with H3Y41ph overlap near the promoter, but not across the rest of the gene.
We applied a functional annotation analysis to these blanketed genes and the three control sets using the DAVID Functional Annotation Tool (Dennis et al., 2003; Huang da et al., 2009). This uses a modified Fisher exact test to test for enrichment of terms from a number of annotation databases, including gene ontology, tissue expression and pathways.
Initial peak calling for the STAT5 and H3Y41ph ChIP-seq data sets was performed using Findpeaks 3.1 (Fejes et al., 2008) and MACs (Zhang et al., 2008). In addition STAT5 peaks were called using PeakSeq (Rozowsky et al., 2009). Only regions called as peaks by all programs used were retained for further filtering. Any regions that were also called as peaks by Findpeaks 3.1 within the control IgG ChIP-seq data set were removed. Of the remaining peaks only those that were located within 100 kb of a RefSeq and a UCSC gene were retained. Finally, any peaks that contained more than 70% of repetitive sequence were removed. This then provided a final list of high confidence regions of STAT5 and H3Y41ph enrichment.
Quantile Normalization of H3K4me3 and H3Y41ph at Transcription Start Sites
The transcription start sites were ranked relative to the quantity of sequence tags, whole/partial, located within ± 500 bp of the start site. Each transcription start site then had rankings determined for both H3K4me3 and H3Y41ph which could then be used for a two-dimensional kernel density plot using the kde2d R function (Venables and Ripley, 2002) Modern Applied Statistics with S. Fourth edition. Springer).
De novo motif discovery was performed with the program MEME (Bailey et al., 2009) using default settings. De novo motif discovery using the central 100bp of the high confidence STAT5 peaks were used to provide a single highly significant motif of 10 bp in length. This was then displayed as a sequence logo.
Analysis of Overlap between H3Y41ph, STAT1, STAT2 and STAT5
To categorize the level of overlap between STAT5 binding sites and H3Y41ph-enriched loci, we identified the number of peaks called in the STAT5 data set which share at least 1 base pair overlap with those called in the H3Y41ph data set. We also included in this analysis STAT1- and STAT2-enriched peaks called as part of the ENCODE project in K562 cells (Zhang et al., 2007). To measure the statistical significance of these overlapping regions, we calculated the level of overlap between the various STAT peak regions (within 100 kb of annotated UCSC/RefSeq genes) and a random permutation of the H3Y41ph peaks (within 100 kb of annotated UCSC/RefSeq genes). A NULL distribution for each STAT protein was generated by performing 1,000 permutations, and a non-parametric Wilcoxon rank-sum test was used to determine the probability of observing an overlap at least as large as that observed in our data set, under the NULL hypothesis that STAT protein and H3Y41ph binding events are independent of one another.
|Bailey T.L.,Boden M.,Buske F.A.,Frith M.,Grant C.E.,Clementi L.,Ren J.,Li W.W.,Noble W.S.. MEME SUITE: tools for motif discovery and searchingNucleic Acids Res.37Web Server issueYear: 2009W202W20819458158|
|Barahmand-Pour F.,Meinke A.,Groner B.,Decker T.. Jak2-Stat5 interactions analyzed in yeastJ. Biol. Chem.273Year: 199812567125759575217|
|Barski A.,Cuddapah S.,Cui K.,Roh T.Y.,Schones D.E.,Wang Z.,Wei G.,Chepelev I.,Zhao K.. High-resolution profiling of histone methylations in the human genomeCell129Year: 200782383717512414|
