Search

The Drosophila brain is a frequently used model in neuroscience. Single-cell transcriptome analysis1-6, three-dimensional morphological classification7 and electron microscopy mapping of the connectome8,9 have revealed an immense diversity of neuronal and glial cell types that underlie an array of functional and behavioural traits in the fly. The identities of these cell types are controlled by gene regulatory networks (GRNs), involving combinations of transcription factors that bind to genomic enhancers to regulate their target genes. Here, to characterize GRNs at the cell-type level in the fly brain, we profiled the chromatin accessibility of 240,919 single cells spanning 9 developmental timepoints and integrated these data with single-cell transcriptomes. We identify more than 95,000 regulatory regions that are used in different neuronal cell types, of which 70,000 are linked to developmental trajectories involving neurogenesis, reprogramming and maturation. For 40 cell types, uniquely accessible regions were associated with their expressed transcription factors and downstream target genes through a combination of motif discovery, network inference and deep learning, creating enhancer GRNs. The enhancer architectures revealed by DeepFlyBrain lead to a better understanding of neuronal regulatory diversity and can be used to design genetic driver lines for cell types at specific timepoints, facilitating their characterization and manipulation. ispartof: NATURE vol:601 issue:7894 pages:630-+ ispartof: location:England status: published

We present cisTopic, a probabilistic framework used to simultaneously discover coaccessible enhancers and stable cell states from sparse single-cell epigenomics data ( http://github.com/aertslab/cistopic ). Using a compendium of single-cell ATAC-seq datasets from differentiating hematopoietic cells, brain and transcription factor perturbations, we demonstrate that topic modeling can be exploited for robust identification of cell types, enhancers and relevant transcription factors. cisTopic provides insight into the mechanisms underlying regulatory heterogeneity in cell populations. As an unsupervised Bayesian framework, cisTopic classifies regions in scATAC-seq data into regulatory topics, which are used for clustering.

Deciphering the genomic regulatory code of enhancers is a key challenge in biology because this code underlies cellular identity. A better understanding of how enhancers work will improve the interpretation of noncoding genome variation and empower the generation of cell type-specific drivers for gene therapy. Here, we explore the combination of deep learning and cross-species chromatin accessibility profiling to build explainable enhancer models. We apply this strategy to decipher the enhancer code in melanoma, a relevant case study owing to the presence of distinct melanoma cell states. We trained and validated a deep learning model, called DeepMEL, using chromatin accessibility data of 26 melanoma samples across six different species. We show the accuracy of DeepMEL predictions on the CAGI5 challenge, where it significantly outperforms existing models on the melanoma enhancer of IRF4 Next, we exploit DeepMEL to analyze enhancer architectures and identify accurate transcription factor binding sites for the core regulatory complexes in the two different melanoma states, with distinct roles for each transcription factor, in terms of nucleosome displacement or enhancer activation. Finally, DeepMEL identifies orthologous enhancers across distantly related species, where sequence alignment fails, and the model highlights specific nucleotide substitutions that underlie enhancer turnover. DeepMEL can be used from the Kipoi database to predict and optimize candidate enhancers and to prioritize enhancer mutations. In addition, our computational strategy can be applied to other cancer or normal cell types. ispartof: GENOME RESEARCH vol:30 issue:12 ispartof: location:United States status: published

Transcriptional enhancers function as docking platforms for combinations of transcription factors (TFs) to control gene expression. How enhancer sequences determine nucleosome occupancy, TF recruitment and transcriptional activation in vivo remains unclear. Using ATAC–seq across a panel of Drosophila inbred strains, we found that SNPs affecting binding sites of the TF Grainy head (Grh) causally determine the accessibility of epithelial enhancers. We show that deletion and ectopic expression of Grh cause loss and gain of DNA accessibility, respectively. However, although Grh binding is necessary for enhancer accessibility, it is insufficient to activate enhancers. Finally, we show that human Grh homologs—GRHL1, GRHL2 and GRHL3—function similarly. We conclude that Grh binding is necessary and sufficient for the opening of epithelial enhancers but not for their activation. Our data support a model positing that complex spatiotemporal expression patterns are controlled by regulatory hierarchies in which pioneer factors, such as Grh, establish tissue-specific accessible chromatin landscapes upon which other factors can act. ispartof: NATURE GENETICS vol:50 issue:7 pages:1011-1020 ispartof: location:United States status: published

