FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2012071760> ?p ?o ?g. }

• W2012071760 endingPage "169" @default.
• W2012071760 startingPage "156" @default.
• W2012071760 abstract "We report the generation and comparative analysis of genome-wide chromatin state maps, PPARγ and CTCF localization maps, and gene expression profiles from murine and human models of adipogenesis. The data provide high-resolution views of chromatin remodeling during cellular differentiation and allow identification of thousands of putative preadipocyte- and adipocyte-specific cis-regulatory elements based on dynamic chromatin signatures. We find that the specific locations of most such elements differ between the two models, including at orthologous loci with similar expression patterns. Based on sequence analysis and reporter assays, we show that these differences are determined, in part, by evolutionary turnover of transcription factor motifs in the genome sequences and that this turnover may be facilitated by the presence of multiple distal regulatory elements at adipogenesis-dependent loci. We also utilize the close relationship between open chromatin marks and transcription factor motifs to identify and validate PLZF and SRF as regulators of adipogenesis. We report the generation and comparative analysis of genome-wide chromatin state maps, PPARγ and CTCF localization maps, and gene expression profiles from murine and human models of adipogenesis. The data provide high-resolution views of chromatin remodeling during cellular differentiation and allow identification of thousands of putative preadipocyte- and adipocyte-specific cis-regulatory elements based on dynamic chromatin signatures. We find that the specific locations of most such elements differ between the two models, including at orthologous loci with similar expression patterns. Based on sequence analysis and reporter assays, we show that these differences are determined, in part, by evolutionary turnover of transcription factor motifs in the genome sequences and that this turnover may be facilitated by the presence of multiple distal regulatory elements at adipogenesis-dependent loci. We also utilize the close relationship between open chromatin marks and transcription factor motifs to identify and validate PLZF and SRF as regulators of adipogenesis. Chromatin state maps were generated throughout murine and human adipogenesis Most active cis-regulatory elements differ between species Many differences are due to evolutionary turnover of transcription factor motifs Motif enrichment predicted that PLZF and SRF play a repressive role in adipogenesis Describing the gene regulatory networks (GRNs) that control development, differentiation, and physiological processes is a major goal of mammalian genome biology. A GRN consists of trans-regulatory factors and cis-regulatory elements whose interactions with each other and the environment govern the expression of genes in the network and ultimately manifest as a complex phenotype such as gastrulation, adipogenesis, or glucose homeostasis (Arnone and Davidson, 1997Arnone M.I. Davidson E.H. The hardwiring of development: organization and function of genomic regulatory systems.Development. 1997; 124: 1851-1864PubMed Google Scholar). The core trans-regulatory factors in a variety of GRNs have been identified by expression profiling and genetic analysis, but the large size and complex architecture of mammalian genomes have prevented systematic identification of cis-regulatory elements. Recent advances in high-throughput DNA sequencing have led to the development of new experimental tools that greatly enhance our ability to study genome function. In particular, chromatin immunoprecipitation and sequencing (ChIP-Seq) allows efficient genome-wide profiling of transcription factor (TF) localization (Johnson et al., 2007Johnson D.S. Mortazavi A. Myers R.M. Wold B. Genome-wide mapping of in vivo protein-DNA interactions.Science. 2007; 316: 1497-1502Crossref PubMed Scopus (1827) Google Scholar, Robertson et al., 2007Robertson G. Hirst M. Bainbridge M. Bilenky M. Zhao Y. Zeng T. Euskirchen G. Bernier B. Varhol R. Delaney A. et al.Genome-wide profiles of STAT1 DNA association using chromatin immunoprecipitation and massively parallel sequencing.Nat. Methods. 