Exploiting transcription factor binding site clustering to identify cis-regulatory modules involved in pattern formation in the Drosophila genome

Genome access is transcription factor-specific and defined by nucleosome position

Mammalian gene expression is controlled by transcription factors (TFs) that engage sequence motifs in a chromatinized genome, where nucleosomes can restrict DNA access. Yet, how nucleosomes affect individual TFs remains unclear. Here, we measure the ability of over one hundred TF motifs to recruit TFs in a defined chromosomal locus in mouse embryonic stem cells. This identifies a set sufficient to enable the binding of TFs with diverse tissue specificities, functions, and DNA-binding domains. These chromatin-competent factors are further classified when challenged to engage motifs within a highly phased nucleosome. The pluripotency factors OCT4-SOX2 preferentially engage non-nucleosomal and entry-exit motifs, but not nucleosome-internal sites, a preference that also guides binding genome wide. By contrast, factors such as BANP, REST, or CTCF engage throughout, causing nucleosomal displacement. This supports that TFs vary widely in their sensitivity to nucleosomes and that genome access is TF specific and influenced by nucleosome position in the cell.

How Transcription Factor Clusters Shape the Transcriptional Landscape

In eukaryotic cells, gene transcription typically occurs in discrete periods of promoter activity, interspersed with intervals of inactivity. This pattern deviates from simple stochastic events and warrants a closer examination of the molecular interactions that activate the promoter. Recent studies have identified transcription factor (TF) clusters as key precursors to transcriptional bursting. Often, these TF clusters form at chromatin segments that are physically distant from the promoter, making changes in chromatin conformation crucial for promoter-TF cluster interactions. In this review, I explore the formation and constituents of TF clusters, examining how the dynamic interplay between chromatin architecture and TF clustering influences transcriptional bursting. Additionally, I discuss techniques for visualizing TF clusters and provide an outlook on understanding the remaining gaps in this field.

The regulation of liquid-liquid phase separated condensates containing nucleic acids

Liquid-liquid phase separation (LLPS) has been recognized as a universal biological phenomenon. It plays an important role in life activities. LLPS is induced by weak interactions between intrinsically disordered regions or low complex domains. Nucleic acids are widely present in cells, and shown to be closely related to LLPS. Their structure and electronegativity provide the excellent platforms for the formation of phase-separated condensates. In this review, we summarize the interconnected regulation between nucleic acids and LLPS demonstrated in in vivo and in vitro studies. Beside homogeneous and single-phase condensates, complicated and multicompartment LLPS induced by nucleic acids is discussed as well. Recent advances about nucleic-acid-induced LLPS as a new pathogenic mechanism and drug design direction are highlighted, especially virus-mediated disease treatment and prevention.

A universal system for boosting gene expression in eukaryotic cell-lines

We demonstrate a transcriptional regulatory design algorithm that can boost expression in yeast and mammalian cell lines. The system consists of a simplified transcriptional architecture composed of a minimal core promoter and a synthetic upstream regulatory region (sURS) composed of up to three motifs selected from a list of 41 motifs conserved in the eukaryotic lineage. The sURS system was first characterized using an oligo-library containing 189,990 variants. We validate the resultant expression model using a set of 43 unseen sURS designs. The validation sURS experiments indicate that a generic set of grammar rules for boosting and attenuation may exist in yeast cells. Finally, we demonstrate that this generic set of grammar rules functions similarly in mammalian CHO-K1 and HeLa cells. Consequently, our work provides a design algorithm for boosting the expression of promoters used for expressing industrially relevant proteins in yeast and mammalian cell lines.

Less-is-more: selecting transcription factor binding regions informative for motif inference

Numerous statistical methods have emerged for inferring DNA motifs for transcription factors (TFs) from genomic regions. However, the process of selecting informative regions for motif inference remains understudied. Current approaches select regions with strong ChIP-seq signal for a given TF, assuming that such strong signal primarily results from specific interactions between the TF and its motif. Additionally, these selection approaches do not account for non-target motifs, i.e. motifs of other TFs; they presume the occurrence of these non-target motifs infrequent compared to that of the target motif, and thus assume these have minimal interference with the identification of the target. Leveraging extensive ChIP-seq datasets, we introduced the concept of TF signal 'crowdedness', referred to as C-score, for each genomic region. The C-score helps in highlighting TF signals arising from non-specific interactions. Moreover, by considering the C-score (and adjusting for the length of genomic regions), we can effectively mitigate interference of non-target motifs. Using these tools, we find that in many instances, strong ChIP-seq signal stems mainly from non-specific interactions, and the occurrence of non-target motifs significantly impacts the accurate inference of the target motif. Prioritizing genomic regions with reduced crowdedness and short length markedly improves motif inference. This 'less-is-more' effect suggests that ChIP-seq region selection warrants more attention.

EnhancerTracker: Comparing cell-type-specific enhancer activity of DNA sequence triplets via an ensemble of deep convolutional neural networks

Motivation

Transcriptional enhancers - unlike promoters - are unrestrained by distance or strand orientation with respect to their target genes, making their computational identification a challenge. Further, there are insufficient numbers of confirmed enhancers for many cell types, preventing robust training of machine-learning-based models for enhancer prediction for such cell types.

Results

We present EnhancerTracker , a novel tool that leverages an ensemble of deep separable convolutional neural networks to identify cell-type-specific enhancers with the need of only two confirmed enhancers. EnhancerTracker is trained, validated, and tested on 52,789 putative enhancers obtained from the FANTOM5 Project and control sequences derived from the human genome. Unlike available tools, which accept one sequence at a time, the input to our tool is three sequences; the first two are enhancers active in the same cell type. EnhancerTracker outputs 1 if the third sequence is an enhancer active in the same cell type(s) where the first two enhancers are active. It outputs 0 otherwise. On a held-out set (15%), EnhancerTracker achieved an accuracy of 64%, a specificity of 93%, a recall of 35%, a precision of 84%, and an F1 score of 49%.

Availability and implementation

https://github.com/BioinformaticsToolsmith/EnhancerTracker.

Contact

hani.girgis@tamuk.edu.

The chromatin signatures of enhancers and their dynamic regulation

Enhancers are cis-regulatory elements that can stimulate gene expression from distance, and drive precise spatiotemporal gene expression profiles during development. Functional enhancers display specific features including an open chromatin conformation, Histone H3 lysine 27 acetylation, Histone H3 lysine 4 mono-methylation enrichment, and enhancer RNAs production. These features are modified upon developmental cues which impacts their activity. In this review, we describe the current state of knowledge about enhancer functions and the diverse chromatin signatures found on enhancers. We also discuss the dynamic changes of enhancer chromatin signatures, and their impact on lineage specific gene expression profiles, during development or cellular differentiation.

