The ENCODE Imputation Challenge: a critical assessment of methods for cross-cell type imputation of epigenomic profiles.

A promising alternative to comprehensively performing genomics experiments is to, instead, perform a subset of experiments and use computational methods to impute the remainder. However, identifying the best imputation methods and what measures meaningfully evaluate performance are open questions. We address these questions by comprehensively analyzing 23 methods from the ENCODE Imputation Challenge. We find that imputation evaluations are challenging and confounded by distributional shifts from differences in data collection and processing over time, the amount of available data, and redundancy among performance measures. Our analyses suggest simple steps for overcoming these issues and promising directions for more robust research.

Tissue context determines the penetrance of regulatory DNA variation.

Functional assessment of disease-associated sequence variation at non-coding regulatory elements is complicated by their high degree of context sensitivity to both the local chromatin and nuclear environments. Allelic profiling of DNA accessibility across individuals has shown that only a select minority of sequence variation affects transcription factor (TF) occupancy, yet low sequence diversity in human populations means that no experimental assessment is available for the majority of disease-associated variants. Here we describe high-resolution in vivo maps of allelic DNA accessibility in liver, kidney, lung and B cells from 5 increasingly diverged strains of F1 hybrid mice. The high density of heterozygous sites in these hybrids enables precise quantification of effect size and cell-type specificity for hundreds of thousands of variants throughout the mouse genome. We show that chromatin-altering variants delineate characteristic sensitivity profiles for hundreds of TF motifs. We develop a compendium of TF-specific sensitivity profiles accounting for genomic context effects. Finally, we link maps of allelic accessibility to allelic transcript levels in the same samples. This work provides a foundation for quantitative prediction of cell-type specific effects of non-coding variation on TF activity, which will facilitate both fine-mapping and systems-level analyses of common disease-associated variation in human genomes.

Index and biological spectrum of human DNase I hypersensitive sites.

DNase I hypersensitive sites (DHSs) are generic markers of regulatory DNA1-5 and contain genetic variations associated with diseases and phenotypic traits6-8. We created high-resolution maps of DHSs from 733 human biosamples encompassing 438 cell and tissue types and states, and integrated these to delineate and numerically index approximately 3.6 million DHSs within the human genome sequence, providing a common coordinate system for regulatory DNA. Here we show that these maps highly resolve the cis-regulatory compartment of the human genome, which encodes unexpectedly diverse cell- and tissue-selective regulatory programs at very high density. These programs can be captured comprehensively by a simple vocabulary that enables the assignment to each DHS of a regulatory barcode that encapsulates its tissue manifestations, and global annotation of protein-coding and non-coding RNA genes in a manner orthogonal to gene expression. Finally, we show that sharply resolved DHSs markedly enhance the genetic association and heritability signals of diseases and traits. Rather than being confined to a small number of distal elements or promoters, we find that genetic signals converge on congruently regulated sets of DHSs that decorate entire gene bodies. Together, our results create a universal, extensible coordinate system and vocabulary for human regulatory DNA marked by DHSs, and provide a new global perspective on the architecture of human gene regulation.

Global reference mapping of human transcription factor footprints.

Combinatorial binding of transcription factors to regulatory DNA underpins gene regulation in all organisms. Genetic variation in regulatory regions has been connected with diseases and diverse phenotypic traits1, but it remains challenging to distinguish variants that affect regulatory function2. Genomic DNase I footprinting enables the quantitative, nucleotide-resolution delineation of sites of transcription factor occupancy within native chromatin3-6. However, only a small fraction of such sites have been precisely resolved on the human genome sequence6. Here, to enable comprehensive mapping of transcription factor footprints, we produced high-density DNase I cleavage maps from 243 human cell and tissue types and states and integrated these data to delineate about 4.5 million compact genomic elements that encode transcription factor occupancy at nucleotide resolution. We map the fine-scale structure within about 1.6 million DNase I-hypersensitive sites and show that the overwhelming majority are populated by well-spaced sites of single transcription factor-DNA interaction. Cell-context-dependent cis-regulation is chiefly executed by wholesale modulation of accessibility at regulatory DNA rather than by differential transcription factor occupancy within accessible elements. We also show that the enrichment of genetic variants associated with diseases or phenotypic traits in regulatory regions1,7 is almost entirely attributable to variants within footprints, and that functional variants that affect transcription factor occupancy are nearly evenly partitioned between loss- and gain-of-function alleles. Unexpectedly, we find increased density of human genetic variation within transcription factor footprints, revealing an unappreciated driver of cis-regulatory evolution. Our results provide a framework for both global and nucleotide-precision analyses of gene regulatory mechanisms and functional genetic variation.

