Conference Scene: Epigenetic regulation: from Mechanism to Intervention

Abstract

The Medical Research Council Clinical Sciences Centre Symposium on Epigenetic Regulation: From Mechanism to Intervention in London, UK, which was held on 20–22 June 2012, attracted 305 participants from around the globe and included 37 speakers and 85 selected poster presentations. The organizing committee, led by Niall Dillon of the Medical Research Council Clinical Sciences Centre (London, UK), consisted of several distinguished researchers in the fields of epigenetics and chromatin organization from across the UK. The meeting covered a diverse range of topics and brought together scientists carrying out fundamental research on epigenetic mechanisms and also researchers who are exploring the role of epigenetics in human diseases and its clinical applications. In addition, the meeting highlighted some emerging aspects in the rapidly evolving field of epigenetics.

Epigenetics is the study of change in gene expression or cellular phenotype that is caused by covalent modification of DNA without altering the sequence composition [1]. The meeting provided new insights into mechanistic aspects of epigenetic phenomena, how this knowledge enhances our understanding of common human diseases, factors that could alter the dynamic epigenome and potential therapeutic targets. Selected highlights from the meetings are summarized in this article.

Nuclear dynamics & the role of different protein complexes in shaping the epigenome

The first lecture of the meeting was given by Danny Reinberg (New York University School of Medicine, NY, USA). His group has previously shown the association of Jarid2 (which encodes a nuclear protein essential for embryogenesis) with PRC2, resulting in an enhanced PRC2 activity at promoter regions in embryonic stem (ES) cells in vitro[2]. He showed Jarid2 binds to ncRNA and also to the upstream of the Dlk1–Gtl2 imprinted domain and these associations of Jarid2 were positively correlated with the presence of an H3K27me3 mark. These findings provided new insights into the role Jarid2 has in nuclear organization events. Bas Van Steensel (Netherlands Cancer Institute, Amsterdam, The Netherlands) highlighted the interaction of nuclear lamina with lamina-associated domains (LADs) [3,4] in individual live cells employing a new microscopic approach. Many of the LADs, also occur inside the nucleus, especially in daughter cells at the G1 phase of mitosis, and the presence of active histone marks (e.g., H3K14ac or H3K27ac) at these regions, provides a good idea of whether they are transcriptionally active or not. Studies in mouse and human cells unveiled that the LADs are highly conserved and the LAD–nuclear lamina interactions are very labile and stochastic to an extent. These results furnished new evidences for dynamic spatial organization of the genome. Bradley E Bernstein (Broad Institute of Harvard and MIT, MA, USA) highlighted his group‘s longstanding efforts to characterize chromatin state dynamics and the regulatory function of different chromatin states [5]. Their current approach is to apply the hidden Markov model to high-throughput sequencing data to scan the genome, identify interactions between the genomic elements, model and then characterize the most frequent combinations. More than 100,000 candidate enhancers were identified; 10% of which are within the close vicinity of protein-coding genes and are cell-type specific. Further work will document the role of enhancers and its regulatory roles in driving tissue-specific gene expression.

External factors influencing the epigenetic state of the genome

Emma Whitelaw (Queensland Institute of Medical Research, Queensland, Australia) presented results from their N-ethyl-N-nitrosourea mutagenesis screen in mice to identify genes that modify epigenomic state [6]. The group has now screened 5000 G1 lines, and identified 50 mutants called modifiers of murine metastable epialleles dominant (MommeDs). More than 20 underlying mutations including some novel cases were identified and these mutant lines will be a valuable resource to study epigenetics–environment interactions. However, it will be important to correlate these interactions with gene-expression changes and address their transgenerational stability.

Michael J Meaney (Douglas Mental Health University Institute, McGill University, QC, Canada) and his group have demonstrated that maternal care influences hypothalamic–pituitary–adrenal function by mediating epigenetic changes of glucocorticoid receptor expression [7]. At the meeting he demonstrated that a similar mechanism operates in Grm1 (which encodes type I metabotropic glutamate receptor or mGluR1) high-performing mothers (more frequent pup licking and grooming [high LG]) that showed increased mGluR1 expression, which was associated with a decrease in methylation levels in the exon 2 region of the Grm1 gene. He also showed evidence that this behavior affects reproductive function in females and is inherited by the next generation – that is, lactating females from high LG mothers also show increased pup licking and grooming. It appears that increased estrogen receptor α (ERα) expression in high LG mothers causes this effect and the differences in ERα expression were associated with changes in the methylation profile at exon 1B of the ERα promoter as well. It will be interesting to translate the concept of behavioral changes altering the epigenome and transgenerational inheritance of the trait in humans experimentally.

Linking mechanism to disease intervention

Peter A Jones (USC Norris Comprehensive Cancer Centre, CA, USA) described a new methodology of mapping nucleosomal positioning and DNA methylation concurrently to assess alterations in nucleosomal positioning during abnormal gene silencing by promoter hypermethylation. Their efforts to treat cancer patients with 5-azanucleoside have been effective so far and results suggest that the effects are mediated by nucleosomal repositioning [8]. He also opined that it is not only about promoter or transcription start site (TSS); gene-body methylation, could be functionally important as well [9]. Manel Esteller (Bellvitage Biomedical Research Institute, Barcelona, Spain) focused on their continuing effort in global epigenetic profiling of 1500 primary tumors, preferring the 450K microarray (Illumina, CA, USA) technique to generate comprehensive methylomes with low cost and effort [10]. There is good evidence for DNA methylation changes outside the promoter CpG islands of tumor suppressor genes and he demonstrated a growing list of mutated chromatin remodeler genes that contribute to tumorigenesis. This knowledge could be applied to a clinical situation to select biomarkers or diagnose primary cancers, and possibly treatment with epigenetic drugs in future.