In-silico identification and comparison of transcription factor binding sites cluster in anterior-posterior patterning genes in Drosophila melanogaster and Tribolium castaneum

The cis-regulatory data that help in transcriptional regulation is arranged into modular pieces of a few hundred base pairs called CRMs (cis-regulatory modules) and numerous binding sites for multiple transcription factors are prominent characteristics of these cis-regulatory modules. The present study was designed to localize transcription factor binding site (TFBS) clusters on twelve Anterior-posterior (A-P) genes in Tribolium castaneum and compare them to their orthologous gene enhancers in Drosophila melanogaster. Out of the twelve A-P patterning genes, six were gap genes (Kruppel, Knirps, Tailless, Hunchback, Giant, and Caudal) and six were pair rule genes (Hairy, Runt, Even-skipped, Fushi-tarazu, Paired, and Odd-skipped). The genes along with 20 kb upstream and downstream regions were scanned for TFBS clusters using the Motif Cluster Alignment Search Tool (MCAST), a bioinformatics tool that looks for set of nucleotide sequences for statistically significant clusters of non-overlapping occurrence of a given set of motifs. The motifs used in the current study were Hunchback, Caudal, Giant, Kruppel, Knirps, and Even-skipped. The results of the MCAST analysis revealed the maximum number of TFBS for Hunchback, Knirps, Caudal, and Kruppel in both D. melanogaster and T. castaneum, while Bicoid TFBS clusters were found only in D. melanogaster. The size of all the predicted TFBS clusters was less than 1kb in both insect species. These sequences revealed more transversional sites (Tv) than transitional sites (Ti) and the average Ti/Tv ratio was 0.75.

Decoding enhancer complexity with machine learning and high-throughput discovery

Enhancers are genomic DNA elements controlling spatiotemporal gene expression. Their flexible organization and functional redundancies make deciphering their sequence-function relationships challenging. This article provides an overview of the current understanding of enhancer organization and evolution, with an emphasis on factors that influence these relationships. Technological advancements, particularly in machine learning and synthetic biology, are discussed in light of how they provide new ways to understand this complexity. Exciting opportunities lie ahead as we continue to unravel the intricacies of enhancer function.

Mechanisms of robustness in gene regulatory networks involved in neural development

The functions of living organisms are affected by different kinds of perturbation, both internal and external, which in many cases have functional effects and phenotypic impact. The effects of these perturbations become particularly relevant for multicellular organisms with complex body patterns and cell type heterogeneity, where transcriptional programs controlled by gene regulatory networks determine, for example, the cell fate during embryonic development. Therefore, an essential aspect of development in these organisms is the ability to maintain the functionality of their genetic developmental programs even in the presence of genetic variation, changing environmental conditions and biochemical noise, a property commonly termed robustness. We discuss the implication of different molecular mechanisms of robustness involved in neurodevelopment, which is characterized by the interplay of many developmental programs at a molecular, cellular and systemic level. We specifically focus on processes affecting the function of gene regulatory networks, encompassing transcriptional regulatory elements and post-transcriptional processes such as miRNA-based regulation, but also higher order regulatory organization, such as gene network topology. We also present cases where impairment of robustness mechanisms can be associated with neurodevelopmental disorders, as well as reasons why understanding these mechanisms should represent an important part of the study of gene regulatory networks driving neural development.

CFA: An explainable deep learning model for annotating the transcriptional roles of cis-regulatory modules based on epigenetic codes

Metazoa gene expression is controlled by modular DNA segments called cis-regulatory modules (CRMs). CRMs can convey promoter/enhancer/insulator roles, generating additional regulation layers in transcription. Experiments for understanding CRM roles are low-throughput and costly. Large-scale CRM function investigation still depends on computational methods. However, existing in silico tools only recognize enhancers or promoters exclusively, thus accumulating errors when considering CRM promoter/enhancer/insulator roles altogether. Currently, no algorithm can concurrently consider these CRM roles. In this research, we developed the CRM Function Annotator (CFA) model. CFA provides complete CRM transcriptional role labeling based on epigenetic profiling interpretation. We demonstrated that CFA achieves high performance (test macro auROC/auPRC = 94.1%/90.3%) and outperforms existing tools in promoter/enhancer/insulator identification. CFA is also inspected to recognize explainable epigenetic codes consistent with previous findings when labeling CRM roles. By considering the higher-order combinations of the epigenetic codes, CFA significantly reduces false-positive rates in CRM transcriptional role annotation. CFA is available at https://github.com/cobisLab/CFA/.

Transcriptional enhancers at 40: evolution of a viral DNA element to nuclear architectural structures

Gene regulation by transcriptional enhancers is the dominant mechanism driving cell type- and signal-specific transcriptional diversity in metazoans. However, over four decades since the original discovery, how enhancers operate in the nuclear space remains largely enigmatic. Recent multidisciplinary efforts combining real-time imaging, genome sequencing, and biophysical strategies provide insightful but conflicting models of enhancer-mediated gene control. Here, we review the discovery and progress in enhancer biology, emphasizing the recent findings that acutely activated enhancers assemble regulatory machinery as mesoscale architectural structures with distinct physical properties. These findings help formulate novel models that explain several mysterious features of the assembly of transcriptional enhancers and the mechanisms of spatial control of gene expression.

Concentration-dependent transcriptional switching through a collective action of cis-elements

Gene expression specificity of homeobox transcription factors has remained paradoxical. WUSCHEL activates and represses CLAVATA3 transcription at lower and higher concentrations, respectively. We use computational modeling and experimental analysis to investigate the properties of the cis-regulatory module. We find that intrinsically each cis-element can only activate CLAVATA3 at a higher WUSCHEL concentration. However, together, they repress CLAVATA3 at higher WUSCHEL and activate only at lower WUSCHEL, showing that the concentration-dependent interactions among cis-elements regulate both activation and repression. Biochemical experiments show that two adjacent functional cis-elements bind WUSCHEL with higher affinity and dimerize at relatively lower levels. Moreover, increasing the distance between cis-elements prolongs WUSCHEL monomer binding window, resulting in higher CLAVATA3 activation. Our work showing a constellation of optimally spaced cis-elements of defined affinities determining activation and repression thresholds in regulating CLAVATA3 transcription provides a previously unknown mechanism of cofactor-independent regulation of transcription factor binding in mediating gene expression specificity.

Generating specificity in genome regulation through transcription factor sensitivity to chromatin

Cell type-specific gene expression relies on transcription factors (TFs) binding DNA sequence motifs embedded in chromatin. Understanding how motifs are accessed in chromatin is crucial to comprehend differential transcriptional responses and the phenotypic impact of sequence variation. Chromatin obstacles to TF binding range from DNA methylation to restriction of DNA access by nucleosomes depending on their position, composition and modification. In vivo and in vitro approaches now enable the study of TF binding in chromatin at unprecedented resolution. Emerging insights suggest that TFs vary in their ability to navigate chromatin states. However, it remains challenging to link binding and transcriptional outcomes to molecular characteristics of TFs or the local chromatin substrate. Here, we discuss our current understanding of how TFs access DNA in chromatin and novel techniques and directions towards a better understanding of this critical step in genome regulation.

Molecular competition can shape enhancer activity in the Drosophila embryo

Transgenic reporters allow the measurement of regulatory DNA activity in vivo and consequently have long been useful tools for studying enhancers. Despite their utility, few studies have investigated the effects these reporters may have on the expression of other genes. Understanding these effects is required to accurately interpret reporter data and characterize gene regulatory mechanisms. By measuring the expression of Kruppel (Kr) enhancer reporters in live Drosophila embryos, we find reporters inhibit one another's expression and that of a nearby endogenous gene. Using synthetic transcription factor (TF) binding site arrays, we present evidence that competition for TFs is partially responsible for the observed transcriptional inhibition. We develop a simple thermodynamic model that predicts competition of the measured magnitude specifically when TF binding is restricted to distinct nuclear subregions. Our findings underline an unexpected role of the non-homogenous nature of the nucleus in regulating gene expression.