Expanded encyclopaedias of DNA elements in the human and mouse genomes.

The human and mouse genomes contain instructions that specify RNAs and proteins and govern the timing, magnitude, and cellular context of their production. To better delineate these elements, phase III of the Encyclopedia of DNA Elements (ENCODE) Project has expanded analysis of the cell and tissue repertoires of RNA transcription, chromatin structure and modification, DNA methylation, chromatin looping, and occupancy by transcription factors and RNA-binding proteins. Here we summarize these efforts, which have produced 5,992 new experimental datasets, including systematic determinations across mouse fetal development. All data are available through the ENCODE data portal (https://www.encodeproject.org), including phase II ENCODE1 and Roadmap Epigenomics2 data. We have developed a registry of 926,535 human and 339,815 mouse candidate cis-regulatory elements, covering 7.9 and 3.4% of their respective genomes, by integrating selected datatypes associated with gene regulation, and constructed a web-based server (SCREEN; http://screen.encodeproject.org) to provide flexible, user-defined access to this resource. Collectively, the ENCODE data and registry provide an expansive resource for the scientific community to build a better understanding of the organization and function of the human and mouse genomes.

Global Regulatory DNA Potentiation by SMARCA4 Propagates to Selective Gene Expression Programs via Domain-Level Remodeling.

The human genome encodes millions of regulatory elements, of which only a small fraction are active within a given cell type. Little is known about the global impact of chromatin remodelers on regulatory DNA landscapes and how this translates to gene expression. We use precision genome engineering to reawaken homozygously inactivated SMARCA4, a central ATPase of the human SWI/SNF chromatin remodeling complex, in lung adenocarcinoma cells. Here, we combine DNase I hypersensitivity, histone modification, and transcriptional profiling to show that SMARCA4 dramatically increases both the number and magnitude of accessible chromatin sites genome-wide, chiefly by unmasking sites of low regulatory factor occupancy. By contrast, transcriptional changes are concentrated within well-demarcated remodeling domains wherein expression of specific genes is gated by both distal element activation and promoter chromatin configuration. Our results provide a perspective on how global chromatin remodeling activity is translated to gene expression via regulatory DNA.

Prelamin A processing, accumulation and distribution in normal cells and laminopathy disorders.

Lamin A is part of a complex structural meshwork located beneath the nuclear envelope and is involved in both structural support and the regulation of gene expression. Lamin A is initially expressed as prelamin A, which contains an extended carboxyl terminus that undergoes a series of post-translational modifications and subsequent cleavage by the endopeptidase ZMPSTE24 to generate lamin A. To facilitate investigations of the role of this cleavage in normal and disease states, we developed a monoclonal antibody (PL-1C7) that specifically recognizes prelamin A at the intact ZMPSTE24 cleavage site, ensuring prelamin A detection exclusively. Importantly, PL-1C7 can be used to determine prelamin A localization and accumulation in cells where lamin A is highly expressed without the use of exogenous fusion proteins. Our results show that unlike mature lamin A, prelamin A accumulates as discrete and localized foci at the nuclear periphery. Furthermore, whereas treatment with farnesylation inhibitors of cells overexpressing a GFP-prelamin A fusion protein results in the formation of large nucleoplasmic clumps, these aggregates are not observed upon similar treatment of cells expressing endogenous prelamin A or in cells lacking ZMPSTE24 expression and/or activity. Finally, we show that specific laminopathy-associated mutations exhibit both positive and negative effects on prelamin A accumulation, indicating that these mutations affect prelamin A processing efficiency in different manners.

