ChromHMM: automating chromatin-state discovery and characterization

Illuminating the genome

DNA methylation has been proved to play important roles in cell development and complex diseases through comparative studies of DNA methylation profiles across different tissues and samples. Current studies indicate that the regulation of DNA methylation to gene expression depends on the genomic locations of CpGs. Common DNA methylation patterns shared across different cell types and tissues are abundant, and they are likely involved in the basic functions of cell development, such as housekeeping functions. By way of contrast, cell type-specific DNA methylation patterns show distinct functional relevance with cell type specificity. Additionally, abnormal DNA methylation patterns are extensively involved in tumour development. Pan-cancer methylation patterns reveal common mechanisms and new similarities of different cancers, while cancer-specific patterns are relating to tumour heterogeneity and patient survival. Moreover, DNA methylation patterns in specific cancer are relevant with diverse regulatory elements such as enhancers and long non-coding RNAs. In this review, we survey the recent advances on DNA methylation patterns in normal or tumour states to illustrate their potential roles in cell development and cell canceration.

Changes in chromatin state play important roles in cell fate transitions. Current computational approaches to analyze chromatin modifications across multiple cell types do not model how the cell types are related on a lineage or over time. To overcome this limitation, we developed a method called Chromatin Module INference on Trees (CMINT), a probabilistic clustering approach to systematically capture chromatin state dynamics across multiple cell types. Compared to existing approaches, CMINT can handle complex lineage topologies, capture higher quality clusters, and reliably detect chromatin transitions between cell types. We applied CMINT to gain novel insights in two complex processes: reprogramming to induced pluripotent stem cells (iPSCs) and hematopoiesis. In reprogramming, chromatin changes could occur without large gene expression changes, different combinations of activating marks were associated with specific reprogramming factors, there was an order of acquisition of chromatin marks at pluripotency loci, and multivalent states (comprising previously undetermined combinations of activating and repressive histone modifications) were enriched for CTCF. In the hematopoietic system, we defined critical decision points in the lineage tree, identified regulatory elements that were enriched in cell-type–specific regions, and found that the underlying chromatin state was achieved by specific erasure of preexisting chromatin marks in the precursor cell or by de novo assembly. Our method provides a systematic approach to model the dynamics of chromatin state to provide novel insights into the relationships among cell types in diverse cell-fate specification processes.

ATP-citrate lyase (Acly) is one of two cytosolic enzymes that synthesize acetyl-coenzyme A (CoA). Because acetyl-CoA is an essential building block for cholesterol and triglycerides, Acly has been considered a therapeutic target for hyperlipidemias and obesity. To define the phenotype of Acly-deficient mice, we created Acly knockout mice in which a beta-galactosidase marker is expressed from Acly regulatory sequences. We also sought to define the cell type-specific expression patterns of Acly to further elucidate the in vivo roles of the enzyme. Homozygous Acly knockout mice died early in development. Heterozygous mice were healthy, fertile, and normolipidemic on both chow and high fat diets, despite expressing half-normal amounts of Acly mRNA and protein. Fibroblasts and hepatocytes from heterozygous Acly mice contained half-normal amounts of Acly mRNA and protein, but this did not perturb triglyceride and cholesterol synthesis or the expression of lipid biosynthetic genes regulated by sterol regulatory element-binding proteins. The expression of acetyl-CoA synthetase 1, another cytosolic enzyme for producing acetyl-CoA, was not up-regulated. As judged by beta-galactosidase staining, Acly was expressed ubiquitously but was expressed particularly highly in tissues with high levels of lipogenesis, such as in the livers of mice fed a high-carbohydrate diet. beta-Galactosidase staining was intense in the developing brain, in keeping with the high levels of de novo lipogenesis of the tissue. In the adult brain, beta-galactosidase staining was in general much lower, consistent with reduced levels of lipogenesis; however, beta-galactosidase expression remained very high in cholinergic neurons, likely reflecting the importance of Acly in generating acetyl-CoA for acetylcholine synthesis. The Acly knockout allele is useful for identifying cell types with a high demand for acetyl-CoA synthesis.

Hippocampal Pyramidal Neurons Comprise Two Distinct Cell Types that Are Countermodulated by Metabotropic Receptors

Endothelial cells are remarkably heterogeneous in both morphology and function, and they play critical roles in the formation of multiple organ systems. In addition endothelial cell dysfunction can contribute to disease processes, including diabetic nephropathy, which is a leading cause of end stage renal disease. In this report we define the comprehensive gene expression programs of multiple types of kidney endothelial cells, and analyze the differences that distinguish them. Endothelial cells were purified from Tie2-GFP mice by cell dissociation and fluorescent activated cell sorting. Microarrays were then used to provide a global, quantitative and sensitive measure of gene expression levels. We examined renal endothelial cells from the embryo and from the adult glomerulus, cortex and medulla compartments, as well as the glomerular endothelial cells of the db/db mutant mouse, which represents a model for human diabetic nephropathy. The results identified the growth factors, receptors and transcription factors expressed by these multiple endothelial cell types. Biological processes and molecular pathways were characterized in exquisite detail. Cell type specific gene expression patterns were defined, finding novel molecular markers and providing a better understanding of compartmental distinctions. Further, analysis of enriched, evolutionarily conserved transcription factor binding sites in the promoters of co-activated genes begins to define the genetic regulatory network of renal endothelial cell formation. Finally, the gene expression differences associated with diabetic nephropathy were defined, providing a global view of both the pathogenic and protective pathways activated. These studies provide a rich resource to facilitate further investigations of endothelial cell functions in kidney development, adult compartments, and disease.

