Epigenetics articles of the month: January 2016

Need to keep up with the latest epigenetics research? Take a look at our summaries of top articles from the past month.

Epigenomic annotation of gene regulatory alteration during evolution of the primate brain

A comparison of regulatory elements sheds light on primate brain evolution

Differences between the human brain and those of closely-related primate species may explain our advances in cognitive function. However, the adaptive genetic changes that have occurred throughout primate evolution have mainly been to non-coding DNA; inducing changes in gene expression rather than altering the coding sequence itself.

To further understand regulatory alterations during primate evolution, a team led by Menno Creyghton from University Medical Center Utrecht in the Netherlands annotated and compared cis-regulatory elements (CREs) in human, chimpanzee and rhesus macaque brain regions. As chimpanzees are more closely related to humans than rhesus macaque, this allowed the researchers to ask whether changes were specific to human, or occurred earlier in primate evolution. Here is what they found:

  • Rhesus macaque and human CREs are in a similar genomic position but differ in H3K27ac enrichment.
  • Few changes to regulatory elements occurred after human-chimp divergence.
  • Some CREs are either newly introduced or specifically deleted in human compared with macaque, although the majority of these changes occurred before the divergence of humans and chimpanzees.

The authors have demonstrated that a fine-tuning of existing CRE activity occurred throughout primate evolution, but that very few changes have occurred since the diversion of humans and chimpanzees.

Read the full paper in Nature Neuroscience, January 2016.

Trim28 haploinsufficiency triggers bi-stable epigenetic obesity

Evidence of polyphenism in obesity

Obesity is a risk factor for heart disease, type 2 diabetes, stroke and cancer. Although obesity is a growing problem worldwide, there is still much more to learn about the factors that cause it. In particular, we know little about the contribution of epigenetic factors to obesity.

Previously, a highly variable body mass and adiposity phenotype was found in mice haploinsuffient for Trim28, a chromatin interacting protein. To understand more about how Trim28 affects obesity, a team led by Andrew Pospisilik from the Max Planck Institute of Immunobiology and Epigenetics in Germany further characterized this variable phenotype. They found the following:

  • Trim28 haploinsufficient mice displayed a non-Mendelian bimodal distribution for obesity; clustering into an obese and a lean population.
  • Trim28 haploinsufficiency-induced obesity is distinct from overfeeding-induced obesity.
  • A cluster of paternal imprinted genes (termed IGN1) shows marked pathway downregulation in obese Trim28 haploinsufficient mice. Deletion of IGN1 genes Nnat and Peg4 induces bi-stable obesity.
  • Obese human children have reduced TRIM28 levels and dysregulated IGN1 compared with lean children.

In this study, the authors find evidence for the existence of polyphenism in human and mouse obesity. They argue that, if substantiated, these findings will have great academic, ethical and therapeutic impact.

Read the full paper in Cell, January 2016.

Parallel single-cell sequencing links transcriptional and epigenetic heterogeneity

New technique for parallel methylome and transcriptome sequencing in single cells

Recently developed technologies to study DNA methylation in single cells have improved our understanding of heterogeneity between populations. In this study, a team led by Wolf Reik take single cell technologies a step further by describing single cell methylome and transcriptome sequencing (scM&T-seq); a technique that enables the entire methylome and transcriptome to be compared within single cells.

To show this new technique in action, the team applied it to mouse embryonic stem cells to investigate the link between epigenetic and transcriptional heterogeneity. Here are their findings:

  • scM&T-seq yields results that agree with stand-alone single-cell bisulfite sequencing.
  • There is primarily a negative association between DNA methylation and transcriptional state for non-CpG island promoters. However, distal regulatory elements had a balance of positive and negative associations.
  • The strength of the relationship between the methylome and the transcriptome varies between individual cells.

This powerful new approach will enable the relationship between DNA methylation and transcription to be probed in more detail in heterogeneous cell populations. Using this technique, the authors have highlighted the complexity of the relationship between the methylome and the transcriptome.

Read the full paper in Nature Methods, January 2016.

Transcriptome-wide distribution and function of RNA hydroxymethylcytosine

Drosophila brain development requires hydroxymethylcytosine modification of RNA

Addition of methyl and hydroxymethyl groups to cytosine is a well-described DNA modification. Although cytosine modification in RNA also occurs in the form of hydroxymethylcytosine (hmrC), the distribution, localization and functional relevance of this modification remains unknown.

A team led by François Fuks from Université Libre de Bruxelles in Belgium characterized this RNA modification in Drosophila melanogaster using an adapted form of methylated RNA immunoprecipitation and sequencing (MeRIP-seq).