UpSET recruits HDAC complexes and restricts chromatin accessibility and acetylation at promoter regions.

Developmental gene expression results from the orchestrated interplay between genetic and epigenetic mechanisms. Here, we describe upSET, a transcriptional regulator encoding a SET domain-containing protein recruited to active and inducible genes in Drosophila. However, unlike other Drosophila SET proteins associated with gene transcription, UpSET is part of an Rpd3/Sin3-containing complex that restricts chromatin accessibility and histone acetylation to promoter regions. In the absence of UpSET, active chromatin marks and chromatin accessibility increase and spread to genic and flanking regions due to destabilization of the histone deacetylase complex. Consistent with this, transcriptional noise increases, as manifest by activation of repetitive elements and off-target genes. Interestingly, upSET mutant flies are female sterile due to upregulation of key components of Notch signaling during oogenesis. Thus UpSET defines a class of metazoan transcriptional regulators required to fine tune transcription by preventing the spread of active chromatin.

H3 K79 dimethylation marks developmental activation of the beta-globin gene but is reduced upon LCR-mediated high-level transcription.

Genome-wide analyses of the relationship between H3 K79 dimethylation and transcription have revealed contradictory results. To clarify this relationship at a single locus, we analyzed expression and H3 K79 modification levels of wild-type (WT) and transcriptionally impaired beta-globin mutant genes during erythroid differentiation. Analysis of fractionated erythroid cells derived from WT/Delta locus control region (LCR) heterozygous mice reveals no significant H3 K79 dimethylation of the beta-globin gene on either allele prior to activation of transcription. Upon transcriptional activation, H3 K79 di-methylation is observed along both WT and DeltaLCR alleles, and both alleles are located in proximity to H3 K79 dimethylation nuclear foci. However, H3 K79 di-methylation is significantly increased along the DeltaLCR allele compared with the WT allele. In addition, analysis of a partial LCR deletion mutant reveals that H3 K79 dimethylation is inversely correlated with beta-globin gene expression levels. Thus, while our results support a link between H3 K79 dimethylation and gene expression, high levels of this mark are not essential for high level beta-globin gene transcription. We propose that H3 K79 dimethylation is destabilized on a highly transcribed template.

The beta -globin locus control region (LCR) functions primarily by enhancing the transition from transcription initiation to elongation.

To investigate the molecular basis of beta-globin gene activation, we analyzed factor recruitment and histone modification at the adult beta-globin gene in wild-type (WT)/locus control region knockout (DeltaLCR) heterozygous mice and in murine erythroleukemia (MEL) cells. Although histone acetylation and methylation (Lys 4) are high before and after MEL differentiation, recruitment of the erythroid-specific activator NF-E2 to the promoter and preinitiation complex (PIC) assembly occur only after differentiation. We reported previously that targeted deletion of the LCR reduces beta-globin gene expression to 1%-4% of WT without affecting promoter histone acetylation. Here, we report that NF-E2 is recruited equally efficiently to the adult beta-globin promoters of the DeltaLCR and WT alleles. Moreover, the LCR deletion reduces PIC assembly only twofold, but has a dramatic effect on Ser 5 phosphorylation of RNA polymerase II and transcriptional elongation. Our results suggest at least three distinct stages in beta-globin gene activation: (1) an LCR-independent chromatin opening stage prior to NF-E2 recruitment to the promoter and PIC assembly; (2) an intermediate stage in which NF-E2 binding (LCR-independent) and PIC assembly (partially LCR-dependent) occur; and (3) an LCR-dependent fully active stage characterized by efficient pol II elongation. Thus, in its native location the LCR functions primarily downstream of activator recruitment and PIC assembly.