Mapping and functional characterisation of a CTCF-dependent insulator element at the 3' border of the murine Scl transcriptional domain.

The Scl gene encodes a transcription factor essential for haematopoietic development. Scl transcription is regulated by a panel of cis-elements spread over 55 kb with the most distal 3' element being located downstream of the neighbouring gene Map17, which is co-regulated with Scl in haematopoietic cells. The Scl/Map17 domain is flanked upstream by the ubiquitously expressed Sil gene and downstream by a cluster of Cyp genes active in liver, but the mechanisms responsible for delineating the domain boundaries remain unclear. Here we report identification of a DNaseI hypersensitive site at the 3' end of the Scl/Map17 domain and 45 kb downstream of the Scl transcription start site. This element is located at the boundary of active and inactive chromatin, does not function as a classical tissue-specific enhancer, binds CTCF and is both necessary and sufficient for insulator function in haematopoietic cells in vitro. Moreover, in a transgenic reporter assay, tissue-specific expression of the Scl promoter in brain was increased by incorporation of 350 bp flanking fragments from the +45 element. Our data suggests that the +45 region functions as a boundary element that separates the Scl/Map17 and Cyp transcriptional domains, and raise the possibility that this element may be useful for improving tissue-specific expression of transgenic constructs.

Pdgfrα and Flk1 are direct target genes of Mixl1 in differentiating embryonic stem cells.

The Mixl1 homeodomain protein plays a key role in mesendoderm patterning during embryogenesis, but its target genes remain to be identified. We compared gene expression in differentiating heterozygous Mixl1(GFP/w) and homozygous null Mixl1(GFP/Hygro) mouse embryonic stem cells to identify potential downstream transcriptional targets of Mixl1. Candidate Mixl1 regulated genes whose expression was reduced in GFP+ cells isolated from differentiating Mixl1(GFP/Hygro) embryoid bodies included Pdgfrα and Flk1. Mixl1 bound to ATTA sequences located in the Pdgfrα and Flk1 promoters and chromatin immunoprecipitation assays confirmed Mixl1 occupancy of these promoters in vivo. Furthermore, Mixl1 transactivated the Pdgfrα and Flk1 promoters through ATTA sequences in a DNA binding dependent manner. These data support the hypothesis that Mixl1 directly regulates Pdgfrα and Flk1 gene expression and strengthens the position of Mixl1 as a key regulator of mesendoderm development during mammalian gastrulation.

A Runx1-Smad6 rheostat controls Runx1 activity during embryonic hematopoiesis.

The oncogenic transcription factor Runx1 is required for the specification of definitive hematopoietic stem cells (HSC) in the developing embryo. The activity of this master regulator is tightly controlled during development. The transcription factors that upregulate the expression of Runx1 also upregulate the expression of Smad6, the inhibitory Smad, which controls Runx1 activity by targeting it to the proteasome. Here we show that Runx1, in conjunction with Fli1, Gata2, and Scl, directly regulates the expression of Smad6 in the aorta-gonad-mesonephros (AGM) region in the developing embryo, where HSCs originate. Runx1 regulates Smad6 activity via a novel upstream enhancer, and Runx1 null embryos show reduced Smad6 transcripts in the yolk-sac and c-Kit-positive fetal liver cells. By directly regulating the expression of Smad6, Runx1 sets up a functional rheostat to control its own activity. The perturbation of this rheostat, using a proteasomal inhibitor, results in an increase in Runx1 and Smad6 levels that can be directly attributed to increased Runx1 binding to tissue-specific regulatory elements of these genes. Taken together, we describe a scenario in which a key hematopoietic transcription factor controls its own expression levels by transcriptionally controlling its controller.

Epigenetic silencing of BIM in glucocorticoid poor-responsive pediatric acute lymphoblastic leukemia, and its reversal by histone deacetylase inhibition.

Glucocorticoids play a critical role in the therapy of lymphoid malignancies, including pediatric acute lymphoblastic leukemia (ALL), although the mechanisms underlying cellular resistance remain unclear. We report glucocorticoid resistance attributable to epigenetic silencing of the BIM gene in pediatric ALL biopsies and xenografts established in immune-deficient mice from direct patient explants as well as a therapeutic approach to reverse resistance in vivo. Glucocorticoid resistance in ALL xenografts was consistently associated with failure to up-regulate BIM expression after dexamethasone exposure despite confirmation of a functional glucocorticoid receptor. Although a comprehensive assessment of BIM CpG island methylation revealed no consistent changes, glucocorticoid resistance in xenografts and patient biopsies significantly correlated with decreased histone H3 acetylation. Moreover, the histone deacetylase inhibitor vorinostat relieved BIM repression and exerted synergistic antileukemic efficacy with dexamethasone in vitro and in vivo. These findings provide a novel therapeutic strategy to reverse glucocorticoid resistance and improve outcome for high-risk pediatric ALL.