Computational biology approaches for mapping transcriptional regulatory networks

Transcriptional Regulatory Networks (TRNs) are mainly responsible for the cell-type- or cell-state-specific expression of gene sets from the same DNA sequence. However, so far there are no precise maps of TRNs available for each cell-type or cell-state, and no ideal tool to map those networks clearly and in full from biological samples. In this review, major approaches and tools to map TRNs from high-throughput data are presented, depending on the type of methods or data used to infer them, and their advantages and limitations are discussed. After summarizing the main principles defining the topology and structure–function relationships in TRNs, an overview of the extensive work done to map TRNs from bulk transcriptomic data will be presented by type of methodological approach. Most recent modellings of TRNs using other types of molecular data or integrating different data types, including single-cell RNA-sequencing and chromatin information, will then be discussed, before briefly concluding with improvements expected to come in the field.

Cis-regulatory variants affect gene expression dynamics in yeast

Evolution of cis-regulatory sequences depends on how they affect gene expression and motivates both the identification and prediction of cis-regulatory variants responsible for expression differences within and between species. While much progress has been made in relating cis-regulatory variants to expression levels, the timing of gene activation and repression may also be important to the evolution of cis-regulatory sequences. We investigated allele-specific expression (ASE) dynamics within and between Saccharomyces species during the diauxic shift and found appreciable cis-acting variation in gene expression dynamics. Within-species ASE is associated with intergenic variants, and ASE dynamics are more strongly associated with insertions and deletions than ASE levels. To refine these associations, we used a high-throughput reporter assay to test promoter regions and individual variants. Within the subset of regions that recapitulated endogenous expression, we identified and characterized cis-regulatory variants that affect expression dynamics. Between species, chimeric promoter regions generate novel patterns and indicate constraints on the evolution of gene expression dynamics. We conclude that changes in cis-regulatory sequences can tune gene expression dynamics and that the interplay between expression dynamics and other aspects of expression is relevant to the evolution of cis-regulatory sequences.

Uncovering the cellular capacity for intensive and specific feedback self-control of the argonautes and MicroRNA targeting activity

The miRNA pathway has three segments—biogenesis, targeting and downstream regulatory effectors. We aimed to better understand their cellular control by exploring the miRNA-mRNA-targeting relationships. We first used human evolutionarily conserved sites. Strikingly, AGOs 1–3 are all among the top 14 mRNAs with the highest miRNA site counts, along with ANKRD52, the phosphatase regulatory subunit of the recently identified AGO phosphorylation cycle; and the AGO phosphorylation cycle mRNAs share much more than expected miRNA sites. The mRNAs for TNRC6, which acts with AGOs to channel miRNA-mediated regulatory actions onto specific mRNAs, are also heavily miRNA-targeted. In contrast, upstream miRNA biogenesis mRNAs are not, and neither are downstream regulatory effectors. In short, binding site enrichment in miRNA targeting machinery mRNAs, but neither upstream biogenesis nor downstream effector mRNAs, was observed, endowing a cellular capacity for intensive and specific feedback control of the targeting activity. The pattern was confirmed with experimentally determined miRNA-mRNA target relationships. Moreover, genetic experiments demonstrated cellular utilization of this capacity. Thus, we uncovered a capacity for intensive, and specific, feedback-regulation of miRNA targeting activity directly by miRNAs themselves, i.e. segment-specific feedback auto-regulation of miRNA pathway, complementing miRNAs pairing with transcription factors to form hybrid feedback-loop.

Transcriptional Enhancers in Drosophila

Key discoveries in Drosophila have shaped our understanding of cellular "enhancers." With a special focus on the fly, this chapter surveys properties of these adaptable cis-regulatory elements, whose actions are critical for the complex spatial/temporal transcriptional regulation of gene expression in metazoa. The powerful combination of genetics, molecular biology, and genomics available in Drosophila has provided an arena in which the developmental role of enhancers can be explored. Enhancers are characterized by diverse low- or high-throughput assays, which are challenging to interpret, as not all of these methods of identifying enhancers produce concordant results. As a model metazoan, the fly offers important advantages to comprehensive analysis of the central functions that enhancers play in gene expression, and their critical role in mediating the production of phenotypes from genotype and environmental inputs. A major challenge moving forward will be obtaining a quantitative understanding of how these cis-regulatory elements operate in development and disease.

How to study enhancers in non-traditional insect models

Transcriptional enhancers are central to the function and evolution of genes and gene regulation. At the organismal level, enhancers play a crucial role in coordinating tissue- and context-dependent gene expression. At the population level, changes in enhancers are thought to be a major driving force that facilitates evolution of diverse traits. An amazing array of diverse traits seen in insect morphology, physiology and behavior has been the subject of research for centuries. Although enhancer studies in insects outside of Drosophila have been limited, recent advances in functional genomic approaches have begun to make such studies possible in an increasing selection of insect species. Here, instead of comprehensively reviewing currently available technologies for enhancer studies in established model organisms such as Drosophila, we focus on a subset of computational and experimental approaches that are likely applicable to non-Drosophila insects, and discuss the pros and cons of each approach. We discuss the importance of validating enhancer function and evaluate several possible validation methods, such as reporter assays and genome editing. Key points and potential pitfalls when establishing a reporter assay system in non-traditional insect models are also discussed. We close with a discussion of how to advance enhancer studies in insects, both by improving computational approaches and by expanding the genetic toolbox in various insects. Through these discussions, this Review provides a conceptual framework for studying the function and evolution of enhancers in non-traditional insect models.

Enhancer Features that Drive Formation of Transcriptional Condensates

Enhancers are DNA elements that are bound by transcription factors (TFs), which recruit coactivators and the transcriptional machinery to genes. Phase-separated condensates of TFs and coactivators have been implicated in assembling the transcription machinery at particular enhancers, yet the role of DNA sequence in this process has not been explored. We show that DNA sequences encoding TF binding site number, density, and affinity above sharply defined thresholds drive condensation of TFs and coactivators. A combination of specific structured (TF-DNA) and weak multivalent (TF-coactivator) interactions allows for condensates to form at particular genomic loci determined by the DNA sequence and the complement of expressed TFs. DNA features found to drive condensation promote enhancer activity and transcription in cells. Our study provides a framework to understand how the genome can scaffold transcriptional condensates at specific loci and how the universal phenomenon of phase separation might regulate this process.

Spatiotemporal control of gene expression boundaries using a feedforward loop

BACKGROUND

A feedforward loop (FFL) is commonly observed in several biological networks. The FFL network motif has been mostly studied with respect to variation of the input signal in time, with only a few studies of FFL activity in a spatially distributed system such as morphogen-mediated tissue patterning. However, most morphogen gradients also evolve in time.