2007; 4: 651-657Crossref PubMed Scopus (989) Google Scholar) and chromatin state (Barski et al., 2007Barski 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 genome.Cell. 2007; 129: 823-837Abstract Full Text Full Text PDF PubMed Scopus (4697) Google Scholar, Mikkelsen et al., 2007aMikkelsen 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.Genome-wide maps of chromatin state in pluripotent and lineage-committed cells.Nature. 2007; 448: 553-560Crossref PubMed Scopus (3034) Google Scholar). Because different classes of cis-regulatory elements display characteristic chromatin signatures when they are active (Hon et al., 2009Hon G.C. Hawkins R.D. Ren B. Predictive chromatin signatures in the mammalian genome.Hum. Mol. Genet. 2009; 18: R195-R201Crossref PubMed Scopus (166) Google Scholar), ChIP-Seq has emerged as a powerful tool for comprehensive discovery of these elements. Identifying the components of a GRN that govern a specific phenotype of interest from ChIP-Seq maps of a given cell type, however, remains challenging for several reasons. First, these maps typically identify tens of thousands of putative regulatory elements, only some of which are likely to be directly relevant to the phenotype. Second, whereas these maps appear to be highly sensitive, their specificity toward biologically relevant elements is less clear (Birney et al., 2007Birney E. Stamatoyannopoulos J.A. Dutta A. Guigó R. Gingeras T.R. Margulies E.H. Weng Z. Snyder M. Dermitzakis E.T. Thurman R.E. et al.ENCODE Project ConsortiumIdentification and analysis of functional elements in 1% of the human genome by the ENCODE pilot project.Nature. 2007; 447: 799-816Crossref PubMed Scopus (3836) Google Scholar). For example, TF localization analyses frequently reveal many binding sites that have no discernable effect on the expression levels of nearby genes (Johnson et al., 2007Johnson D.S. Mortazavi A. Myers R.M. Wold B. Genome-wide mapping of in vivo protein-DNA interactions.Science. 2007; 316: 1497-1502Crossref PubMed Scopus (1827) Google Scholar, Robertson et al., 2007Robertson G. Hirst M. Bainbridge M. Bilenky M. Zhao Y. Zeng T. Euskirchen G. Bernier B. Varhol R. Delaney A. et al.Genome-wide profiles of STAT1 DNA association using chromatin immunoprecipitation and massively parallel sequencing.Nat. Methods. 2007; 4: 651-657Crossref PubMed Scopus (989) Google Scholar, Zhang et al., 2005Zhang X. Odom D.T. Koo S.H. Conkright M.D. Canettieri G. Best J. Chen H. Jenner R. Herbolsheimer E. Jacobsen E. et al.Genome-wide analysis of cAMP-response element binding protein occupancy, phosphorylation, and target gene activation in human tissues.Proc. Natl. Acad. Sci. USA. 2005; 102: 4459-4464Crossref PubMed Scopus (711) Google Scholar). Third, practical considerations often necessitate the use of in vitro cell culture models that might be subject to aberrant genetic or epigenetic changes. This raises the possibility that some chromatin state components observed in an in vitro model may not be representative of the analogous cell type in vivo (Noer et al., 2009Noer A. Lindeman L.C. Collas P. Histone H3 modifications associated with differentiation and long-term culture of mesenchymal adipose stem cells.Stem Cells Dev. 2009; 18: 725-736Crossref PubMed Scopus (84) Google Scholar). We reasoned that comparative profiling of multiple cell culture models that display similar, inducible phenotypes might help shed light on these issues. Profiling closely related cell types before and after induction should help identify regulatory elements that are directly related to the phenotype. Classification of these elements as either model-specific or shared should then provide a foundation for understanding their relative importance and therefore help prioritize in-depth functional studies. To explore this approach, we focused on adipogenesis. Adipocytes play a central role in systemic metabolism, coordinating lipid and glucose homeostasis (Rosen and Spiegelman, 2006Rosen E.D. Spiegelman B.M. Adipocytes as regulators of energy balance and glucose homeostasis.Nature. 