Phosphorylation and acetylation of histone H3 and autoregulation by early growth response 1 mediate interleukin 1beta induction of early growth response 1 transcription.

OBJECTIVE: The transcription factor early growth response (EGR)-1 has been implicated as a key vascular phenotypic switch through its control of inducible transcription. EGR-1 autoregulation, and histone modification in the EGR-1 promoter, represent key mechanisms in EGR-1 control, but have not been explored. METHODS AND RESULTS: We demonstrate that EGR-1 regulates its own transcription and that this involves histone H3 phosphorylation and acetylation. EGR-1 transactivates its promoter in smooth muscle cells exposed to interleukin (IL) 1beta through a novel cis-acting element (-211/-203). PD98059, which inhibits mitogen-activated protein kinase kinase/extracellular regulated kinase (MEK/ERK) attenuates IL-1beta-inducible phosphorylation of extracellular signal-regulated kinase 1/2 and mitogen and stress-activated protein kinases 1/2; and reduces levels of phosphorylated and acetylated histone H3. Histone deacetylase inhibition enhances EGR-1 transcription in response to cytokine. Conversely, suppression of histone modification with mitogen and stress-activated protein kinase 1/2 short interfering RNA, or the histone H3 acetyltransferase inhibitor Garcinol, inhibits IL-1beta-inducible EGR-1 transcription. EGR-1 interacts with the acetyltransferase p300. Acetylated H3 and phosphorylated H3 are enriched at the promoter of EGR-1; and EGR-1 is enriched at the promoters of tissue factor and plasminogen activator inhibitor 1 in response to IL-1beta, and attenuated by PD98059, Garcinol, and mitogen and stress-activated protein kinase 1/2 short interfering RNA. CONCLUSIONS: IL-1beta induction of EGR-1 transcription involves histone H3 phosphorylation, acetylation, and autoregulation by EGR-1.

Expression of the leukemia oncogene Lmo2 is controlled by an array of tissue-specific elements dispersed over 100 kb and bound by Tal1/Lmo2, Ets, and Gata factors.

The Lmo2 gene encodes a transcriptional cofactor critical for the development of hematopoietic stem cells. Ectopic LMO2 expression causes leukemia in T-cell acute lymphoblastic leukemia (T-ALL) patients and severe combined immunodeficiency patients undergoing retroviral gene therapy. Tightly controlled Lmo2 expression is therefore essential, yet no comprehensive analysis of Lmo2 regulation has been published so far. By comparative genomics, we identified 17 highly conserved noncoding elements, 9 of which revealed specific acetylation marks in chromatin-immunoprecipitation and microarray (ChIP-chip) assays performed across 250 kb of the Lmo2 locus in 11 cell types covering different stages of hematopoietic differentiation. All candidate regulatory regions were tested in transgenic mice. An extended LMO2 proximal promoter fragment displayed strong endothelial activity, while the distal promoter showed weak forebrain activity. Eight of the 15 distal candidate elements functioned as enhancers, which together recapitulated the full expression pattern of Lmo2, directing expression to endothelium, hematopoietic cells, tail, and forebrain. Interestingly, distinct combinations of specific distal regulatory elements were required to extend endothelial activity of the LMO2 promoter to yolk sac or fetal liver hematopoietic cells. Finally, Sfpi1/Pu.1, Fli1, Gata2, Tal1/Scl, and Lmo2 were shown to bind to and transactivate Lmo2 hematopoietic enhancers, thus identifying key upstream regulators and positioning Lmo2 within hematopoietic regulatory networks.

Endoglin expression in blood and endothelium is differentially regulated by modular assembly of the Ets/Gata hemangioblast code.

Endoglin is an accessory receptor for TGF-beta signaling and is required for normal hemangioblast, early hematopoietic, and vascular development. We have previously shown that an upstream enhancer, Eng -8, together with the promoter region, mediates robust endothelial expression yet is inactive in blood. To identify hematopoietic regulatory elements, we used array-based methods to determine chromatin accessibility across the entire locus. Subsequent transgenic analysis of candidate elements showed that an endothelial enhancer at Eng +9 when combined with an element at Eng +7 functions as a strong hemato-endothelial enhancer. Chromatin immunoprecipitation (ChIP)-chip analysis demonstrated specific binding of Ets factors to the promoter as well as to the -8, +7+9 enhancers in both blood and endothelial cells. By contrast Pu.1, an Ets factor specific to the blood lineage, and Gata2 binding was only detected in blood. Gata2 was bound only at +7 and GATA motifs were required for hematopoietic activity. This modular assembly of regulators gives blood and endothelial cells the regulatory freedom to independently fine-tune gene expression and emphasizes the role of regulatory divergence in driving functional divergence.