For an embryo to successfully develop into an adult animal, specific genes must act in different types of cells. Though all the cells have the same genes encoded within their DNA, looking at the way that the DNA is packaged can indicate which parts of the DNA are important for that particular cell type. If regions of DNA are “open” one can infer that those regions are actively involved in gene regulation, whereas “closed” regions are considered less important. It is currently difficult to determine which parts of the DNA are open within an individual cell type in a complex organ, such as the brain. Existing methods require the cells to be physically isolated from the tissue, which is technically challenging. To overcome this issue, Aughey et al. have now developed a method that does not require isolation of the cells. The new technique involves using genetic engineering to introduce an enzyme called Dam into specific cell types in living fruit flies. This enzyme adds a chemical label on regions of open DNA, which can then be detected. Aughey et al. tested this technique on various cells of the developing brain and gut, and were able to see differences in the openness of DNA that corresponded to the action of genes that are important in each cell type. The data also contain trends that help to understand the role of open DNA in development. For example, mature cells were shown to overall have less open DNA than the stem cells that divide to generate them. Aughey et al. hope their new technique will be of use to other researchers working with either fruit flies or mammalian tissues. The knowledge that scientists will gain from identifying how open DNA contributes to gene regulation, in both healthy and diseased tissues, will further our understanding of human development and the biology of diseases such as cancer.

Bio-acoustic levitational assembly of heterocellular 3D constructs: 3D model establishment for cells radiation effect studies in 3D microenvironment

Tight control of gene expression is crucial to govern cell function and identity at any developmental stage. Epigenetic modifications of chromatin have emerged as important determinants for chromatin structure and gene expression. The aim of this work was to test the hypothesis that epigenetic mechanisms contribute to the establishment and maintenance of cell type specific gene expression patterns and to delimiting the developmental potential of somatic cells. Towards this goal we defined genome-wide targets of epigenetic reprogramming during neuronal differentiation of mouse embryonic stem cells. DNA methylation, which is a potent and stable repressive modification, is increasing during differentiation of embryonic stem cells into neurons. Many de novo methylation targets encode pluripotency-associated and germline specific genes and only few appear to be specific for alternative lineages. Polycomb-mediated repression, a distinct epigenetic repression pathway, was previously shown to be essential for embryonic patterning and maintaining developmental potential in stem cells. Unlike DNA methylation, Polycomb targets are very dynamic during neuronal differentiation. Repression is resolved at activated genes while novel targets appear at both the multipotential neuronal progenitor state and the terminally differentiated neuron state. As in stem cells, many Polycomb targets in neuronal progenitor cells will be activated upon further differentiation. Polycomb could therefore serve as a general safe-guard system for genes that can be activated at later stages but need to be tightly controlled to avoid precocious and uncontrolled cell fate changes. In summary, there are at least two distinct epigenetic modes of repression, which nonetheless might crosstalk for target specification. Stable repression of the pluripotency program is conferred by DNA methylation. In turn, Polycomb mediates a more transient repression mechanism with cell type and developmental stage specific targets. Together, this argues that epigenetic mechanisms contribute to cellular differentiation and development via stabilizing gene expression programs initiated by transcription factors. Hence, epigenetic mechanisms could be viewed as additional regulatory layer for balancing gene regulation in order to confer robustness to cellular states and gene expression programs rather than as key drivers for setting up such cell type specific gene expression patterns.

Highly purified GH-receptor preparations from 3T3-F442A fibroblasts, whose differentiation into adipocytes is promoted by GH, have been shown to contain a tyrosine kinase capable of phosphorylating GH receptors. In the current work, characteristics of the tyrosine kinase responsible for the in vitro phosphorylation of GH receptors from cultured 3T3-F442A fibroblasts were examined, and the presence of this GH receptor-associated tyrosine kinase activity was demonstrated in multiple cell types. GH-receptor complexes from GH-treated cells were partially purified by immunoprecipitation using anti-GH antibodies and then incubated as an immune complex with [gamma 32P] ATP. Incorporation of 32P into the GH receptor from 3T3-F442A fibroblasts was apparent within 1 min at 30 C after the addition of [gamma 32P]ATP (5-10 microM). A divalent cation was requisite for the phosphorylation; Mn2+ was significantly more effective than Mg2+ and Co2+; Ba2+, Ca2+, or Zn2+ had no effect. Excess unlabeled ATP, but not cytosine triphosphate, GTP, or uridine triphosphate, abolished 32P incorporation into the GH receptor and [gamma 32P]GTP could not replace [gamma 32P]ATP as a source of 32P. At 5.5 mM Mn2+, phosphorylation exhibited a biphasic dose response to ATP, with maximal phosphorylation occurring at a concentration of 10 microM ATP. At more physiological concentrations of ATP (1 mM), phosphorylation of the GH receptor was also stimulated by lower concentrations of Mn2+ (as low as 500 nM). Optimal reaction conditions determined for the phosphorylation reaction in 3T3-F442A fibroblasts were used to demonstrate incorporation of 32P from [gamma 32P]ATP into partially purified GH receptors from cultured human IM-9 lymphocytes, murine 3T3-F442A adipocytes, rat H-35 hepatoma cells, and freshly isolated rat adipocytes. The 32P was shown to be incorporated into tyrosyl residues in receptors from the two cell types tested (IM-9 lymphocytes and rat adipocytes). Cross-linked [125I] hGH-receptor complexes solubilized from the four cell types (IM-9 lymphocytes, 3T3-F442A adipocytes, H-35 hepatoma cells, and freshly isolated rat adipocytes) bound to and could be eluted from phosphotyrosyl antibody, suggesting that tyrosyl phosphorylation of GH receptors in all of these cells occurs in vivo. The presence of tyrosine kinase activity associated with GH receptors in multiple cell types from different species is consistent with tyrosine kinase activity playing a role in the actions of GH.