|Chen E.,Staudt L.M.,Green A.R.. Janus kinase deregulation in leukemia and lymphomaImmunity36Year: 201252954122520846|
|Dawson M.A.,Bannister A.J.,Göttgens B.,Foster S.D.,Bartke T.,Green A.R.,Kouzarides T.. JAK2 phosphorylates histone H3Y41 and excludes HP1alpha from chromatinNature461Year: 200981982219783980|
|Dawson M.A.,Curry J.E.,Barber K.,Beer P.A.,Graham B.,Lyons J.F.,Richardson C.J.,Scott M.A.,Smyth T.,Squires M.S.. AT9283, a potent inhibitor of the Aurora kinases and Jak2, has therapeutic potential in myeloproliferative disordersBr. J. Haematol.150Year: 2010465720507304|
|Decker T.,Müller M.. Jak-stat signaling: from basics to diseaseFirst EditionYear: 2012SpringerNew York|
|Dennis G. Jr.,Sherman B.T.,Hosack D.A.,Yang J.,Gao W.,Lane H.C.,Lempicki R.A.. DAVID: Database for Annotation, Visualization, and Integrated DiscoveryGenome Biol.4Year: 20033|
|Dumon S.,Santos S.C.,Debierre-Grockiego F.,Gouilleux-Gruart V.,Cocault L.,Boucheron C.,Mollat P.,Gisselbrecht S.,Gouilleux F.. IL-3 dependent regulation of Bcl-xL gene expression by STAT5 in a bone marrow derived cell lineOncogene18Year: 19994191419910435632|
|Fejes A.P.,Robertson G.,Bilenky M.,Varhol R.,Bainbridge M.,Jones S.J.. FindPeaks 3.1: a tool for identifying areas of enrichment from massively parallel short-read sequencing technologyBioinformatics24Year: 20081729173018599518|
|Griffiths D.S.,Li J.,Dawson M.A.,Trotter M.W.,Cheng Y.H.,Smith A.M.,Mansfield W.,Liu P.,Kouzarides T.,Nichols J.. LIF-independent JAK signalling to chromatin in embryonic stem cells uncovered from an adult stem cell diseaseNat. Cell Biol.13Year: 2011132121151131|
|Hu S.,Xie Z.,Onishi A.,Yu X.,Jiang L.,Lin J.,Rho H.S.,Woodard C.,Wang H.,Jeong J.S.. Profiling the human protein-DNA interactome reveals ERK2 as a transcriptional repressor of interferon signalingCell139Year: 200961062219879846|
|Huang da W.,Sherman B.T.,Lempicki R.A.. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resourcesNat. Protoc.4Year: 2009445719131956|
|Kamakura S.,Oishi K.,Yoshimatsu T.,Nakafuku M.,Masuyama N.,Gotoh Y.. Hes binding to STAT3 mediates crosstalk between Notch and JAK-STAT signallingNat. Cell Biol.6Year: 200454755415156153|
|Lefrancois-Martinez A.M.,Blondet-Trichard A.,Binart N.,Val P.,Chambon C.,Sahut-Barnola I.,Pointud J.C.,Martinez A.. Transcriptional control of adrenal steroidogenesis: novel connection between Janus kinase (JAK) 2 protein and protein kinase A (PKA) through stabilization of cAMP response element-binding protein (CREB) transcription factorJ. Biol. Chem.286Year: 2011329763298521808064|
|Lim C.P.,Cao X.. Structure, function, and regulation of STAT proteinsMol. Biosyst.2Year: 200653655017216035|
|Liu F.,Zhao X.,Perna F.,Wang L.,Koppikar P.,Abdel-Wahab O.,Harr M.W.,Levine R.L.,Xu H.,Tefferi A.. JAK2V617F-mediated phosphorylation of PRMT5 downregulates its methyltransferase activity and promotes myeloproliferationCancer Cell19Year: 201128329421316606|
|Matsumoto A.,Masuhara M.,Mitsui K.,Yokouchi M.,Ohtsubo M.,Misawa H.,Miyajima A.,Yoshimura A.. CIS, a cytokine inducible SH2 protein, is a target of the JAK-STAT5 pathway and modulates STAT5 activationBlood89Year: 1997314831549129017|
|McLean C.Y.,Bristor D.,Hiller M.,Clarke S.L.,Schaar B.T.,Lowe C.B.,Wenger A.M.,Bejerano G.. GREAT improves functional interpretation of cis-regulatory regionsNat. Biotechnol.28Year: 201049550120436461|
|Min I.M.,Waterfall J.J.,Core L.J.,Munroe R.J.,Schimenti J.,Lis J.T.. Regulating RNA polymerase pausing and transcription elongation in embryonic stem cellsGenes Dev.25Year: 201174275421460038|