Real-time PCR mapping of DNaseI-hypersensitive sites using a novel ligation-mediated amplification technique.

Mapping sites within the genome that are hypersensitive to digestion with DNaseI is an important method for identifying DNA elements that regulate transcription. The standard approach to locating these DNaseI-hypersensitive sites (DHSs) has been to use Southern blotting techniques, although we, and others, have recently published alternative methods using a range of technologies including high-throughput sequencing and genomic array tiling paths. In this article, we describe a novel protocol to use real-time PCR to map DHS. Advantages of the technique reported here include the small cell numbers required for each analysis, rapid, relatively low-cost experiments with minimal need for specialist equipment. Presented examples include comparative DHS mapping of known TAL1/SCL regulatory elements between human embryonic stem cells and K562 cells.

Endocrine cells develop within pancreatic bud-like structures derived from mouse ES cells differentiated in response to BMP4 and retinoic acid.

We have examined factors affecting the in vitro differentiation of Pdx1(GFP/w) ESCs to pancreatic endocrine cells. Inclusion of Bone Morphogenetic Protein 4 (BMP4) during the first four days of differentiation followed by a 24-hour pulse of retinoic acid (RA) induced the formation of GFP(+) embryoid bodies (EBs). GFP expression was restricted to E-cadherin(+) tubes and GFP bright (GFP(br)) buds, reminiscent of GFP(+) early foregut endoderm and GFP(br) pancreatic buds observed in Pdx1(GFP/w) embryos. These organoid structures developed without further addition of exogenous factors between days 5 and 12, suggesting that day 5 EBs contained a template for the subsequent phase of development. EBs treated with nicotinamide after day 12 of differentiation expressed markers of endocrine and exocrine differentiation, but only in cells within the GFP(br) buds. Analysis of Pdx1(GFP/w) ESCs modified by targeting a dsRed1 gene to the Ins1 locus (Pdx1(GFP/w)Ins1(RFP/w) ESCs) provided corroborating evidence that insulin positive cells arose from GFP(br) buds, mirroring the temporal relationship between pancreatic bud development and the formation of endocrine cells in the developing embryo. The readily detectable co-expression of GFP and RFP in grafts derived from transplanted EBs demonstrated the utility of Pdx1(GFP/w)Ins1(RFP/w) ESCs for investigating pancreatic differentiation in vitro and in vivo.

Identifying gene regulatory elements by genomic microarray mapping of DNaseI hypersensitive sites.

The identification of cis-regulatory elements is central to understanding gene transcription. Hypersensitivity of cis-regulatory elements to digestion with DNaseI remains the gold-standard approach to locating such elements. Traditional methods used to identify DNaseI hypersensitive sites are cumbersome and can only be applied to short stretches of DNA at defined locations. Here we report the development of a novel genomic array-based approach to DNaseI hypersensitive site mapping (ADHM) that permits precise, large-scale identification of such sites from as few as 5 million cells. Using ADHM we identified all previously recognized hematopoietic regulatory elements across 200 kb of the mouse T-cell acute lymphocytic leukemia-1 (Tal1) locus, and, in addition, identified two novel elements within the locus, which show transcriptional regulatory activity. We further validated the ADHM protocol by mapping the DNaseI hypersensitive sites across 250 kb of the human TAL1 locus in CD34+ primary stem/progenitor cells and K562 cells and by mapping the previously known DNaseI hypersensitive sites across 240 kb of the human alpha-globin locus in K562 cells. ADHM provides a powerful approach to identifying DNaseI hypersensitive sites across large genomic regions.

Retinoic acid induces Pdx1-positive endoderm in differentiating mouse embryonic stem cells.

We have generated an embryonic stem (ES) cell line in which sequences encoding green fluorescent protein (GFP) were targeted to the locus of the pancreatic-duodenal homeobox gene (Pdx1). Analysis of chimeric embryos derived from blastocyst injection of Pdx1(GFP/w) ES cells demonstrated that the pattern of GFP expression was consistent with that reported for the endogenous Pdx1 gene. By monitoring GFP expression during the course of ES cell differentiation, we have shown that retinoic acid (RA) can regulate the commitment of ES cells to form Pdx1(+) pancreatic endoderm. RA was most effective at inducing Pdx1 expression when added to cultures at day 4 of ES differentiation, a period corresponding to the end of gastrulation in the embryo. RT-PCR analysis showed that Pdx1-positive cells from day 8 cultures expressed the early endoderm markers Ptf1a, Foxa2, Hnf4alpha, Hnf1beta, and Hnf6, consistent with the notion that they corresponded to the early pancreatic endoderm present in the embryonic day 9.5 mouse embryo. These results demonstrate the utility of Pdx1(GFP/w) ES cells as a tool for monitoring the effects of factors that influence pancreatic differentiation from ES cells.