Genome control is operated by transcription factors (TFs) controlling their target genes by binding to promoters and enhancers. Conceptually, the interactions between TFs, their binding sites, and their functional targets are represented by gene regulatory networks (GRNs). Deciphering in vivo GRNs underlying organ development in an unbiased genome-wide setting involves identifying both functional TF-gene interactions and physical TF-DNA interactions. To reverse engineer the GRNs of eye development in Drosophila, we performed RNA-seq across 72 genetic perturbations and sorted cell types and inferred a coexpression network. Next, we derived direct TF-DNA interactions using computational motif inference, ultimately connecting 241 TFs to 5,632 direct target genes through 24,926 enhancers. Using this network, we found network motifs, cis-regulatory codes, and regulators of eye development. We validate the predicted target regions of Grainyhead by ChIP-seq and identify this factor as a general cofactor in the eye network, being bound to thousands of nucleosome-free regions. publisher: Elsevier articletitle: Mapping Gene Regulatory Networks in Drosophila Eye Development by Large-Scale Transcriptome Perturbations and Motif Inference journaltitle: Cell Reports articlelink: http://dx.doi.org/10.1016/j.celrep.2014.11.038 content_type: article copyright: Copyright © 2014 The Authors. Published by Elsevier Inc. ispartof: Cell Reports vol:9 issue:6 pages:2290-303 ispartof: location:United States status: published

Background CD36 is a membrane glycoprotein, contributing to the pathogenesis of metabolic disorders, like obesity, which has become a major health concern worldwide. Methods A potential functional role of the scavenger receptor CD36 was investigated in in vitro adipocyte differentiation and in vivo adipogenesis. Results During differentiation of 3T3-F442A preadipocytes into mature adipocytes, expression of CD36 was upregulated and CD36 gene silencing resulted in impaired differentiation, as monitored by Oil Red O staining and expression of adipogenic markers. De novo fat pad formation in NUDE mice following injection of preadipocytes was significantly reduced upon CD36 gene silencing as compared to control. This was associated with marked adipocyte hypotrophy and reduced adipose tissue adipocyte content. Macrophage infiltration in de novo fat tissues derived from preadipocytes with CD36 gene silencing was not significantly different from controls. Collagen content was significantly higher in de novo fat with CD36 gene silencing. In a nutritionally induced obesity model, total body weight as well as subcutaneous and gonadal adipose tissue mass were significantly lower in CD36 deficient mice as compared to wild-type littermates. General significance Thus, our data support a functional role of CD36 in promoting adipogenesis in vitro as well as in vivo.

Type 2 diabetes is causally related to obesity and characterized by dysfunctional pancreatic beta cells. It is so far unclear whether direct interactions exist between adipocytes and beta cells and possibly raise any pathogenic relevance. In this study, we examined whether 9-day co-cultured 3T3-F442A (pre)adipocytes and primary rat pancreatic beta cells exert an influence on each other's function. In the presence of beta cells, 3T3-F442A cells became lipid-storing cells expressing markers of differentiated adipocytes and releasing adiponectin. This effect was attributed to the medium insulin levels (around 0.1 μM) and was associated with an elevated glucose consumption by the 3T3-F442A cells. The subsequent decrease in medium glucose concentration reduced the rate of insulin release by beta cells cultured at 10 mM glucose, and thus suppressed their degranulation during culture. These changes in beta cell function did not occur at 20 mM glucose and were reversible upon removal of the 3T3-F422A cells. They could not be reproduced by 3T3-F422A-conditioned medium containing varying adiponectin concentrations. These data indicate that insulin secreted by beta cells is sufficient to induce differentiation of preadipocytes without addition of exogenous adipogenic factors. Over 9 days culture, (pre)adipocytes did not directly and irreversibly affect beta cell functions.