|Mui A.L.,Wakao H.,Kinoshita T.,Kitamura T.,Miyajima A.. Suppression of interleukin-3-induced gene expression by a C-terminal truncated Stat5: role of Stat5 in proliferationEMBO J.15Year: 1996242524338665850|
|Nagy Z.S.,LeBaron M.J.,Ross J.A.,Mitra A.,Rui H.,Kirken R.A.. STAT5 regulation of BCL10 parallels constitutive NFkappaB activation in lymphoid tumor cellsMol. Cancer8Year: 20096719709433|
|Nilsson J.,Bjursell G.,Kannius-Janson M.. Nuclear Jak2 and transcription factor NF1-C2: a novel mechanism of prolactin signaling in mammary epithelial cellsMol. Cell. Biol.26Year: 20065663567416847321|
|Rozowsky J.,Euskirchen G.,Auerbach R.K.,Zhang Z.D.,Gibson T.,Bjornson R.,Carriero N.,Snyder M.,Gerstein M.B.. PeakSeq enables systematic scoring of ChIP-seq experiments relative to controlsNat. Biotechnol.27Year: 2009667519122651|
|Rui L.,Emre N.C.,Kruhlak M.J.,Chung H.J.,Steidl C.,Slack G.,Wright G.W.,Lenz G.,Ngo V.N.,Shaffer A.L.. Cooperative epigenetic modulation by cancer amplicon genesCancer Cell18Year: 201059060521156283|
|Teglund S.,McKay C.,Schuetz E.,van Deursen J.M.,Stravopodis D.,Wang D.,Brown M.,Bodner S.,Grosveld G.,Ihle J.N.. Stat5a and Stat5b proteins have essential and nonessential, or redundant, roles in cytokine responsesCell93Year: 19988418509630227|
|Venables W.N.,Ripley B.D.. Modern applied statistics with SFourth EditionYear: 2002SpringerNew York|
|Wilson N.K.,Miranda-Saavedra D.,Kinston S.,Bonadies N.,Foster S.D.,Calero-Nieto F.,Dawson M.A.,Donaldson I.J.,Dumon S.,Frampton J.. The transcriptional program controlled by the stem cell leukemia gene Scl/Tal1 during early embryonic hematopoietic developmentBlood113Year: 20095456546519346495|
|Wood A.D.,Chen E.,Donaldson I.J.,Hattangadi S.,Burke K.A.,Dawson M.A.,Miranda-Saavedra D.,Lodish H.F.,Green A.R.,Göttgens B.. ID1 promotes expansion and survival of primary erythroid cells and is a target of JAK2V617F-STAT5 signalingBlood114Year: 20091820183019571317|
|Zhang Y.,Liu T.,Meyer C.A.,Eeckhoute J.,Johnson D.S.,Bernstein B.E.,Nusbaum C.,Myers R.M.,Brown M.,Li W.,Liu X.S.. Model-based analysis of ChIP-Seq (MACS)Genome Biol.9Year: 2008R13718798982|
|Zhang Y.,Laz E.V.,Waxman D.J.. Dynamic, sex-differential STAT5 and BCL6 binding to sex-biased, growth hormone-regulated genes in adult mouse liverMol. Cell. Biol.32Year: 201288089622158971|
|Zhang Z.D.,Paccanaro A.,Fu Y.,Weissman S.,Weng Z.,Chang J.,Snyder M.,Gerstein M.B.. Statistical analysis of the genomic distribution and correlation of regulatory elements in the ENCODE regionsGenome Res.17Year: 200778779717567997|
|Bailey, T.L., Boden, M., Buske, F.A., Frith, M., Grant, C.E., Clementi, L., Ren, J., Li, W.W., and Noble, W.S. (2009). MEME SUITE: tools for motif discovery and searching. Nucleic Acids Res. 37 (Web Server issue), W202-8.|
|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.|
|Dawson, M.A., Bannister, A.J., Göttgens, B., Foster, S.D., Bartke, T., Green, A.R., and Kouzarides, T. (2009). JAK2 phosphorylates histone H3Y41 and excludes HP1alpha from chromatin. Nature 461, 819–822.|
|Dennis, G., Jr., Sherman, B.T., Hosack, D.A., Yang, J., Gao, W., Lane, H.C., and Lempicki, R.A. (2003). DAVID: Database for Annotation, Visualization, and Integrated Discovery. Genome Biol. 4, 3.|
|Fejes, A.P., Robertson, G., Bilenky, M., Varhol, R., Bainbridge, M., and Jones, S.J. (2008). FindPeaks 3.1: a tool for identifying areas of enrichment from massively parallel short-read sequencing technology. Bioinformatics 24, 1729–1730.|