Recent studies with myosin heavy chain mutants in the slime mold Dictyostelium discoideum and the yeast Saccharomyces cerevisiae indicate that the myosin heavy chain gene is not essential for cell survival under laboratory growth conditions. However, cells lacking a normal myosin heavy chain gene demonstrate substantial alterations in growth and cell division. In this study, we report that a disruption mutant in the rod portion of the yeast myosin heavy chain gene, MYO1, produces abnormal chitin distribution and cell wall organization at the mother-bud neck in a high proportion of dividing cells. It is suggested that this phenotype is the cause of the cell division defect and the osmotic sensitivity of yeast MYO1 mutants. In the absence of a normal MYO1 polypeptide, yeast cells alter their cell type specific budding pattern. It is concluded that an intact myosin heavy chain gene is required to maintain the cell type specific budding pattern and the correct localization and deposition of chitin and cell wall components during cell growth and division.

BackgroundGenome-wide maps of chromatin marks such as histone modifications and open chromatin sites provide valuable information for annotating the non-coding genome, including identifying regulatory elements. Computational approaches such as ChromHMM have been applied to discover and annotate chromatin states defined by combinatorial and spatial patterns of chromatin marks within the same cell type. An alternative “stacked modeling” approach was previously suggested, where chromatin states are defined jointly from datasets of multiple cell types to produce a single universal genome annotation based on all datasets. Despite its potential benefits for applications that are not specific to one cell type, such an approach was previously applied only for small-scale specialized purposes. Large-scale applications of stacked modeling have previously posed scalability challenges.ResultsUsing a version of ChromHMM enhanced for large-scale applications, we apply the stacked modeling approach to produce a universal chromatin state annotation of the human genome using over 1000 datasets from more than 100 cell types, with the learned model denoted as the full-stack model. The full-stack model states show distinct enrichments for external genomic annotations, which we use in characterizing each state. Compared to per-cell-type annotations, the full-stack annotations directly differentiate constitutive from cell type-specific activity and is more predictive of locations of external genomic annotations.ConclusionsThe full-stack ChromHMM model provides a universal chromatin state annotation of the genome and a unified global view of over 1000 datasets. We expect this to be a useful resource that complements existing per-cell-type annotations for studying the non-coding human genome.

The Drosophila Dysfusion basic-helix-loop-helix-PAS (bHLH-PAS) protein controls the transcription of genes that mediate tracheal fusion. Dysfusion is highly related to the mammalian Nxf protein that has been implicated in nervous system gene regulation. Toward the goal of understanding how Dysfusion controls fusion cell gene expression, the biochemical properties of Dysfusion were investigated using protein interaction experiments, cell culture-based transcription assays, and in vivo transgenic analyses. Dysfusion dimerizes with the Tango bHLH-PAS protein, and together they act as a DNA binding transcriptional activator. Dysfusion/Tango binds multiple NCGTG binding sites, with the following preference: TCGTG > GCGTG > ACGTG > CCGTG. This binding site promiscuity differs from the restricted binding site preferences of other bHLH-PAS/Tango heterodimers. However, it is identical to the binding site preferences of mammalian Nxf/Arnt, indicating that the specificity is evolutionarily conserved. Germ line transformation experiments using a fragment of the CG13196 Dysfusion target gene allowed identification of a fusion cell enhancer. Experiments in which NCGTG sites were mutated individually and in combination revealed that TCGTG sites were required for fusion cell expression but that the single ACGTG and GCGTG sites present were not. Finally, a reporter transgene containing four tandemly arranged TCGTG elements has strong expression in tracheal fusion cells. Transgenic misexpression of dysfusion further revealed that Dysfusion has the ability to activate transcription in multiple cell types, although it does this most effectively in tracheal cells and can only function at mid-embryogenesis and later.

Sequence and chromatin determinants of transcription factor binding and the establishment of cell type-specific binding patterns

The Neuron Identity Problem: Form Meets Function