RESULTS

We studied the spatiotemporal behavior of a coherent FFL in two contexts: (a) a generic, oscillating morphogen gradient and (b) the dorsal-ventral patterning of the early Drosophila embryo by a gradient of the NF-κB homolog dorsal with its early target Twist. In both models, we found features in the dynamics of the intermediate node-phase difference and noise filtering-that were largely independent of the parameterization of the models, and thus were functions of the structure of the FFL itself. In the dorsal gradient model, we also found that proper target gene expression was not possible without including the effect of maternal pioneer factor Zelda.

CONCLUSIONS

An FFL buffers fluctuation to changes in the morphogen signal ensuring stable gene expression boundaries.

Formation, interpretation, and regulation of the Drosophila Dorsal/NF-κB gradient

The morphogen gradient of the transcription factor Dorsal in the early Drosophila embryo has become one of the most widely studied tissue patterning systems. Dorsal is a Drosophila homolog of mammalian NF-κB and patterns the dorsal-ventral axis of the blastoderm embryo into several tissue types by spatially regulating upwards of 100 zygotic genes. Recent studies using fluorescence microscopy and live imaging have quantified the Dorsal gradient and its target genes, which has paved the way for mechanistic modeling of the gradient. In this review, we describe the mechanisms behind the initiation of the Dorsal gradient and its regulation of target genes. The main focus of the review is a discussion of quantitative and computational studies of the Dl gradient system, including regulation of the Dl gradient. We conclude with a discussion of potential future directions.

Exploring a Drosophila Transcription Factor Interaction Network to Identify Cis-Regulatory Modules

Multiple transcription factors (TFs) bind to specific sites in the genome and interact among themselves to form the cis-regulatory modules (CRMs). They are essential in modulating the expression of genes, and it is important to study this interplay to understand gene regulation. In the present study, we integrated experimentally identified TF binding sites collected from published studies with computationally predicted TF binding sites to identify Drosophila CRMs. Along with the detection of the previously known CRMs, this approach identified novel protein combinations. We determined high-occupancy target sites, where a large number of TFs bind. Investigating these sites revealed that Giant, Dichaete, and Knirp are highly enriched in these locations. A common TAG team motif was observed at these sites, which might play a role in recruiting other TFs. While comparing the binding sites at distal and proximal promoters, we found that certain regulatory TFs, such as Zelda, were highly enriched in enhancers. Our study has shown that, from the information available concerning the TF binding sites, the real CRMs could be predicted accurately and efficiently. Although we only may claim co-occurrence of these proteins in this study, it may actually point to their interaction (as known interaction proteins typically co-occur together). Such an integrative approach can, therefore, help us to provide a better understanding of the interplay among the factors, even though further experimental verification is required.

An in silico analysis of robust but fragile gene regulation links enhancer length to robustness

Organisms must ensure that expression of genes is directed to the appropriate tissues at the correct times, while simultaneously ensuring that these gene regulatory systems are robust to perturbation. This idea is captured by a mathematical concept called r-robustness, which says that a system is robust to a perturbation in up to r - 1 randomly chosen parameters. r-robustness implies that the biological system has a small number of sensitive parameters and that this number can be used as a robustness measure. In this work we use this idea to investigate the robustness of gene regulation using a sequence level model of the Drosophila melanogaster gene even-skipped. We consider robustness with respect to mutations of the enhancer sequence and with respect to changes of the transcription factor concentrations. We find that gene regulation is r-robust with respect to mutations in the enhancer sequence and identify a number of sensitive nucleotides. In both natural and in silico predicted enhancers, the number of nucleotides that are sensitive to mutation correlates negatively with the length of the sequence, meaning that longer sequences are more robust. The exact degree of robustness obtained is dependent not only on DNA sequence, but also on the local concentration of regulatory factors. We find that gene regulation can be remarkably sensitive to changes in transcription factor concentrations at the boundaries of expression features, while it is robust to perturbation elsewhere.

Analysis of Genetic Variation Indicates DNA Shape Involvement in Purifying Selection

Noncoding DNA sequences, which play various roles in gene expression and regulation, are under evolutionary pressure. Gene regulation requires specific protein–DNA binding events, and our previous studies showed that both DNA sequence and shape readout are employed by transcription factors (TFs) to achieve DNA binding specificity. By investigating the shape-disrupting properties of single nucleotide polymorphisms (SNPs) in human regulatory regions, we established a link between disruptive local DNA shape changes and loss of specific TF binding. Furthermore, we described cases where disease-associated SNPs may alter TF binding through DNA shape changes. This link led us to hypothesize that local DNA shape within and around TF binding sites is under selection pressure. To verify this hypothesis, we analyzed SNP data derived from 216 natural strains of Drosophila melanogaster. Comparing SNPs located in functional and nonfunctional regions within experimentally validated cis-regulatory modules (CRMs) from D. melanogaster that are active in the blastoderm stage of development, we found that SNPs within functional regions tended to cause smaller DNA shape variations. Furthermore, SNPs with higher minor allele frequency were more likely to result in smaller DNA shape variations. The same analysis based on a large number of SNPs in putative CRMs of the D. melanogaster genome derived from DNase I accessibility data confirmed these observations. Taken together, our results indicate that common SNPs in functional regions tend to maintain DNA shape, whereas shape-disrupting SNPs are more likely to be eliminated through purifying selection.

Inhibition of the Hypoxia-Inducible Factor 1α-Induced Cardiospecific HERNA1 Enhance-Templated RNA Protects From Heart Disease

Supplemental Digital Content is available in the text.

Background:

Enhancers are genomic regulatory elements conferring spatiotemporal and signal-dependent control of gene expression. Recent evidence suggests that enhancers can generate noncoding enhancer RNAs, but their (patho)biological functions remain largely elusive.

Methods:

We performed chromatin immunoprecipitation–coupled sequencing of histone marks combined with RNA sequencing of left ventricular biopsies from experimental and genetic mouse models of human cardiac hypertrophy to identify transcripts revealing enhancer localization, conservation with the human genome, and hypoxia-inducible factor 1α dependence. The most promising candidate, hypoxia-inducible enhancer RNA (HERNA)1, was further examined by investigating its capacity to modulate neighboring coding gene expression by binding to their gene promoters by using chromatin isolation by RNA purification and λN–BoxB tethering–based reporter assays. The role of HERNA1 and its neighboring genes for pathological stress–induced growth and contractile dysfunction, and the therapeutic potential of HERNA1 inhibition was studied in gapmer-mediated loss-of-function studies in vitro using human induced pluripotent stem cell–derived cardiomyocytes and various in vivo models of human pathological cardiac hypertrophy.

Results:

HERNA1 is robustly induced on pathological stress. Production of HERNA1 is initiated by direct hypoxia-inducible factor 1α binding to a hypoxia-response element in the histoneH3-lysine27acetylation marks–enriched promoter of the enhancer and confers hypoxia responsiveness to nearby genes including synaptotagmin XVII, a member of the family of membrane-trafficking and Ca²⁺-sensing proteins and SMG1, encoding a phosphatidylinositol 3-kinase–related kinase. Consequently, a substrate of SMG1, ATP-dependent RNA helicase upframeshift 1, is hyperphoshorylated in a HERNA1- and SMG1-dependent manner. In vitro and in vivo inactivation of SMG1 and SYT17 revealed overlapping and distinct roles in modulating cardiac hypertrophy. Finally, in vivo administration of antisense oligonucleotides targeting HERNA1 protected mice from stress-induced pathological hypertrophy. The inhibition of HERNA1 postdisease development reversed left ventricular growth and dysfunction, resulting in increased overall survival.