2006; 444: 847-853Crossref PubMed Scopus (1446) Google Scholar). The burgeoning human and financial costs of obesity, type 2 diabetes, and other metabolic disorders have therefore thrust adipocyte biology into the forefront of biomedical research priorities (Camp et al., 2002Camp H.S. Ren D. Leff T. Adipogenesis and fat-cell function in obesity and diabetes.Trends Mol. Med. 2002; 8: 442-447Abstract Full Text Full Text PDF PubMed Scopus (161) Google Scholar). Adipogenesis is also one of the most intensively studied examples of cellular differentiation, and several cell culture models that appear to closely approximate events that occur during adipogenesis in vivo are available (Rosen and MacDougald, 2006Rosen E.D. MacDougald O.A. Adipocyte differentiation from the inside out.Nat. Rev. Mol. Cell Biol. 2006; 7: 885-896Crossref PubMed Scopus (1772) Google Scholar). Here, we report the generation and analysis of genome-wide chromatin state maps, TF localization maps, and gene expression profiles from multiple stages of differentiation in two established models of adipogenesis, murine 3T3-L1 cells (L1s) and human adipose stromal cells (hASCs). 3T3-L1 is a cell line originally subcloned from embryonic fibroblasts (Green and Meuth, 1974Green H. Meuth M. An established pre-adipose cell line and its differentiation in culture.Cell. 1974; 3: 127-133Abstract Full Text PDF PubMed Scopus (774) Google Scholar), and hASCs are primary cells derived from adult subcutaneous lipoaspirates (Aust et al., 2004Aust L. Devlin B. Foster S.J. Halvorsen Y.D. Hicok K. du Laney T. Sen A. Willingmyre G.D. Gimble J.M. Yield of human adipose-derived adult stem cells from liposuction aspirates.Cytotherapy. 2004; 6: 7-14Abstract Full Text Full Text PDF PubMed Scopus (529) Google Scholar). Undifferentiated L1s and hASCs (“preadipocytes”) have similar fibroblast-like morphologies. When induced to undergo terminal differentiation in adipogenic media, both change into round cells that exhibit properties typical of adipocytes in vivo, such as insulin-stimulated glucose uptake, lipogenesis, catecholamine-stimulated lipolysis, and adipokine secretion. These two models therefore provide the opportunity to study GRNs that govern similar adipogenic phenotypes against a background of phylogenetic, ontogenetic and technical differences. To facilitate comprehensive epigenomic profiling of cells undergoing adipogenesis, we expanded L1 and hASC preadipocytes and induced differentiation in adipogenic media. We selected four matched time points that represented similar stages of differentiation, as judged by morphology and lipid droplet accumulation. These time points corresponded to proliferating (day −2) and confluent (day 0) preadipocytes, immature adipocytes (day 2 for L1s, day 3 for hASCs), and mature adipocytes (day 7 for L1s, day 9 for hASCs). We generated genome-wide chromatin state maps using ChIP-Seq, profiling six histone modifications (H3K4me3/me2/me1, H3K27me3/ac, and H3K36me3) and the CCCTC-binding factor (CTCF) at all four time points. We also profiled the adipogenic TF peroxisome proliferators-activated receptor γ (PPARγ) at the last time point. The resulting data consist of 60 ChIP-Seq experiments and two negative controls. We also measured mRNA and miRNA expression levels in both models using microarrays. All data have been deposited in public databases. To visualize the data, we generated histograms of normalized densities of ChIP fragments across the genomes. Figure 1 shows these densities near the murine Pparg gene, which is strongly upregulated during adipogenesis (Figure S1, available online, shows the human PPARG locus). The profiled histone modifications and TFs showed spatial and temporal density distributions that are qualitatively consistent with their known functions (Hon et al., 2009Hon G.C. Hawkins R.D. Ren B. Predictive chromatin signatures in the mammalian genome.Hum. Mol. Genet. 2009; 18: R195-R201Crossref PubMed Scopus (166) Google Scholar). For example, H3K4me3, which is associated with transcriptional initiation, was primarily found near known promoters. A case in point is the gain of H3K4me3 observed near the adipocyte-specific alternative promoter of Pparg (P2, Figure 1). H3K4me2/me1 and H3K27ac, which are associated with “open” chromatin and cis-regulatory activity, showed dynamic distributions in promoter, intronic, and intergenic regions. H3K36me3, which is associated with transcriptional elongation, was distributed across active gene bodies and increased markedly across Pparg as it was upregulated. H3K27me3, which is associated with Polycomb-mediated repression, was distributed broadly across the inactive flanking regions. The PPARγ and CTCF densities showed