The potential role of placental growth factor (PlGF) in early stages of adipogenesis was investigated in vivo using a murine model of obesity, as well as in vitro using cultured preadipocytes. PlGF-deficient (PlGF-/-) and wild-type (WT) mice, kept on high fat diet (HFD) for 3 weeks, had comparable body weight and weight of subcutaneous (SC) and gonadal (GON) adipose tissues. Blood vessel size and blood vessel density, normalized to adipocyte number, were not significantly different in SC and GON adipose tissues of both genotypes. Differentiation of embryonic fibroblasts derived from WT or PlGF-/- mice into mature adipocytes was comparable. Furthermore, addition of recombinant PlGF, of the PlGF neutralizing MAb PL5D11D4 or of the anti-Flk-1 MAb DC101 to cultured 3T3-F442A preadipocytes did not significantly affect their differentiation into mature adipocytes. Ex vivo blood vessel outgrowth following seeding of adipose tissue-derived microvessel fragments in 3D-collagen gels was not affected by PlGF deficiency. Thus, in murine model systems, PlGF does not seem to play an important role in early adipogenesis.

Scoring the impact of noncoding variation on the function of cis-regulatory regions, on their chromatin state, and on the qualitative and quantitative expression levels of target genes is a fundamental problem in evolutionary genomics. A particular challenge is how to model the divergence of quantitative traits and to identify relationships between the changes across the different levels of the genome, the chromatin activity landscape, and the transcriptome. Here, we examine the use of the Ornstein-Uhlenbeck (OU) model to infer selection at the level of predicted cis-regulatory modules (CRMs), and link these with changes in transcription factor binding and chromatin activity. Using publicly available cross-species ChIP-Seq and STARR-Seq data we show how OU can be applied genome-wide to identify candidate transcription factors for which binding site and CRM turnover is correlated with changes in regulatory activity. Next, we profile open chromatin in the developing eye across three Drosophila species. We identify the recognition motifs of the chromatin remodelers, Trithorax-like and Grainyhead as mostly correlating with species-specific changes in open chromatin. In conclusion, we show in this study that CRM scores can be used as quantitative traits and that motif discovery approaches can be extended towards more complex models of divergence. ispartof: Molecular Biology and Evolution vol:32 issue:9 pages:2441-55 ispartof: location:United States status: published

Genomic enhancers regulate spatio-temporal gene expression by recruiting specific combinations of transcription factors (TFs). When TFs are bound to active regulatory regions, they displace canonical nucleosomes, making these regions biochemically detectable as nucleosome-depleted regions or accessible/open chromatin. Here we ask whether open chromatin profiling can be used to identify the entire repertoire of active promoters and enhancers underlying tissue-specific gene expression during normal development and oncogenesis in vivo. To this end, we first compare two different approaches to detect open chromatin in vivo using the Drosophila eye primordium as a model system: FAIRE-seq, based on physical separation of open versus closed chromatin; and ATAC-seq, based on preferential integration of a transposon into open chromatin. We find that both methods reproducibly capture the tissue-specific chromatin activity of regulatory regions, including promoters, enhancers, and insulators. Using both techniques, we screened for regulatory regions that become ectopically active during Ras-dependent oncogenesis, and identified 3778 regions that become (over-)activated during tumor development. Next, we applied motif discovery to search for candidate transcription factors that could bind these regions and identified AP-1 and Stat92E as key regulators. We validated the importance of Stat92E in the development of the tumors by introducing a loss of function Stat92E mutant, which was sufficient to rescue the tumor phenotype. Additionally we tested if the predicted Stat92E responsive regulatory regions are genuine, using ectopic induction of JAK/STAT signaling in developing eye discs, and observed that similar chromatin changes indeed occurred. Finally, we determine that these are functionally significant regulatory changes, as nearby target genes are up- or down-regulated. In conclusion, we show that FAIRE-seq and ATAC-seq based open chromatin profiling, combined with motif discovery, is a straightforward approach to identify functional genomic regulatory regions, master regulators, and gene regulatory networks controlling complex in vivo processes., Author Summary The functional expression of all genes is regulated by proteins, namely transcription factors that bind to specific areas of DNA known as regulatory regions. Whereas most DNA in our genome is normally bound by other proteins (histones) and packaged into units called nucleosomes, a specific subset of tissue-specific regulatory regions is responsible for tissue-specific gene expression; these active regions are nucleosome-depleted and bound by transcription factors. We use two techniques to identify these open chromatin regions, in a normal tissue and a RasV12 induced cancer tissue. We discovered a remarkable change in the accessible regulatory landscape between these two tissues, with several thousand regions becoming more accessible in the cancer tissue. We identified two transcription factors known to be involved in cancer (AP-1 and Stat92E) controlling these newly accessible regulatory regions. Finally, we introduced a mutation resulting in Stat92E becoming non-functional in the cancer tissue, which decreased the severity of the tumor. Our study shows that open chromatin profiling can be used to identify complex in vivo processes, and we shed new light on Ras dependent cancer development.