Conclusions:

HERNA1 is a novel heart-specific noncoding RNA with key regulatory functions in modulating the growth, metabolic, and contractile gene program in disease, and reveals a molecular target amenable to therapeutic exploitation.

Neuroblast-specific open chromatin allows the temporal transcription factor, Hunchback, to bind neuroblast-specific loci

Spatial and temporal cues are required to specify neuronal diversity, but how these cues are integrated in neural progenitors remains unknown. Drosophila progenitors (neuroblasts) are a good model: they are individually identifiable with relevant spatial and temporal transcription factors known. Here we test whether spatial/temporal factors act independently or sequentially in neuroblasts. We used Targeted DamID to identify genomic binding sites of the Hunchback temporal factor in two neuroblasts (NB5-6 and NB7-4) that make different progeny. Hunchback targets were different in each neuroblast, ruling out the independent specification model. Moreover, each neuroblast had distinct open chromatin domains, which correlated with differential Hb-bound loci in each neuroblast. Importantly, the Gsb/Pax3 spatial factor, expressed in NB5-6 but not NB7-4, had genomic binding sites correlated with open chromatin in NB5-6, but not NB7-4. Our data support a model in which early-acting spatial factors like Gsb establish neuroblast-specific open chromatin domains, leading to neuroblast-specific temporal factor binding and the production of different neurons in each neuroblast lineage.

The human brain is considered to be the most complicated object in the universe, but it only takes a handful of stem cells to make one. The process depends on two types of information: signals separated across space and time. Spatial cues tell a stem cell what type of cell it is going to be, while temporal cues work as molecular clocks to generate a sequence of different neurons over time. Together, these cues generate the large array of cell types in the nervous system.

Each stem cell occupies its own space in the developing body and receives its own spatial cues, but they all follow the same timeline. For example, proteins called transcription factors act as molecular clocks and interact with specific genes, telling the cell when to turn them on or off. The same series of transcription factors operates in different stem cells, but they have different effects. So far, it has been unclear whether spatial and temporal signals work independently or sequentially to generate new cell types.

To find out, Sen et al. studied two distinct, developing stem cells in fruit flies, which receive different spatial signals. Transcription factors only work if they are able to get to their target genes. Cells can open or close access to different genes by changing the structure of the chromatin wrapping that surrounds the genes. In the experiments, a marker was used to reveal the areas of open chromatin in each of the cells. Another marker was used to track the transcription factors. The results showed that the areas of open chromatin varied between stem cells. Moreover, although both cells used the same transcription factor called Hunchback, it targeted different genes in each stem cell. This was due to changes in the chromatin wrapping: Hunchback only acted in areas where the chromatin was open. This suggests that the spatial cues first sculpt the chromatin, making some genes easier to get to than others. Then, the same transcription factors go to the accessible gene, which will differ from one stem cell to another.

These findings help us to understand how different types of brain cells develop, which may also aid us in finding a way how to engineer specific cell types. If we could turn stem cells into different types of brain cells, it might help us to treat brain diseases. This may involve giving the right spatial signal before starting the temporal cues.

HACER: an atlas of human active enhancers to interpret regulatory variants

Recent studies have shown that disease-susceptibility variants frequently lie in cell-type-specific enhancer elements. To identify, interpret, and prioritize such risk variants, we must identify the enhancers active in disease-relevant cell types, their upstream transcription factor (TF) binding, and their downstream target genes. To address this need, we built HACER (http://bioinfo.vanderbilt.edu/AE/HACER/), an atlas of Human ACtive Enhancers to interpret Regulatory variants. The HACER atlas catalogues and annotates in-vivo transcribed cell-type-specific enhancers, as well as placing enhancers within transcriptional regulatory networks by integrating ENCODE TF ChIP-Seq and predicted/validated chromatin interaction data. We demonstrate the utility of HACER in (i) offering a mechanistic hypothesis to explain the association of SNP rs614367 with ER-positive breast cancer risk, (ii) exploring tumor-specific enhancers in selective MYC dysregulation and (iii) prioritizing/annotating non-coding regulatory regions targeting CCND1. HACER provides a valuable resource for studies of GWAS, non-coding variants, and enhancer-mediated regulation.

Transcription factor binding site clusters identify target genes with similar tissue-wide expression and buffer against mutations

Background: The distribution and composition of cis-regulatory modules composed of transcription factor (TF) binding site (TFBS) clusters in promoters substantially determine gene expression patterns and TF targets. TF knockdown experiments have revealed that TF binding profiles and gene expression levels are correlated. We use TFBS features within accessible promoter intervals to predict genes with similar tissue-wide expression patterns and TF targets using Machine Learning (ML). Methods: Bray-Curtis Similarity was used to identify genes with correlated expression patterns across 53 tissues. TF targets from knockdown experiments were also analyzed by this approach to set up the ML framework. TFBSs were selected within DNase I-accessible intervals of corresponding promoter sequences using information theory-based position weight matrices (iPWMs) for each TF. Features from information-dense clusters of TFBSs were input to ML classifiers which predict these gene targets along with their accuracy, specificity and sensitivity. Mutations in TFBSs were analyzed in silico to examine their impact on TFBS clustering and predict changes in gene regulation. Results: The glucocorticoid receptor gene ( NR3C1), whose regulation has been extensively studied, was selected to test this approach. SLC25A32 and TANK exhibited the most similar expression patterns to NR3C1. A Decision Tree classifier exhibited the best performance in detecting such genes, based on Area Under the Receiver Operating Characteristic curve (ROC). TF target gene prediction was confirmed using siRNA knockdown, which was more accurate than CRISPR/CAS9 inactivation. TFBS mutation analyses revealed that accurate target gene prediction required  at least 1  information-dense TFBS cluster.  Conclusions: ML based on TFBS information density, organization, and chromatin accessibility accurately identifies gene targets with comparable tissue-wide expression patterns. Multiple information-dense TFBS clusters in promoters appear to protect promoters from effects of deleterious binding site mutations in a single TFBS that would otherwise alter regulation of these genes.

De novo recruitment of Polycomb-group proteins in Drosophila embryos

Polycomb-group (PcG)-mediated transcriptional repression of target genes can be delineated into two phases. First, following initial repression of target genes by gene-specific transcription factors, PcG proteins recognize the repressed state and assume control of the genes' repression. Second, once the silenced state is established, PcG proteins may maintain repression through an indefinite number of cell cycles. Little is understood about how PcG proteins initially recognize the repressed state of target genes and the steps leading to de novo establishment of PcG-mediated repression. We describe a genetic system in which a Drosophila PcG target gene, giant (gt), is ubiquitously repressed during early embryogenesis by a maternally expressed transcription factor, and show the temporal recruitment of components of three PcG protein complexes: PhoRC, PRC1 and PRC2. We show that de novo PcG recruitment follows a temporal hierarchy in which PhoRC stably localizes at the target gene at least 1 h before stable recruitment of PRC2 and concurrent trimethylation of histone H3 at lysine 27 (H3K27me3). The presence of PRC2 and increased levels of H3K27me3 are found to precede stable binding by PRC1.