sharp peaks, consistent with individual TF binding sites.Figure S1Chromatin State and TF Localization Near PPARG during hASC Adipogenesis, Related to Figure 1Show full captionHistograms of ChIP fragments across the human PPARG locus, normalized to fragments per 10 million aligned reads, for each of the profiled histone modifications and transcription factors at four time points during hASC adipogenesis. All histograms are shown on the same scale and high values were truncated as necessary.View Large Image Figure ViewerDownload Hi-res image Download (PPT) Histograms of ChIP fragments across the human PPARG locus, normalized to fragments per 10 million aligned reads, for each of the profiled histone modifications and transcription factors at four time points during hASC adipogenesis. All histograms are shown on the same scale and high values were truncated as necessary. To support quantitative analyses, we identified significant clusters of ChIP fragments using a sliding window approach for histone modifications and QuEST (Valouev et al., 2008Valouev A. Johnson D.S. Sundquist A. Medina C. Anton E. Batzoglou S. Myers R.M. Sidow A. Genome-wide analysis of transcription factor binding sites based on ChIP-Seq data.Nat. Methods. 2008; 5: 829-834Crossref PubMed Scopus (530) Google Scholar) for TF-binding sites. Each such region or binding site was assigned an “enrichment score,” which represents the ratio of observed over expected fragments. Their genome-wide distributions are consistent with the qualitative patterns described above (Table S1). mRNA and miRNA expression analyses revealed correlated expression dynamics that are consistent with efficient adipogenic differentiation (Figure S2, Table S2, and Extended Experimental Procedures). To compare data from the two models, we first attempted to map each enriched region in the mouse genome to corresponding regions of orthologous sequence in the human genome, and vice versa, using previously computed whole-genome alignments. About 80%–90% of these regions could be mapped to the other genome. We then asked whether these orthologous regions overlapped the same chromatin marks or TF-binding sites in the other model (conservatively requiring an overlap of ≥ 1 bp). We will refer to such overlaps as “shared” marks or binding sites and the remainder as “model-specific.” We conclude that the data provide a rich resource for studies of chromatin remodeling and gene regulation in two key models of adipogenesis. In the following sections, we focus on detection and functional analysis of cis-regulatory elements in the adipogenic GRN and the sequence-specific TFs that interact with them. We began our analysis by characterizing open chromatin marks in regions distal to (>2 kb from) known promoters. H3K4me2/me1 and H3K27ac were distributed in highly correlated patterns at each time point and changed dynamically in thousands of genomic regions in each cell culture model (Table S1). These “dynamic” regions were often clustered near genes with adipogenesis-dependent expression patterns, suggesting that they represent cooperative or redundant distal enhancers. Orthologous genes with similar expression in L1s and hASCs frequently showed similar chromatin marks, but the specific location of these marks was often model specific; this suggests that the expression pattern of genes is better conserved between the models than the specific elements controlling the expression. To identify putative distal enhancers, we focused on H3K27ac because recruitment of histone acetyltransferases (HATs) is the most specific signature known for these elements (Ghisletti et al., 2010Ghisletti S. Barozzi I. Mietton F. Polletti S. De Santa F. Venturini E. Gregory L. Lonie L. Chew A. Wei C.L. et al.Identification and characterization of enhancers controlling the inflammatory gene expression program in macrophages.Immunity. 2010; 32: 317-328Abstract Full Text Full Text PDF PubMed Scopus (441) Google Scholar, Heintzman et al., 2009Heintzman N.D. Hon G.C. Hawkins R.D. Kheradpour P. Stark A. Harp L.F. Ye Z. Lee L.K. Stuart R.K. Ching C.W. et al.Histone modifications at human enhancers reflect global cell-type-specific gene expression.Nature. 2009; 459: 108-112Crossref PubMed Scopus (1647) Google Scholar, Wang et al., 2008Wang Z. Zang C. Rosenfeld J.A. Schones D.E. Barski A. Cuddapah S. Cui K. Roh T.Y. Peng W. Zhang M.Q. Zhao K. Combinatorial patterns of histone acetylations and methylations in the human genome.Nat. Genet. 