Obesity is a common disorder and related diseases such as diabetes, atherosclerosis, hypertension, cardiovascular disease and cancer are a major cause of mortality and morbidity in Western-type societies. Development of obesity is associated with extensive modifications in adipose tissue involving adipogenesis, angiogenesis and extracellular matrix proteolysis. The fibrinolytic (plasminogen/plasmin) and matrix metalloproteinase (MMP) systems cooperate in these processes. A nutritionally induced obesity model in transgenic mice has been used extensively to study the role of the fibrinolytic and MMP systems in the development of obesity. These studies support a role of both systems in adipogenesis and obesity; the role of specific members of these families, however, remains to be determined.

Steroid receptors are transcription factors that, upon binding to their response elements, regulate the expression of several target genes via direct protein interactions with transcriptional coactivators. For the androgen receptor, additional interactions between the amino- and carboxyl-terminal regions have been reported. The first amino acids of the amino-terminal domain are necessary for this amino/carboxyl-terminal interaction. Deletion of a FQNLF core sequence in this region blunts the interaction, as does a G21E mutation. We investigated the effect of the aforementioned mutations in the context of the full size androgen receptor on a series of selective and nonselective androgen response elements. Strikingly, the FQNLF deletion strongly reduced the androgen receptor capacity to transactivate through nonselective motifs but did not affect its activity on selective elements. Although the G21E mutation strongly impairs the amino/carboxyl-terminal interaction, it does not significantly influence androgen receptor activity on either selective or nonselective elements. Surprisingly, this mutation leads to an increased binding of the amino-terminal domain to the glutamine-rich region of the steroid receptor coactivator-1 of the p160 family. Taken together, these data suggest that the amino-terminal amino acids of the androgen receptor play a key role in determining its transcriptional activity by modulating the interaction with the ligand-binding domain as well as interaction with p160 coactivators.

The androgen receptor has a subset of target DNA sequences, which are not recognized by any other steroid receptors. The androgen selectivity of these sequences was proposed to be the consequence of the ability of the androgen receptor to dimerize on direct repeats of 5′-TGTTCT-3′-like sequences. This is in contrast with the classical non-selective elements consisting of inverted repeats of the 5′-TGTTCT-3′ elements separated by three nucleotides and which are recognized by other steroid receptors in addition to the androgen receptor. We demonstrate that while the DNA-binding domain of the oestrogen receptor is unable to dimerize on direct repeats, dimeric binding can be rescued by replacing the second Zn finger and part of the hinge region by the corresponding fragment of the androgen receptor, but not the glucocorticoid receptor. In this study, we investigate the androgen receptor binding to all natural androgen-selective response elements described so far. We show that a 12-amino acid C-terminal extension of the DNA-binding domain is required for high-affinity binding of the androgen receptor to all these elements. For one androgen-specific low-affinity binding site, the flanking sequences do not contribute to the invitro affinity of the androgen receptor DNA-binding domain. Surprisingly, however, they control the transcriptional activity of the androgen receptor in transient transfection experiments. In conclusion, we give evidence that the alternative DNA-dependent dimerization of the androgen receptor on direct repeats is a general mechanism for androgen specificity in which the second Zn finger and hinge region are involved. In addition, the sequences flanking an androgen-response element can control the activity of the androgen receptor.