Canonical and single-cell Hi-C reveal distinct chromatin interaction sub-networks of mammalian transcription factors

Background

Transcription factor (TF) binding to regulatory DNA sites is a key determinant of cell identity within multi-cellular organisms and has been studied extensively in relation to site affinity and chromatin modifications. There has been a strong focus on the inference of TF-gene regulatory networks and TF-TF physical interaction networks. Here, we present a third type of TF network, the spatial network of co-localized TF binding sites within the three-dimensional genome.

Results

Using published canonical Hi-C data and single-cell genome structures, we assess the spatial proximity of a genome-wide array of potential TF-TF co-localizations in human and mouse cell lines. For individual TFs, the abundance of occupied binding sites shows a positive correspondence with their clustering in three dimensions, and this is especially apparent for weak TF binding sites and at enhancer regions. An analysis between different TF proteins identifies significantly proximal pairs, which are enriched in reported physical interactions. Furthermore, clustering of different TFs based on proximity enrichment identifies two partially segregated co-localization sub-networks, involving different TFs in different cell types. Using data from both human lymphoblastoid cells and mouse embryonic stem cells, we find that these sub-networks are enriched within, but not exclusive to, different chromosome sub-compartments that have been identified previously in Hi-C data.

Conclusions

This suggests that the association of TFs within spatial networks is closely coupled to gene regulatory networks. This applies to both differentiated and undifferentiated cells and is a potential causal link between lineage-specific TF binding and chromosome sub-compartment segregation.

Electronic supplementary material

The online version of this article (10.1186/s13059-018-1558-2) contains supplementary material, which is available to authorized users.

Thirty biologically interpretable clusters of transcription factors distinguish cancer type

Background

Transcription factors are essential regulators of gene expression and play critical roles in development, differentiation, and in many cancers. To carry out their regulatory programs, they must cooperate in networks and bind simultaneously to sites in promoter or enhancer regions of genes. We hypothesize that the mRNA co-expression patterns of transcription factors can be used both to learn how they cooperate in networks and to distinguish between cancer types.

Results

We recently developed a new algorithm, Thresher, that combines principal component analysis, outlier filtering, and von Mises-Fisher mixture models to cluster genes (in this case, transcription factors) based on expression, determining the optimal number of clusters in the process. We applied Thresher to the RNA-Seq expression data of 486 transcription factors from more than 10,000 samples of 33 kinds of cancer studied in The Cancer Genome Atlas (TCGA). We found that 30 clusters of transcription factors from a 29-dimensional principal component space were able to distinguish between most cancer types, and could separate tumor samples from normal controls. Moreover, each cluster of transcription factors could be either (i) linked to a tissue-specific expression pattern or (ii) associated with a fundamental biological process such as cell cycle, angiogenesis, apoptosis, or cytoskeleton. Clusters of the second type were more likely also to be associated with embryonically lethal mouse phenotypes.

Conclusions

Using our approach, we have shown that the mRNA expression patterns of transcription factors contain most of the information needed to distinguish different cancer types. The Thresher method is capable of discovering biologically interpretable clusters of genes. It can potentially be applied to other gene sets, such as signaling pathways, to decompose them into simpler, yet biologically meaningful, components.

Electronic supplementary material

The online version of this article (10.1186/s12864-018-5093-z) contains supplementary material, which is available to authorized users.

Nascent RNA sequencing analysis provides insights into enhancer-mediated gene regulation

BACKGROUND

Enhancers are distal cis-regulatory elements that control gene expression. Despite an increasing appreciation of the importance of enhancers in cellular function and disease, our knowledge of enhancer-regulated transcription is very limited. Nascent RNA sequencing technologies, such as global nuclear run-on sequencing (GRO-seq) and precision run-on sequencing (PRO-seq), not only provide a direct and reliable measurement of enhancer activity, but also allow for quantifying transcription of enhancers and target genes simultaneously, making these technologies extremely useful for exploring enhancer-mediated regulation.

RESULTS

Nascent RNA sequencing analysis (NRSA) provides a comprehensive view of enhancer-mediated gene regulation. NRSA not only outperforms existing methods for enhancer identification, but also enables annotation and quantification of active enhancers, and prediction of their target genes. Furthermore, NRSA identifies functionally important enhancers by integrating 1) nascent transcriptional changes in enhancers and their target genes and 2) binding profiles from regulator(s) of interest. Applied to wildtype and histone deacetylase 3 (Hdac3) knockout mouse livers, NRSA showed that HDAC3 regulates RNA polymerase recruitment through both proximal (promoter) and distal (enhancer) regulatory elements. Integrating ChIP-seq with PRO-seq data, NRSA prioritized enhancers based on their potential contribution to mediating HDAC3 regulation.

CONCLUSIONS

NRSA will greatly facilitate the usage of nascent RNA sequencing techniques and accelerate the study of enhancer-mediated regulation.

Species-Specific Changes in a Primate Transcription Factor Network Provide Insights into the Molecular Evolution of the Primate Prefrontal Cortex

The human prefrontal cortex (PFC) differs from that of other primates with respect to size, histology, and functional abilities. Here, we analyzed genome-wide expression data of humans, chimpanzees, and rhesus macaques to discover evolutionary changes in transcription factor (TF) networks that may underlie these phenotypic differences. We determined the co-expression networks of all TFs with species-specific expression including their potential target genes and interaction partners in the PFC of all three species. Integrating these networks allowed us inferring an ancestral network for all three species. This ancestral network as well as the networks for each species is enriched for genes involved in forebrain development, axonogenesis, and synaptic transmission. Our analysis allows us to directly compare the networks of each species to determine which links have been gained or lost during evolution. Interestingly, we detected that most links were gained on the human lineage, indicating increase TF cooperativity in humans. By comparing network changes between different tissues, we discovered that in brain tissues, but not in the other tissues, the human networks always had the highest connectivity. To pinpoint molecular changes underlying species-specific phenotypes, we analyzed the sub-networks of TFs derived only from genes with species-specific expression changes in the PFC. These sub-networks differed significantly in structure and function between the human and chimpanzee. For example, the human-specific sub-network is enriched for TFs implicated in cognitive disorders and for genes involved in synaptic plasticity and cognitive functions. Our results suggest evolutionary changes in TF networks that might have shaped morphological and functional differences between primate brains, in particular in the human PFC.

Iterative random forests to discover predictive and stable high-order interactions

We developed a predictive, stable, and interpretable tool: the iterative random forest algorithm (iRF). iRF discovers high-order interactions among biomolecules with the same order of computational cost as random forests. We demonstrate the efficacy of iRF by finding known and promising interactions among biomolecules, of up to fifth and sixth order, in two data examples in transcriptional regulation and alternative splicing.