2008; 40: 897-903Crossref PubMed Scopus (1605) Google Scholar). We detected 29,092 distal H3K27ac regions in L1 adipocytes (day 7), with enrichment scores spanning an order of magnitude (Table S3). Of these, 6096 (∼21%) showed a ≥ 5-fold increase in enrichment scores relative to preadipocytes (days −2 and 0), suggesting that they harbor regulatory elements that recruit HATs during adipogenesis. Conversely, we identified 5159 H3K27ac regions in L1 preadipocytes (day −2) whose enrichment scores decreased at least 5-fold. We observed similar dynamics in hASCs (Table S3). Dynamic changes in open chromatin marks were significantly correlated with changes in the expression levels of linked genes. For simplicity, we assumed that each H3K27ac region was associated with the closest known gene (although there are counterexamples, as described below). Roughly 15% of all genes on our microarrays showed a ≥ 2-fold change in expression between L1 adipocytes and preadipocytes. We found that the more the expression level of a gene increased or decreased, the more likely it was to be associated with adipocyte- or preadipocyte-specific H3K27ac, respectively (Figure 2A ). Conversely, the likelihood that the expression level of a gene changed ≥ 2-fold was positively correlated with both the enrichment scores (Figure 2B) and the total number (Figure 2C) of dynamic H3K27ac regions associated with it. By contrast, association with invariant H3K27ac (enriched in both adipocytes and preadipocytes) had little predictive value with respect to changes in expression. We observed similar patterns in hASCs (Figures 2D–2F). Distal regions that show changes in open chromatin marks during adipogenesis are therefore likely to be enriched for cell type-specific enhancers. Moreover, genes with dynamic expression patterns appear to frequently be located near multiple such enhancers (see below and Figure S3).Figure S3Coordinated Chromatin in Multiple Distinct Regions Surrounding a Regulated Gene, Related to Figure 2Show full caption(A) Histograms of ChIP fragments near the murine Cebpa gene, which is strongly upregulated.(B) Histograms of ChIP fragments near the murine Fabp4 gene, which is strongly upregulated.(C) Histograms of ChIP fragments near the murine Osr1 gene, which is strongly downregulated.View Large Image Figure ViewerDownload Hi-res image Download (PPT) (A) Histograms of ChIP fragments near the murine Cebpa gene, which is strongly upregulated. (B) Histograms of ChIP fragments near the murine Fabp4 gene, which is strongly upregulated. (C) Histograms of ChIP fragments near the murine Osr1 gene, which is strongly downregulated. Comparing open chromatin marks between L1s and hASCs, we found that ∼15%–30% of marks identified in one model were shared with the other model (that is, orthologous sequences contained the same chromatin marks). Given that regions enriched for each open chromatin mark only covered ∼2%–4% of each genome, this represents a highly significant degree of overlap. Regions with the same size distributions randomly placed across the two genomes would have an expected overlap of less than 0.5%. Nevertheless, the majority (70%–85%) of distal open chromatin marks were model specific. Orthologs that were only associated with dynamic open chromatin marks in one of the models often showed discordant expression patterns. For example, orthologs whose expression increased more in L1s than in hASCs were also more likely to be associated with adipocyte-specific H3K27ac only in L1s and vice versa (Figure 2G). This suggests that model-specific open chromatin marks correlate with model-specific enhancers. Of interest, orthologous genes with similar expression patterns often had similar chromatin marks nearby, but the specific locations of these marks were typically model specific. For example, at orthologous loci induced ≥ 2-fold in both models, the majority (84%) of adipocyte-specific H3K27ac regions in L1s were not shared with hASCs and vice versa. Their expression patterns therefore appear to be better conserved than the specific enhancers that regulate them (below, we verify this observation through functional analyses). We next analyzed the distribution of binding sites for PPARγ in mature L1 