A low-molecular-weight receptor tyrosine kinase inhibitor, 1-(6,7-dihydro-5H-benzo(6,7)cyclohepta(1,2-c)pyridazin-3-yl)-N3-((7-pyrrolidin-1-yl)-6,7,8,9-tetrahydro-5H-benzo(7)annulene-2-yl)-1H-1,2,4-triazole-3,5-diamine (R428) with high affinity and selectivity for the growth arrest-specific protein 6 (GAS6) receptor Axl was used to study a potential role of GAS6 signaling in adiposity. In vitro, R428 caused a concentration-dependent inhibition of preadipocyte differentiation into mature adipocytes, as evidenced by reduced lipid uptake. Inhibition of Axl-mediated signaling was confirmed by reduced levels of phospho-Akt activity. In vivo, oral administration of R428 for 5 weeks to mice kept on a high-fat diet resulted in significantly reduced weight gain and subcutaneous and gonadal fat mass. This was associated with marked adipocyte hypotrophy, enhanced macrophage infiltration, and apoptosis. Thus, affecting GAS6 signaling through receptor antagonism using a low-molecular-weight Axl antagonist impairs adipocyte differentiation and reduces adipose tissue development in a murine model of nutritionally induced obesity.

Summary. Objective: To substantiate a potential role of plasminogen activator inhibitor-1 (PAI-1) in adipogenesis, we have studied its effects on in vitro adipocyte differentiation and on in vivo adipose tissue formation. Results: Our in vitro data do not support a functional role of PAI-1, as substantiated by our findings that: (i) inhibition of PAI-1 with a neutralizing antibody did not affect differentiation of 3T3-F442A preadipocytes; (ii) overexpression of murine PAI-1 in 3T3-F442A cells had no effect on differentiation; and (iii) differentiation of PAI-1-deficient murine embryonic fibroblasts into mature adipocytes was comparable to wild-type (WT) cells. Furthermore, our in vivo studies did not reveal an important role for PAI-1, as suggested by our findings that: (i) de novo fat pad formation in NUDE mice following injection of 3T3-F442A cells was not affected by a PAI-1 neutralizing antibody; and (ii) adipose tissue formation following combined injection of Matrigel and basic fibroblast growth factor was comparable in WT and in PAI-1 deficient mice. Conclusion: Taken together, these in vitro and in vivo studies in murine model systems do not support an important functional role of PAI-1 in adipogenesis.

Placental growth factor (PlGF)-deficient (PlGF-/-) and wild-type mice were kept on a standard-fat or high-fat diet for 15 weeks. With the standard-fat diet, the body weights of PlGF-/- and wild-type mice were comparable, whereas the combined weight of subcutaneous and gonadal adipose tissues was lower in PlGF-/- mice (P = 0.02). With the high-fat diet, PlGF-/- mice had a lower body weight (P < 0.05) and less total subcutaneous plus gonadal adipose tissue (P < 0.0001). Blood vessel size was lower in gonadal adipose tissue of PlGF-/- mice with both the standard-fat and high-fat diet (P < 0.05). Blood vessel density, normalized to adipocyte number, was significantly lower in subcutaneous adipose tissue of PlGF-/- mice fed the high-fat diet (P < 0.01). De novo adipose tissue development in nude mice injected with 3T3-F442A preadipocytes was reduced (P < 0.005) by administration of a PlGF-neutralizing antibody. Bone marrow transplantation from wild-type or PlGF-/- mice to wild-type or PlGF-/- recipient mice revealed significantly lower blood vessel density in PlGF-/- recipient mice without an effect on adipose tissue growth. Thus, in murine models of diet-induced obesity, inactivation of PlGF impairs adipose tissue development, at least in part as a result of reduced angiogenesis. ispartof: Diabetes vol:55 issue:10 pages:2698-2704 status: published

Endocrine disruptors pose a growing threat to human and wildlife health. Validated test systems are required to study the mechanisms by which chemicals may interfere with the endocrine system. In order to identify compounds with (anti)androgenic activity, we used several in vitro bioassays, based on different androgen receptor (AR) functions, including AR transactivation and interaction between the amino-terminal domain and the ligand-binding domain. The AR activity assay, based on activation of the transcription of an androgen-responsive reporter gene in the presence of androgen, proved to excel in terms of high fold induction range and low minimal detection limit. The EC50 value, defined as the concentration that leads to half the maximal response and reflecting the potency of the synthetic androgen R1881 (methyltrienolone), was found to be 250 pM, consistent with the high affinity of this ligand to the human AR. A number of environmental samples were tested in the bioassay. The bioassay also could be used to detect antiandrogenic activity, because known AR-antagonists were able to inhibit R1881-mediated transactivation.