Genomics has revolutionized biology, enabling the interrogation of whole transcriptomes, genome-wide binding sites for proteins, and many other molecular processes. However, individual genomic assays measure elements that interact in vivo as components of larger molecular machines. Understanding how these high-order interactions drive gene expression presents a substantial statistical challenge. Building on random forests (RFs) and random intersection trees (RITs) and through extensive, biologically inspired simulations, we developed the iterative random forest algorithm (iRF). iRF trains a feature-weighted ensemble of decision trees to detect stable, high-order interactions with the same order of computational cost as the RF. We demonstrate the utility of iRF for high-order interaction discovery in two prediction problems: enhancer activity in the early Drosophila embryo and alternative splicing of primary transcripts in human-derived cell lines. In Drosophila, among the 20 pairwise transcription factor interactions iRF identifies as stable (returned in more than half of bootstrap replicates), 80% have been previously reported as physical interactions. Moreover, third-order interactions, e.g., between Zelda (Zld), Giant (Gt), and Twist (Twi), suggest high-order relationships that are candidates for follow-up experiments. In human-derived cells, iRF rediscovered a central role of H3K36me3 in chromatin-mediated splicing regulation and identified interesting fifth- and sixth-order interactions, indicative of multivalent nucleosomes with specific roles in splicing regulation. By decoupling the order of interactions from the computational cost of identification, iRF opens additional avenues of inquiry into the molecular mechanisms underlying genome biology.

Genome variation and conserved regulation identify genomic regions responsible for strain specific phenotypes in rat

Background

The genomes of laboratory rat strains are characterised by a mosaic haplotype structure caused by their unique breeding history. These mosaic haplotypes have been recently mapped by extensive sequencing of key strains. Comparison of genomic variation between two closely related rat strains with different phenotypes has been proposed as an effective strategy for the discovery of candidate strain-specific regions involved in phenotypic differences. We developed a method to prioritise strain-specific haplotypes by integrating genomic variation and genomic regulatory data predicted to be involved in specific phenotypes. Specifically, we aimed to identify genomic regions associated with Metabolic Syndrome (MetS), a disorder of energy utilization and storage affecting several organ systems.

Results

We compared two Lyon rat strains, Lyon Hypertensive (LH) which is susceptible to MetS, and Lyon Low pressure (LL), which is susceptible to obesity as an intermediate MetS phenotype, with a third strain (Lyon Normotensive, LN) that is resistant to both MetS and obesity. Applying a novel metric, we ranked the identified strain-specific haplotypes using evolutionary conservation of the occupancy three liver-specific transcription factors (HNF4A, CEBPA, and FOXA1) in five rodents including rat. Consideration of regulatory information effectively identified regions with liver-associated genes and rat orthologues of human GWAS variants related to obesity and metabolic traits. We attempted to find possible causative variants and compared them with the candidate genes proposed by previous studies. In strain-specific regions with conserved regulation, we found a significant enrichment for published evidence to obesity—one of the metabolic symptoms shown by the Lyon strains—amongst the genes assigned to promoters with strain-specific variation.

Conclusions

Our results show that the use of functional regulatory conservation is a potentially effective approach to select strain-specific genomic regions associated with phenotypic differences among Lyon rats and could be extended to other systems.

Electronic supplementary material

The online version of this article (doi:10.1186/s12864-017-4351-9) contains supplementary material, which is available to authorized users.

The Pattern of microRNA Binding Site Distribution

Micro-RNA (miRNA or miR) regulates at least 60% of the genes in the human genome through their target sites at mRNA 3'-untranslated regions (UTR), and defects in miRNA expression regulation and target sites are frequently observed in cancers. We report here a systematic analysis of the distribution of miRNA target sites. Using the evolutionarily conserved miRNA binding sites in the TargetScan database (release 7.1), we constructed a miRNA co-regulation network by connecting genes sharing common miRNA target sites. The network possesses characteristics of the ubiquitous small-world network. Non-hub genes in the network-those sharing miRNA target sites with small numbers of genes-tend to form small cliques with their neighboring genes, while hub genes exhibit high levels of promiscuousness in their neighboring genes. Additionally, miRNA target site distribution is extremely uneven. Among the miRNAs, the distribution concentrates on a small number of miRNAs, in that their target sites occur in an extraordinarily large number of genes, that is, they have large numbers of target genes. The distribution across the genes follows a similar pattern; the mRNAs of a small proportion of the genes contain extraordinarily large numbers of miRNA binding sites. Quantitatively, the patterns fit into the P_(K) ∝ K^-α relationship (P_(K): the number of miRNAs with K target genes or genes with K miRNA sites; α: a positive constant), the mathematical description of connection distribution among the nodes and a defining characteristic of the so-called scale-free networks-a subset of small-world networks. Notably, well-known tumor-suppressive miRNAs (Let-7, miR-15/16, 26, 29, 31, 34, 145, 200, 203-205, 223, and 375) collectively have more than expected target genes, and well-known cancer genes contain more than expected miRNA binding sites. In summary, miRNA target site distribution exhibits characteristics of the small-world network. The potential to use this pattern to better understand miRNA function and their oncological roles is discussed.

HOXA1 and TALE proteins display cross-regulatory interactions and form a combinatorial binding code on HOXA1 targets

Hoxa1 has diverse functional roles in differentiation and development. We identify and characterize properties of regions bound by HOXA1 on a genome-wide basis in differentiating mouse ES cells. HOXA1-bound regions are enriched for clusters of consensus binding motifs for HOX, PBX, and MEIS, and many display co-occupancy of PBX and MEIS. PBX and MEIS are members of the TALE family and genome-wide analysis of multiple TALE members (PBX, MEIS, TGIF, PREP1, and PREP2) shows that nearly all HOXA1 targets display occupancy of one or more TALE members. The combinatorial binding patterns of TALE proteins define distinct classes of HOXA1 targets, which may create functional diversity. Transgenic reporter assays in zebrafish confirm enhancer activities for many HOXA1-bound regions and the importance of HOX-PBX and TGIF motifs for their regulation. Proteomic analyses show that HOXA1 physically interacts on chromatin with PBX, MEIS, and PREP family members, but not with TGIF, suggesting that TGIF may have an independent input into HOXA1-bound regions. Therefore, TALE proteins appear to represent a wide repertoire of HOX cofactors, which may coregulate enhancers through distinct mechanisms. We also discover extensive auto- and cross-regulatory interactions among the Hoxa1 and TALE genes, indicating that the specificity of HOXA1 during development may be regulated though a complex cross-regulatory network of HOXA1 and TALE proteins. This study provides new insight into a regulatory network involving combinatorial interactions between HOXA1 and TALE proteins.

Crystal Structure of the CLOCK Transactivation Domain Exon19 in Complex with a Repressor

In the canonical clock model, CLOCK:BMAL1-mediated transcriptional activation is feedback regulated by its repressors CRY and PER and, in association with other coregulators, ultimately generates oscillatory gene expression patterns. How CLOCK:BMAL1 interacts with coregulator(s) is not well understood. Here we report the crystal structures of the mouse CLOCK transactivating domain Exon19 in complex with CIPC, a potent circadian repressor that functions independently of CRY and PER. The Exon19:CIPC complex adopts a three-helical coiled-coil bundle conformation containing two Exon19 helices and one CIPC. Unique to Exon19:CIPC, three highly conserved polar residues, Asn341 of CIPC and Gln544 of the two Exon19 helices, are located at the mid-section of the coiled-coil bundle interior and form hydrogen bonds with each other. Combining results from protein database search, sequence analysis, and mutagenesis studies, we discovered for the first time that CLOCK Exon19:CIPC interaction is a conserved transcription regulatory mechanism among mammals, fish, flies, and other invertebrates.