and hASC adipocytes (day 7/9). PPARγ is a nuclear receptor that is recruited to PPAR response elements (PPREs) during adipogenesis as a heterodimer with retinoid X receptors (RXRs) (IJpenberg et al., 1997IJpenberg A. Jeannin E. Wahli W. Desvergne B. Polarity and specific sequence requirements of peroxisome proliferator-activated receptor (PPAR)/retinoid X receptor heterodimer binding to DNA. A functional analysis of the malic enzyme gene PPAR response element.J. Biol. Chem. 1997; 272: 20108-20117Crossref PubMed Scopus (276) Google Scholar) and primarily functions as a transcriptional activator (Lefterova et al., 2008Lefterova M.I. Zhang Y. Steger D.J. Schupp M. Schug J. Cristancho A. Feng D. Zhuo D. Stoeckert Jr., C.J. Liu X.S. Lazar M.A. PPARgamma and C/EBP factors orchestrate adipocyte biology via adjacent binding on a genome-wide scale.Genes Dev. 2008; 22: 2941-2952Crossref PubMed Scopus (544) Google Scholar, Nielsen et al., 2008Nielsen R. Pedersen T.A. Hagenbeek D. Moulos P. Siersbaek R. Megens E. Denissov S. Børgesen M. Francoijs K.J. Mandrup S. Stunnenberg H.G. Genome-wide profiling of PPARgamma:RXR and RNA polymerase II occupancy reveals temporal activation of distinct metabolic pathways and changes in RXR dimer composition during adipogenesis.Genes Dev. 2008; 22: 2953-2967Crossref PubMed Scopus (410) Google Scholar). We found that PPARγ was largely localized to distal regions enriched for open chromatin marks. The vast majority of PPARγ-binding sites were not shared between L1s and hASCs, and this could be explained, in part, by turnover of its motif in the genome sequences. Loci with PPARγ-binding sites in both L1s and hASCs were, however, highly enriched for genes with functions relevant to known adipocyte biology. We detected 7,142 and 39,986 PPARγ-binding sites in L1s and hASCs, respectively (1% FDR; Table S4), with enrichment scores spanning two orders of magnitude. The excess number of human sites primarily reflects the identification of more weak binding sites (Extended Experimental Procedures). Performing ChIP-Seq with a different PPARγ antibody yielded similar results, and the L1 sites reported here showed good concordance with 5299 sites previously detected in this model using ChIP-chip (Lefterova et al., 2008Lefterova M.I. Zhang Y. Steger D.J. Schupp M. Schug J. Cristancho A. Feng D. Zhuo D. Stoeckert Jr., C.J. Liu X.S. Lazar M.A. PPARgamma and C/EBP factors orchestrate adipocyte biology via adjacent binding on a genome-wide scale.Genes Dev. 2008; 22: 2941-2952Crossref PubMed Scopus (544) Google Scholar; Figure S4). The PPARγ-binding sites followed qualitatively similar patterns in L1s and hASCs, with the vast majority (85%–95%) overlapping open chromatin marks (Figure 3A ). Ab initio motif discovery recovered motifs that were similar to the canonical PPARγ/RXR DR1 motif (Figure 3B). There are, however, ∼1.5 million instances of these motifs in each genome; this implies that we detected PPARγ binding at only ∼1 in 200 motifs in the mouse genome. Other factors must therefore contribute to binding site selectivity. Of note, a motif instance was ∼15 times more likely be bound by PPARγ in L1 adipocytes if it overlapped a region enriched for open chromatin marks in preadipocytes (pFisher < 10−60). In fact, ∼77% of all PPARγ-binding sites detected in L1s were located in such regions. This suggests that PPARγ recruitment during adipogenesis is strongly influenced by the preadipocyte chromatin state. The majority (79%) of sites bound by PPARγ in L1s were not shared with hASCs, despite the larger number of sites detected in the latter model. Of note, 34% of L1 PPARγ-binding sites that could not be mapped to the human genome resided within rodent-specific transposable element insertions, which implies that they evolved after the mouse and human lineages diverged. If an L1-binding site could be mapped to an orthologous human sequence, the presence of PPARγ binding in hASCs correlated with the presence a conserved motif and open chromatin marks (Figure 3C). Evolutionary turnover of DR1-like motifs is therefore likely to contribute to the differential recruitment of PPARγ and open chromatin marks between the two models. To explore the correlation between PPARγ recruitment and gene expression, we again assumed that each binding site was associated with the closest known protein-coding gene. We