Top Epigenetics articles: April 2015

Loss of TET2 in hematopoietic cells leads to DNA hypermethylation of active enhancers and induction of leukemogenesis

TET2 loss leads to leukemogenesis by DNA hypermethlation

The TET family of proteins play an essential role in DNA demethylation by converting 5-methylcytosine (5-mC) to 5-hydroxymethylcytosine (5-hmC). TET2 is frequently mutated in hematological cancers including acute myeloid leukemia (AML).

A team led by Kristian Helin from Biotech Research and Innovation Centre, University of Copenhagen and The Danish Stem Cell Centre sought to understand the role of TET2 mutations in DNA methylation and development of leukemia. Using a mouse model for TET2-deficient AML found that:

• Loss of TET2 leads to hypermethylation of active enhancer elements, resulting in loss of enhancer activity and reduced expression of nearby genes.
• CpG island and promoter methylation does not change in a TET2-dependent manner, but increases relative to population doublings. 
• Human AML patients with TET2 mutations had enhanced enrichment of hypermethylated CpG sites.
• There is a rapid deregulation of a large number of genes implicated in tumorigenesis, including many down-regulated tumor suppressor genes.

The authors demonstrate that loss of TET2 leads to genome-wide increase in DNA methylation of active enhancers, and propose that combined silencing of tumor suppressor genese contributes to tumorigenesis in AML.

Read the full paper in Genes and Development, April 2015.

Interested in DNA methylation? Take a look at all our DNA methylation resources.

​Systematic discovery of Xist binding proteins

New technique for the discovery of lncRNA binding proteins

RNA binding proteins play an important part in long non-coding RNA (lncRNA)-mediated gene regulation, and identification of such proteins is critical for understanding the functioning of lncRNAs. Although tools have been developed to identify lncRNAs that bind to specific proteins, there is no ideal strategy to identify the proteins that bind to specific lncRNAs.

A team led by Howard Chang from Stanford University School of Medicine has developed comprehensive identification of RNA binding proteins by mass spectrometry (ChIRP-MS) to identify endogenous protein partners associated with specific RNAs.

Using this technique, the team investigated proteins that bind to Xist, a lncRNA essential for X inactivation. They found that:

• Xist interacts with 81 proteins from chromatin modification, the majority of these proteins were found in all four cell types examined; however 19 of these interact with Xist only in certain differentiated cells. 
• HnrnpK, a specific interactor with Xist, is important for Xist-mediated gene silencing and Xist-mediated chromatin modifications, but does not contribute to Xist biogenesis or localization. 
• Xist also interacts with Spen via its A-repeat domain to mediate gene silencing. 

These data show that the Xist lncRNA binds with numerous proteins to coordinate chromatin spreading and silencing. This work paves the way for future structure-function analysis of Xist and its interacting proteins.

Read the full paper in Cell, April 2015.

Transcriptional co-repressor function of the Hippo pathway transducers YAP and TAZ

YAP and TAZ act as transcriptional repressors as well as promoters

​YAP and TAZ are transcriptional co-activators that are opposed by the Hippo tumor-suppressor pathway and overexpression of these factors leads to cancer development. Numerous genes have been identified that are upregulated by YAP/TAZ; however, these fail to completely account for the YAP/TAZ overexpression phenotype. 

To further understand how YAP/TAZ functions in oncogenesis, Minchul Kim, Dai-Sik Lim and colleagues from the Korea Advanced Institute of Science and Technology and M.D. Anderson Cancer Center in Texas investigated whether YAP/TAZ could act as transcriptional co-repressors of antiproliferative and cell-death inducing genes. 

They found that:

• Approximately 100 genes were acutely suppressed by YAP/TAZ, including tumor-suppressor genes such as DDIT4 (DNA-damage-inducible transcript 4) and Trail (TNF-related apoptosis-inducing ligand).
• Gene repression by YAP/TAZ requires TEAD (TEA domain) transcription factors. 
• Lysine acetylation and histone H3 occupancy are reduced with constitutive YAP expression, but decreased in cells depleted of YAP/TAZ.
• A NuRD (nucleosome remodeling and histone deacetylase) complex is recruited to target genes to mediate repressor function of YAP/TAZ.
• DDIT4 and Trail repression by YAP/TAZ is required for mTORC1 (mechanistic target of rapamycin complex 1) activation and promotes cell survival, underpinning oncogenic function of YAP/TAZ.

The data presented in this paper demonstrate that YAP/TAZ can act as transcriptional co-repressors. this suggests that they can function as oncogenes by repressing antiproliferative and cell-death inducing genes. and opens a new avenue for understanding the Hippo signaling pathway.

Read the full paper in Cell Reports, April 2015.

Pioneer transcription factors target partial DNA motifs on nucleosomes to initiate reprogramming

Mechanistic basis of target recognition by pioneer transcription factors​

​Pioneer transcription factors, including FoxA, have the ability to access silent chromatin to initiate cell fate changes. Although it is known that FoxA binds directly to DNA using a DNA binding domain that resembles linker histones, whether other pioneer transcription factors contain structures allowing direct binding to nucleosomes has not been assessed.

A team led by Kenneth Zaret at the University of Pennsylvania has investigated the nucleosome and chromatin targeting activities of the transcription factors that reprogram somatic cells to pluripotency; Oct4, Sox2, Klf4 and c-Myc.