Tim Spector (Kings College London, London, UK) showed interesting data from a large-scale monozygotic twin study including 1500 pairs (1000 of them are disease–discordant). The work identified 490 age-related differentially methylated regions [11,12], and ongoing research aims to identify differential methylation profiles in diabetes for discordant twins. In conjunction with the Beijing Genomics Institute (Beijing, China), a total of 5000 adult twin methylomes (EpiTWIN study) will be sequenced and then expression studies will be performed on the same individuals [13]. This large-scale study will be a great resource to address whether epigenetic events confer phenotypic variation in twins and if nongenetic factors influence the changes [14]. This study will also inform the role of DNA methylation changes in disease discordance, which will also be of high therapeutic interest. To this end, Art Petronis (The Krembil Family Epigenetics Laboratory, ON, Canada) further illuminated on monozygotic twin discordance in major psychiatric diseases and the possibility of identifying disease-specific epigenetic traits. Jonathan Mill (Kings College London) described intra- and inter-individual difference in methylation profiles across whole blood and several regions of human brain [15] and elucidated how these data can be of value in identifying differential methylation patterns in neurological diseases such as autism, schizophrenia and Alzheimer‘s disease.

Emerging aspects & conclusion

Two new emerging aspects of modern epigenetics were also discussed in the meeting. The first one is the ‘sixth base‘ in the genome, 5-hydroxymethylcytosine (5-hmC) and 10-11 translocation (TET) family proteins. Wolf Reik (Babraham Institute, Cambridge, UK) focused on active demethylation patterns during the early stages of epigenetic reprogramming in the cell. Interestingly, imprinting control regions and a quarter of promoter-associated CpG islands (CGIs) demethylate later than other regions in the genome. He indicated that demethylation is probably indirect with cytosine methylation events and is important in attaining pluripotency but not necessarily for maintaining it. It is possible that 5-hmC plays an important role in the demethylation events (reaction mediated by TET3 and TET1 hydroxylase family) in the genome. The group has recently devised a method (oxidative bisulfite sequencing) for mapping 5-hmC at single-base resolution for the first time [16], and aims to elucidate the role of hydroxymethylation in reprogramming of the epigenetic state in zygotes and ES cells and also whether these events interact with other pathways (e.g., deamination) to generate a cumulative effect on erasure of DNA methylation at early stages of life.

Kristian Helin (University of Copenhagen, Copenhagen, Denmark) presented genome-wide mapping and functional analysis of TET1 and TET2 protein and hydroxymethylation. He stated that throughout the genome TET1 and TET2 have the highest density around the TSS of genes and significantly higher density at CGIs [17]. Loss of TET proteins leads to local but not global increase in DNA methylation and a very few transcriptional changes. Furthermore, he speculated on the role of TET proteins in protecting CGIs from being methylated and that in cancer, TET fails to generate 5-hmC, which probably contributes to aberrant methylation of CGIs. However, the question remains that if TET proteins are present at high density around TSS, but loss of them does not cause significant change in transcriptional activity, then what are the other functions that could they be involved in? In addition, the hypothesis of TET preventing CGI methylation might not be as straightforward and future work needs to be carried out to reach a conclusive answer.

The second aspect was the role and function of ncRNA in effecting gene function at a specific locus. Richard Jenner‘s (University College London, London, UK) group recently identified a class of short (50–200 nucleotide) ncRNAs transcribed from the 5´ end of PRC2 genes in the absence of mRNA transcription [18]. The prevalence of these ncRNA are higher in TSS-associated CGIs, which are largely unmethylated, suggesting they probably regulate PRC2 genes targeted by CGIs. Furthermore, he demonstrated that the PRC2 subunit Suz12 directly interacts with these ncRNAs and other RNA molecules in the cell. These results provide a new layer of understanding about RNA-mediated epigenetic changes and trigger new possibilities of studying the role and mechanism of ncRNAs in regulating DNA methylation and also whether DNA methylation regulates RNA transcription.

The plenary lecture was given by Richard Young (Whitehead Institute, MA, USA and MIT, MA, USA) who showed the importance of understanding different regulatory elements, for example, enhancers, transcription factors and chromatin regulators that together control gene transcription in ES-cells as well as in tumor cells. He focused on a specific example of c-Myc, which is bound to a large proportion of enhancers for the majority of the active ES-cell genes (increased c-Myc level demonstrated increased transcriptional activity of 750 genes) and has a different target signature in different cell types. c-Myc plays an important role in deregulation of ES cells for differentiation and is overexpressed in 40% of cancers [19]. He suggested that the cooperativity phenomena (binding of one protein enhances binding of another protein) leads to additive effects that regulate gene expression. The complete understanding of this multi-dimensional mechanism will be important to finally develop effective therapeutics for human diseases.

Acknowledgements

The author would like to thank the organizers and administrators of the conference for a stimulating meeting and great organization. The author would also like to gratefully acknowledge the support and encouragement provided by IM Morison for this work.

Financial & competing interests disclosure

The author is grateful to the National Research Centre for Growth and Development, Auckland, New Zealand and the Department of Pathology, Division of Health Sciences, University of Otago, Dunedin, New Zealand for providing funding to attend the conference. The author has no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.