Taking promoters out of enhancers in sequence based predictions of tissue-specific mammalian enhancers

BACKGROUND

Many genetic diseases are caused by mutations in non-coding regions of the genome. These mutations are frequently found in enhancer sequences, causing disruption to the regulatory program of the cell. Enhancers are short regulatory sequences in the non-coding part of the genome that are essential for the proper regulation of transcription. While the experimental methods for identification of such sequences are improving every year, our understanding of the rules behind the enhancer activity has not progressed much in the last decade. This is especially true in case of tissue-specific enhancers, where there are clear problems in predicting specificity of enhancer activity.

RESULTS

We show a random-forest based machine learning approach capable of matching the performance of the current state-of-the-art methods for enhancer prediction. Then we show that it is, similarly to other published methods, frequently cross-predicting enhancers as active in different tissues, making it less useful for predicting tissue specific activity. Then we proceed to show that the problem is related to the fact that the enhancer predicting models exhibit a bias towards predicting gene promoters as active enhancers. Then we show that using a two-step classifier can lead to lower cross-prediction between tissues.

CONCLUSIONS

We provide whole-genome predictions of human heart and brain enhancers obtained with two-step classifier.

An Enhancer's Length and Composition Are Shaped by Its Regulatory Task

Enhancers drive the gene expression patterns required for virtually every process in metazoans. We propose that enhancer length and transcription factor (TF) binding site composition—the number and identity of TF binding sites—reflect the complexity of the enhancer's regulatory task. In development, we define regulatory task complexity as the number of fates specified in a set of cells at once. We hypothesize that enhancers with more complex regulatory tasks will be longer, with more, but less specific, TF binding sites. Larger numbers of binding sites can be arranged in more ways, allowing enhancers to drive many distinct expression patterns, and therefore cell fates, using a finite number of TF inputs. We compare ~100 enhancers patterning the more complex anterior-posterior (AP) axis and the simpler dorsal-ventral (DV) axis in Drosophila and find that the AP enhancers are longer with more, but less specific binding sites than the (DV) enhancers. Using a set of ~3,500 enhancers, we find enhancer length and TF binding site number again increase with increasing regulatory task complexity. Therefore, to be broadly applicable, computational tools to study enhancers must account for differences in regulatory task.

Multiple binding sites for transcriptional repressors can produce regular bursting and enhance noise suppression

Cells may control fluctuations in protein levels by means of negative autoregulation, where transcription factors bind DNA sites to repress their own production. Theoretical studies have assumed a single binding site for the repressor, while in most species it is found that multiple binding sites are arranged in clusters. We study a stochastic description of negative autoregulation with multiple binding sites for the repressor. We find that increasing the number of binding sites induces regular bursting of gene products. By tuning the threshold for repression, we show that multiple binding sites can also suppress fluctuations. Our results highlight possible roles for the presence of multiple binding sites of negative autoregulators.

Systematic dissection of genomic features determining transcription factor binding and enhancer function

Enhancers regulate gene expression through the binding of sequence-specific transcription factors (TFs) to cognate motifs. Various features influence TF binding and enhancer function-including the chromatin state of the genomic locus, the affinities of the binding site, the activity of the bound TFs, and interactions among TFs. However, the precise nature and relative contributions of these features remain unclear. Here, we used massively parallel reporter assays (MPRAs) involving 32,115 natural and synthetic enhancers, together with high-throughput in vivo binding assays, to systematically dissect the contribution of each of these features to the binding and activity of genomic regulatory elements that contain motifs for PPARγ, a TF that serves as a key regulator of adipogenesis. We show that distinct sets of features govern PPARγ binding vs. enhancer activity. PPARγ binding is largely governed by the affinity of the specific motif site and higher-order features of the larger genomic locus, such as chromatin accessibility. In contrast, the enhancer activity of PPARγ binding sites depends on varying contributions from dozens of TFs in the immediate vicinity, including interactions between combinations of these TFs. Different pairs of motifs follow different interaction rules, including subadditive, additive, and superadditive interactions among specific classes of TFs, with both spatially constrained and flexible grammars. Our results provide a paradigm for the systematic characterization of the genomic features underlying regulatory elements, applicable to the design of synthetic regulatory elements or the interpretation of human genetic variation.

Capitalizing on disaster: Establishing chromatin specificity behind the replication fork

Eukaryotic genomes are packaged into nucleosomal chromatin, and genomic activity requires the precise localization of transcription factors, histone modifications and nucleosomes. Classic work described the progressive reassembly and maturation of bulk chromatin behind replication forks. More recent proteomics has detailed the molecular machines that accompany the replicative polymerase to promote rapid histone deposition onto the newly replicated DNA. However, localized chromatin features are transiently obliterated by DNA replication every S phase of the cell cycle. Genomic strategies now observe the rebuilding of locus-specific chromatin features, and reveal surprising delays in transcription factor binding behind replication forks. This implies that transient chromatin disorganization during replication is a central juncture for targeted transcription factor binding within genomes. We propose that transient occlusion of regulatory elements by disorganized nucleosomes during chromatin maturation enforces specificity of factor binding.

Distinguishing Functional DNA Words; A Method for Measuring Clustering Levels

Functional DNA sub-sequences and genome elements are spatially clustered through the genome just as keywords in literary texts. Therefore, some of the methods for ranking words in texts can also be used to compare different DNA sub-sequences. In analogy with the literary texts, here we claim that the distribution of distances between the successive sub-sequences (words) is q-exponential which is the distribution function in non-extensive statistical mechanics. Thus the q-parameter can be used as a measure of words clustering levels. Here, we analyzed the distribution of distances between consecutive occurrences of 16 possible dinucleotides in human chromosomes to obtain their corresponding q-parameters. We found that CG as a biologically important two-letter word concerning its methylation, has the highest clustering level. This finding shows the predicting ability of the method in biology. We also proposed that chromosome 18 with the largest value of q-parameter for promoters of genes is more sensitive to dietary and lifestyle. We extended our study to compare the genome of some selected organisms and concluded that the clustering level of CGs increases in higher evolutionary organisms compared to lower ones.

Identification of a novel distal regulatory element of the human Neuroglobin gene by the chromosome conformation capture approach

Neuroglobin (NGB) is predominantly expressed in the brain and retina. Studies suggest that NGB exerts protective effects to neuronal cells and is implicated in reducing the severity of stroke and Alzheimer's disease. However, little is known about the mechanisms which regulate the cell type-specific expression of the gene. In this study, we hypothesized that distal regulatory elements (DREs) are involved in optimal expression of the NGB gene. By chromosome conformation capture we identified two novel DREs located -70 kb upstream and +100 kb downstream from the NGB gene. ENCODE database showed the presence of DNaseI hypersensitive and transcription factors binding sites in these regions. Further analyses using luciferase reporters and chromatin immunoprecipitation suggested that the -70 kb region upstream of the NGB gene contained a neuronal-specific enhancer and GATA transcription factor binding sites. Knockdown of GATA-2 caused NGB expression to drop dramatically, indicating GATA-2 as an essential transcription factor for the activation of NGB expression. The crucial role of the DRE in NGB expression activation was further confirmed by the drop in NGB level after CRISPR-mediated deletion of the DRE. Taken together, we show that the NGB gene is regulated by a cell type-specific loop formed between its promoter and the novel DRE.