The androgen receptor (AR) is a member of the nuclear receptor superfamily. These ligand-activated transcription factors usually contain two activation functions, a ligand-independent activation function 1(AF1) in the divergent N-terminal domain and a ligand-dependent AF2 in the more conserved C-terminal ligand-binding domain. To promote transcription from target promoters, DNA-bound nuclear receptors recruit coactivator proteins that promote transcription by modifying histones within nucleosomes, resulting in altered topology of chromatin to allow access of the basal transcriptional machinery, or stabilising the pre-initiation complex. It is well known that most coactivators interact with AF2 of many nuclear receptors via conserved, helical LxxLL motifs (where L is leucine and x is any amino acid). The AF2 of the AR is very weak, but we were able to demonstrate that its intrinsic ligand-dependent activity is potentiated by steroid receptor coactivator-1 (SRC1) and that this region interacts with coactivators via LxxLL motifs. However, a mutant SRC1 coactivator with no functional LxxLL motifs was still able to potentiate AR activity. We found that SRC1 can also be recruited to (and increase activity of) AF1 of the AR via a conserved, glutamine-rich region. Point mutations within this region abolish SRC1 interaction with AF1 and also abolish or severely impair its ability to potentiate AR activity on all promoters tested. Thus the AR interacts with SRC1 via two different regions and the AF1 interaction is functionally the more important, although the contribution of the two interactions varies in a promoter-dependent fashion. SRC1 then potentiates receptor activity via recruitment of CBP/p300, a histone acetyltranferase. This is important in the context of prostate cancer as SRC1 and other coactivators including CBP are coexpressed with AR in the luminal epithelial cells of the prostate, where over 90% of prostate tumours arise. There is a need for effective second-line prostate cancer therapy aimed at blocking the AR pathway when anti-androgen therapy has failed. Since there is growing evidence that nuclear receptor cofactors may be implicated in the progression of hormone-dependent tumours to hormone-independent states, novel targets could include the interaction of AR with coactivator proteins. We suggest that the N-terminal interaction would be a more specific and effective target in the case of prostate cancer than the LxxLL/AF2 interaction.

The androgen receptor (AR) is a ligand-dependent transcription factor and belongs to the nuclear receptor family. The AR gene contains a long polymorphic CAG repeat, coding for a polyglutamine tract. In the full size AR, the deletion of the polyglutamine tract results in an increase in the transactivation through canonical AREs. However, this effect is clearly dependent on the response elements, since it is not observed on selective elements. In our assays, a deletion of the repeat positively affected the interactions of the ligand-binding domain with the amino-terminal domain as well as the recruitment of the p160 coactivator SRC-1e to the amino-terminal domain of the AR. This is reflected by an enhanced coactivation of the AR by SRC-1e.

The androgen receptor interacts with the p160 coactivators via two surfaces, one in the ligand binding domain and one in the amino-terminal domain. The ligand binding domain interacts with the nuclear receptor signature motifs, whereas the amino-terminal domain has a high affinity for a specific glutamine-rich region in the p160s. We here describe the implication of two conserved motifs in the latter interaction. The amino-terminal domain of the androgen receptor is a very strong activation domain constituent of Tau5, which is mainly active in the absence of the ligand binding domain, and Tau1, which is only active in the presence of the ligand binding domain. Both domains are, however, implicated in the recruitment of the p160s. Mutation analysis of the p160s has shown that the relative contribution of the two recruitment mechanisms via the signature motifs or via the glutamine-rich region depend on the nature of the enhancers tested. We propose, therefore, that the androgen receptor-coactivator complex has several alternative conformations, depending partially on the context of the enhancer.