They found that:

• Purified Oct4, Sox2, Klf4 and c-Myc bind nucleosomes in vitro with differential affinity and through specific and non-specific DNA interactions. 
• In vivo, the factors preferentially bind regions enriched in nucleosomes.
• Oct4, Sox2 and Klf4 factors target nucleosomes using partial or degenerate motifs, but use full motifs when DNA is not bound to nucleosomes at the target site. c-Myc binds cannot bind nucleosomes independently, but associates with other factors to target nucleosomes.
• Oct4 targets partial domains using its bipartite POU domain, Sox2 affinity for nucleosomes is conferred by the pre-bent conformation of the DNA and Klf4 uses two out of its three zinc fingers to recognize a hexameric motif.

This paper has demonstrated how pioneer factors are able to target sites within silent chromatin. Understanding the mechanistic basis behind target recognition by transcription factors paves the way to being able to control the process of transcription factor binding and cell fate determination.

Read the full paper in Cell, April 2015.

Restricted epigenetic inheritance of H3K9 methylation

H3K9 methylation marks are actively removed to prevent inheritance

Histone H3 lysine 9 methylation (H3K9me) is involved in the formation of constitutive heterochromatin. However, the heritability of H3K9me has not previously been demonstrated.

Fission yeast does not have DNA methylation, and has a single methylatransferase (Clr4) responsible for H3K9me-dependant heterochromatin, making analysis of heritability more straightforward than in eukaryotic systems.

Pauline Audergon, Robin Allshire and colleagues from the University of Edinburgh used fission yeast as a model to investigate epigenetic heritablility of H3K9me. By constitutively tethering Clr4 to euchromatin, an extensive domain of H3K9me-dependent heterochromatin is assembled. Using this system, the authors found that:

• H3K9 is rapidly demethylated at the Clr4 tethering site through the cell cycle and in non-cycling cells, suggesting demethylation is by an active process.
• Inactivation of Epe1, a putative H3K9 demethylase, allows H3K9 methylation and heterochromatin retention at the tethering site through mitotic divisions and transgenerationally through meiosis.

The data presented in this paper demonstrate that H3K9 methylation is a heritable epigenetic mark whose transmission is usually countered by its active removal, preventing the unauthorized inheritance of heterochromatin. This represents a built-in safety mechanism to avoid potentially deleterious gene silencing.

Read the full paper in Science, April 2015.

FBXL10 protects Polycomb-bound genes from hypermethylation

FBXL10 deficiency leads to aberrant DNA methylation

​FBXL10 is a multidomain chromosomal protein that binds to CpG-dense promoters, and is bound at the majority of promoter-associated CpG islands in the mouse genome. Mutations affecting FBXL10 are commonly found in diseases including human diffuse large B cell lymphoma and transposon-induced mouse lymphoma.

Mathieu Boulard and colleagues from the College of Physicians and Surgeons of Columbia University and Washington University School of Medicine investigated the biological function of FBXL10. Using FBXL10-mutant mouse embryos, they found that:

• Inactivation of the FBXL10​ gene results in lethal developmental defects and death of homozygous FBXL10​-mutant embryos midgestation.
• Embryonic stem (ES) cells deficient for FBXL10 have marked hypermethylation of genes that are bound by both FBXL10 and Polycomb repressive complex (PRC) 1 and 2. Most of the affected genes are involved in patterning of the early embryo.
• Abberant DNA methylation induced by FBXL10 deficiency results in silencing of genes bound by both FBXL10 and PRC1 or 2, but not genes bound by FBXL10 in the absence of PRCs.
• Deletion of key components of PRC1 and PRC2 did not lead to ectopic genome methylation.

The data presented in this paper indicate that Polycomb domains recruit DNA methyltransferase, and that FBXL10 prevents de novo methylation at PRC1 and PRC2 bound DNA sequences. This exciting development is the first example of a protein necessary for protection of DNA from hypermethlation.

Read the full paper in Nature Genetics, April 2015.

Find out more about the role of Polycombs in cancer.

Epigenetic priming of enhancers predicts developmental competence of hESC-derived endodermal lineage intermediates

A poised epigenetic state at enhancers confers developmental competence

Developmental competence refers to the ability of progenitor cells to appropriately interpret and respond to inductive cues from their environment. The mechanisms that render cells developmentally competent is currently unknown. 

To investigate whether enhancer epigenetic state determines developmental competence, Allen Wang and colleagues from the University of California, San Diego, the University of Pennsylvania and Pennsylvania State University developed maps of enhancer-related chromatin modifications throughout human embryonic stem cell differentiation.

By looking at endodermal and pancreatic development, the team found that:

• Poised and active enhancers can be predicted by looking at epigenetic state, with active enhancers enriched in H3K4me1 and H3K27ac, and poised promoters defined by just H3K4me1 deposition.
• Acquisition of poised state predicts the ability of cells to interpret and respond to inductive cues.
• Poised enhancers are first recognized by pioneer transcription factors followed by subsequent recruitment of lineage-specific transcription factors.

​

The data presented in this paper suggest that developmental competence is conferred by the establishment of a poised chromatin state at enhancers that allows recognition and binding by pioneer transcription factors.

Find the full paper in Cell Stem Cell​, April 2015.

​​