|Huang, W., Sherman, B.T., and Lempicki, R.A. (2009). Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat. Protoc. 4, 44–57.|
|Rozowsky, J., Euskirchen, G., Auerbach, R.K., Zhang, Z.D., Gibson, T., Bjornson, R., Carriero, N., Snyder, M., and Gerstein, M.B. (2009). PeakSeq enables systematic scoring of ChIP-seq experiments relative to controls. Nat. Biotechnol. 27, 66–75.|
|Wilson, N.K., Miranda-Saavedra, D., Kinston, S., Bonadies, N., Foster, S.D., Calero-Nieto, F., Dawson, M.A., Donaldson, I.J., Dumon, S., Frampton, J., et al. (2009). The transcriptional program controlled by the stem cell leukemia gene Scl/Tal1 during early embryonic hematopoietic development. Blood 113, 5456–5465.|
|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). Model-based analysis of ChIP-Seq (MACS). Genome Biol. 9, R137.|
|Zhang, Z.D., Paccanaro, A., Fu, Y., Weissman, S., Weng, Z., Chang, J., Snyder, M., and Gerstein, M.B. (2007). Statistical analysis of the genomic distribution and correlation of regulatory elements in the ENCODE regions. Genome Res. 17, 787–797.|
The ArrayExpress accession number for the ChIP-seq data reported in this paper is E-MTAB-1096.
Table S1. Genes Marked with H3Y41ph/STAT5 and Blanketing Analysis with H3Y41ph and H3K4me3, Related to Figures 1, 2, and 3
This excel table contains a list of all the genes marked by H3Y41ph, STAT5, and both together in HEL cells. It also contains a list of the top H3Y41ph blanketed genes and control analyses performed with H3K4me3 to discern if “blanketing” pattern observed with H3Y41ph was also seen with H3K4me3. This excel table lists the top 100 genes marked by H3Y41ph and whose expression is significantly downregulated following JAK2 inhibition. Each gene is also annotated for STAT5 binding within 50 Kb of the JAK2 responsive gene.
Click here for additional data file (mmc1.xls)
Table S2. Relationship between Genes Blanketed with H3Y41ph and STAT5 Binding, Related to Figure 3
Click here for additional data file (mmc2.xls)
Document S1. Article plus Supplemental Information
This excel table contains a list of all the genes marked by H3Y41ph, STAT5, and both together in HEL cells. It also contains a list of the top H3Y41ph blanketed genes and control analyses performed with H3K4me3 to discern if “blanketing” pattern observed with H3Y41ph was also seen with H3K4me3.
This excel table lists the top 100 genes marked by H3Y41ph and whose expression is significantly downregulated following JAK2 inhibition. Each gene is also annotated for STAT5 binding within 50 Kb of the JAK2 responsive gene.
We would like to thank Yongjun Zhao for overseeing the sequencing and Sarah Teichmann for helpful discussions. Work in the authors’ laboratories is supported by grants from Cancer Research UK, Leukaemia & Lymphoma Research UK, the UK Medical Research Council, the Wellcome Trust, Leukemia & Lymphoma Society of America, NIHR Cambridge Biomedical Research Centre, and the 6th Research Framework Programme of the European Union (Epitron and SMARTER). M.A.D. is a Wellcome-Beit Intermediate Clinical Fellow. S.D.F. was funded by a PhD fellowship grant from the UK Medical Research Council to the Cambridge Institute for Medical Research. T.K. is a director of Abcam Ltd and he is on the scientific advisory board of GlaxoSmithKline. A.R.G. is on the clinical advisory board of Astex Therapeutics, Cambridge, UK.
Previous Document: Regulating contractility of the actomyosin cytoskeleton by pH.
Next Document: Memory Deficits Induced by Inflammation Are Regulated by ?5-Subunit-Containing GABA(A) Receptors.