  • hmrC peaks are frequently within coding sequences and over-represented in UC-rich sequences.
  • mRNA transcript abundance is positively correlated with the presence of hmrC peaks, and hmrC abundance is correlated with active mRNA translation.
  • dTet – an enzyme responsible for depositing hmrC – is expressed in the larval brain along with elevated hmrC levels, suggesting a role for hmrC in brain development.
  • dTet loss-of-function flies display morphological defects in the brain during development.

These results highlight a role for hmrC in regulating translation of certain mRNAs, and demonstrate the importance of the RNA modification in fruit fly brain development.

Read the full text in Science, January 2016.

The H3K4me3/2 histone demethylase RBR-2 controls axon guidance by repressing the actin-remodeling gene wsp-1

Demethylation of histone 3 is required for normal neuronal development

Trimethylation of lysine 4 on histone H3 (H3K4) is enriched at the transcriptional start sites of actively transcribed genes. Mutations in enzymes that deposit (KMT2 proteins) and remove (KDM5 proteins) H3K4 methylation marks are associated with neurological disorders.

Relatively little is known about the biological functions of KDM5 proteins apart from their catalytic activity. In this paper, a team led by Anna Elisabetta Salcini from the University of Copenhagen, Denmark describe the role of H3K4 demethylation in Caenorhabditis elegans neural development. To do this, the team looked at the RBR-2 protein; the sole C. elegans homolog of KDM5.

  • RBR-2 regulates expression of the WSP-1 actin regulator by decreasing H3K4me3 at the transcriptional start site and repressing expression.
  • Loss of RBR-2 results in increased WSP-1 expression and defective axon guidance.
  • NURF-1 – a H3K4me3-binding protein – is required to promote aberrant wsp-1 transcription in rbr-2 mutants.

In this paper, the authors have established a precise role for epigenetic regulation in neuronal development. They demonstrate a functional link between RBR-2 activity, H3K4me3 levels, the NURF complex and expression of WSP-1.

Read the full paper in Development, January 2016.

Response and resistance to BET bromodomain inhibitors in triple-negative breast cancer

Triple-negative breast cancer cells show sensitivity and resistance to BET inhibitors

Triple negative breast cancer (TNBC) is an aggressive form of cancer, but no targeted therapies are currently available. Inhibitors of BET (bromodomain and extraterminal domain family) proteins have shown effectiveness in other forms of cancer. These molecules inhibit oncogenic pathways by preventing BET proteins binding to chromatin.

Despite showing promise in certain cancers, BET inhibitors have not previously been evaluated in TNBC. In this paper, a team led by Kornelia Polyak from Dana-Farber Cancer Institute and Harvard Medical School in Boston show that TNBC cells are sensitive to BET bromodomain inhibition and assess the mechanisms that lead to resistance. Here are their findings:

  • TNBC cell treatment with the BET bromodomain inhibitor JQ1 induces morphological changes consistent with senescence, and inhibits tumor growth.
  • Long-term cell culture results in TBNC cells that are resistant to BET inhibitors.
  • Resistance to BET inhibitors is not dependent on drug export, changes in bromodomain-encoding genes or activation of parallel pathways
  • BET resistant TNBC cells are dependent on wild-type BET protein BRD4 and display increased BRD4 binding to superenhancers, indicating that a bromodomain-independent binding mechanism may be occurring.
  • Resistance is dependent on increased BRD4 binding to MED1 in a bromodomain-independent manner that is enhanced by BRD4 phosphorylation.

This paper shows that although TNBC cells are sensitive to BET inhibition, resistance develops over time. The authors find that resistance is dependent on BRD4 phosphorylation-dependent binding to targets and present a rationale for combination therapy to combat resistance.

Read the full paper in Nature, January 2016.

Chromatin topology is coupled to Polycomb group protein subnuclear organization

Polycomb protein clustering impacts long-range chromatin interactions

Polycomb group (PcG) proteins assemble into multiprotein complexes to modify and compact chromatin and repress gene expression. However, how these proteins and their chromatin substrate are organized in the nucleus is currently unknown.

By combining a variety of different methods, including super-resolution microscopy and chromosome conformation capture, a team led by Nicole Francis from Institut de Recherches Cliniques de Montréal in Canada studied PcG protein subnuclear organization and its impact on chromatin topology and gene expression. The found the following:

  • Polycomb (Pc) and Polyhomeotic (Ph) components of Polycomb repressive complex 1 (PRC1) form nanoscale clusters in the nucleus.
  • Organization of PRC1 into clusters is dependent on the polymerization ability of Ph SAM (sterile alpha motif) domains and the overall quantity of Ph.
  • Long-range chromatin interactions are increased when clustering is enhanced, and decreased when clustering is disrupted.

In this study, the authors have shown that PcG proteins from small clusters that impact long-range chromatin interactions. They suggest that these architectural effects may contribute to the efficiency of PcG-mediated gene expression.

Read the full paper in Nature Communications, January 2016.