found that genes associated with PPARγ in L1s were approximately three times more likely than nonassociated genes to be upregulated ≥ 2-fold (pFisher < 10−60). The majority (84%) of genes associated with PPARγ-binding sites were not upregulated, but the likelihood that a gene was upregulated increased when an associated PPARγ binding site had a higher ChIP enrichment score; was shared with hASCs; or overlapped adipocyte-specific H3K27ac (Figure 3D). The correlation between upregulation and gain of H3K27ac is notable. It suggests that, whereas PPARγ binding is biased toward regions that were already acetylated in preadipocytes, PPARγ-binding sites that recruit HATs to new locations are more likely to be functionally relevant. Concentrating on genes with “dynamic” PPARγ-binding sites that gain H3K27ac in L1s, we found that those for which the orthologous human gene was not associated with H3K27ac in hASCs were significantly more likely to show greater upregulation in L1s than in hASCs and vice versa (pFisher < 10−4; Figure 3E). Model-specific PPREs are therefore likely to contribute to differential gene regulation in the two models. Annotation enrichment analysis (Dennis et al., 2003Dennis Jr., G. Sherman B.T. Hosack D.A. Yang J. Gao W. Lane H.C. Lempicki R.A. DAVID: Database for Annotation, Visualization, and Integrated Discovery.Genome Biol. 2003; 4: P3Crossref PubMed Google Scholar) revealed, however, that genes that were associated with PPARγ-binding sites and upregulated in both models were strongly enriched for components of the classic PPARγ signaling pathway, as well as essential adipocyte functions related to lipid metabolism and cellular respiration (Figure 3F). Of note, only ∼57% of these concordantly upregulated genes actually shared orthologous PPARγ-binding sites. Thus, PPARγ targeting of adipocyte genes appears to be better conserved than the specific PPREs that mediate PPARγ recruitment to these genes. We next analyzed the distribution of binding sites for CTCF, a DNA-binding protein that plays a key role in higher-order organization of chromatin and is associated with insulator and enhancer-blocking activities (Phillips and Corces, 2009Phillips J.E. Corces V.G. CTCF: master weaver of the genome.Cell. 2009; 137: 1194-1211Abstract Full Text Full Text PDF PubMed Scopus (1037) Google Scholar). We found that CTCF recruitment was relatively invariant during differentiation in each model but that the specific binding sites differed significantly between the two models. These differences appear to be largely caused by evolutionary turnover of CTCF motifs. We detected ∼43,000 CTCF-binding sites at each time point in each model (1% FDR). The sites followed largely intergenic distributions similar to those described in other cell types (Barski et al., 2007Barski 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 genome.Cell. 2007; 129: 823-837Abstract Full Text Full Text PDF PubMed Scopus (4697) Google Scholar, Kim et al., 2007Kim T.H. Abdullaev Z.K. Smith A.D. Ching K.A. Loukinov D.I. Green R.D. Zhang M.Q. Lobanenkov V.V. Ren B. Analysis of the vertebrate insulator protein CTCF-binding sites in the human genome.Cell. 2007; 128: 1231-1245Abstract Full Text Full Text PDF PubMed Scopus (748) Google Scholar, Xi et al., 2007Xi H. Shulha H.P. Lin J.M. Vales T." @default.
• W2012071760 created "2016-06-24" @default.
• W2012071760 creator A5000661697 @default.
• W2012071760 creator A5012278873 @default.
• W2012071760 creator A5020748592 @default.
• W2012071760 creator A5026038537 @default.
• W2012071760 creator A5061926947 @default.
• W2012071760 creator A5078653034 @default.
• W2012071760 creator A5090678260 @default.
• W2012071760 date "2010-10-01" @default.
• W2012071760 modified "2023-10-11" @default.
• W2012071760 title "Comparative Epigenomic Analysis of Murine and Human Adipogenesis" @default.
• W2012071760 cites W1548779692 @default.
• W2012071760 cites W1971279197 @default.
• W2012071760 cites W1987414043 @default.
• W2012071760 cites W1987935765 @default.
• W2012071760 cites W1988871238 @default.
• W2012071760 cites W1990362235 @default.
• W2012071760 cites W1998464620 @default.
• W2012071760 cites W2007684630 @default.
• W2012071760 cites W2011869493 @default.
• W2012071760 cites W2015416439 @default.
• W2012071760 cites W2017840637 @default.
• W2012071760 cites W2020672093 @default.
• W2012071760 cites W2022110934 @default.
• W2012071760 cites W2029722028 @default.
• W2012071760 cites W2038239151 @default.
• W2012071760 cites W2048002474 @default.
• W2012071760 cites W2049073862 @default.
• W2012071760 cites W2049361027 @default.
• W2012071760 cites W2054965892 @default.
• W2012071760 cites W2055143633 @default.
• W2012071760 cites W2057390103 @default.
• W2012071760 cites W2059699217 @default.
• W2012071760 cites W2059992589 @default.
• W2012071760 cites W2065485581 @default.
• W2012071760 cites W2073026107 @default.
• W2012071760 cites W2078526277 @default.
• W2012071760 cites W2083065209 @default.
• W2012071760 cites W2083245183 @default.
• W2012071760 cites W2088986234 @default.
• W2012071760 cites W2095005830 @default.
• W2012071760 cites W2097220641 @default.
• W2012071760 cites W2101951807 @default.
• W2012071760 cites W2102304447 @default.
• W2012071760 cites W2102638899 @default.
• W2012071760 cites W2106726734 @default.
• W2012071760 cites W2110065044 @default.
• W2012071760 cites W2122683221 @default.
• W2012071760 cites W2123564569 @default.
• W2012071760 cites W2124025352 @default.
• W2012071760 cites W2129795133 @default.
• W2012071760 cites W2130618859 @default.
• W2012071760 cites W2131820599 @default.
• W2012071760 cites W2163775594 @default.
• W2012071760 cites W2164207497 @default.
• W2012071760 cites W2166094946 @default.
• W2012071760 doi "https://doi.org/10.1016/j.cell.2010.09.006" @default.
• W2012071760 hasPubMedCentralId "https://www.ncbi.nlm.nih.gov/pmc/articles/2950833" @default.
• W2012071760 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/20887899" @default.
• W2012071760 hasPublicationYear "2010" @default.
• W2012071760 type Work @default.
• W2012071760 sameAs 2012071760 @default.
• W2012071760 citedByCount "470" @default.
• W2012071760 countsByYear W20120717602012 @default.
• W2012071760 countsByYear W20120717602013 @default.
• W2012071760 countsByYear W20120717602014 @default.
• W2012071760 countsByYear W20120717602015 @default.
• W2012071760 countsByYear W20120717602016 @default.
• W2012071760 countsByYear W20120717602017 @default.
• W2012071760 countsByYear W20120717602018 @default.
• W2012071760 countsByYear W20120717602019 @default.
• W2012071760 countsByYear W20120717602020 @default.
• W2012071760 countsByYear W20120717602021 @default.
• W2012071760 countsByYear W20120717602022 @default.
• W2012071760 countsByYear W20120717602023 @default.
• W2012071760 crossrefType "journal-article" @default.
• W2012071760 hasAuthorship W2012071760A5000661697 @default.
• W2012071760 hasAuthorship W2012071760A5012278873 @default.
• W2012071760 hasAuthorship W2012071760A5020748592 @default.
• W2012071760 hasAuthorship W2012071760A5026038537 @default.
• W2012071760 hasAuthorship W2012071760A5061926947 @default.
• W2012071760 hasAuthorship W2012071760A5078653034 @default.
• W2012071760 hasAuthorship W2012071760A5090678260 @default.
• W2012071760 hasBestOaLocation W20120717601 @default.
• W2012071760 hasConcept C104317684 @default.
• W2012071760 hasConcept C121912465 @default.
• W2012071760 hasConcept C122927707 @default.
• W2012071760 hasConcept C150194340 @default.
• W2012071760 hasConcept C190727270 @default.
• W2012071760 hasConcept C54355233 @default.
• W2012071760 hasConcept C70721500 @default.
• W2012071760 hasConcept C86803240 @default.
• W2012071760 hasConceptScore W2012071760C104317684 @default.
• W2012071760 hasConceptScore W2012071760C121912465 @default.
• W2012071760 hasConceptScore W2012071760C122927707 @default.
• W2012071760 hasConceptScore W2012071760C150194340 @default.
• W2012071760 hasConceptScore W2012071760C190727270 @default.

• next