Histone acetylation is a particularly intriguing epigenetic mark, in part because it changes dynamically. When histone acetyltransferases (HATs) add acetyl groups to a histone protein, previously coiled DNA opens up, exposing nearby genes for transcription. Histone deacetylases (HDACs) can clip off acetyl groups, turning off genes. Histone acetylation can interact with DNA methylation: for example, methylation-associated proteins can recruit HDACs, leading to gene silencing. HDACs have already garnered intense interest as therapeutic targets. Histone acetylation may have greater biological effects than methylation, according to workshop discussion, but it’s harder to study. Several dozen types of histone acetylation sites exist, since acetylation can take place at numerous lysine residues scattered across many different histone proteins. Most studies look at only one or two sites. Despite the interest in histone acetylation, only a few such studies were discussed at the workshop.

One study, reported by Philip Landfield, University of Kentucky, Lexington, examined the effects of normal aging in the hippocampus in both rats and rhesus monkeys. Landfield found increased gene transcription in the hippocampus of normally aged brains. Most of this increase in transcription seemed to be in glial cells, rather than neurons. In rats, this appeared to be due to a decrease in HDACs, while in monkeys, Landfield saw an increase in HATs (see ARF related news story and Blalock et al., 2003; Rowe et al., 2007; Kadish et al., 2009; and Blalock et al., 2010). He speculated that these shifts in epigenetic regulators might have potential as therapeutic targets in AD. It’s worth noting, however, that Landfield examined only normally aged brains, not brains with dementia. Also, because epigenetic marks vary tremendously by species, these results may not reflect what happens in humans.

An epigenome-wide association study in human AD brains is underway, however. It will use data from two long-running cohort groups, the Religious Orders Study and the Rush Memory and Aging Project, said David Bennett of Rush University, Chicago, Illinois. This project will examine acetylation of lysine residue 9 on histone protein 3 (H3K9 site) in prefrontal cortex, using donated AD and control brains, and correlate the results with genetic and clinical information. Starting next year, the study will examine methylation and mRNA expression in the same brains, as well.

Finally, Stephen Haggarty of Harvard University discussed a potential therapeutic application for histone acetylation. In a mouse model, knockdown of HDAC1 in neurons leads to DNA damage, cell cycle re-entry, and cell death, perhaps because open, accessible DNA is more vulnerable to damage. Overexpression of HDAC1 can rescue these neurons (see ARF related news story on Kim et al., 2008). Haggarty hypothesized that activation of HDAC1 might therefore prevent cell cycle re-entry and cell death in AD. He is currently conducting a high-throughput assay to discover small molecules that selectively activate HDAC1 and can be administered in vivo. Haggarty said his group has discovered an effective in vitro activator that also reduces DNA damage in vivo in mice.

Dealing With a Deluge of Data
One of the biggest challenges facing the fledgling field of epigenetics, the workshop participants agreed, will be the management of the vast amount of data that is sure to be generated in the coming years. Epigenetics is an extension of the Genome Project, and adds yet another layer of data on top of the already complex map of the genome. What’s more, the huge variety of epigenetic marks known to exist make the epigenome more complicated than the genome. These marks vary not only from person to person, but also between tissue types, and even between cells of the same subclass. On top of that, they can change over a person’s lifespan.

Epigenome data must also be correlated with expression data and clinical data. The “omes” keep multiplying, in dizzying layer after layer: the genome, the epigenome, the transcriptome, the proteome, the phenome. Amanda Myers, University of Miami, Florida, used the term “brainome” to describe the interactions of all of these layers in determining the health of the brain. She reported her work on a “Human Brainome” project that seeks to correlate genetic, expression, and protein data in the brain, using a computer algorithm to identify promising networks that might play a role in AD. She predicted this approach might enable scientists to define subclasses of patients and lead to more precise therapeutic approaches, as well as help identify biomarkers.

Given the importance of epigenome mapping, one of the goals of the Roadmap program is to support four multi-institutional epigenome mapping centers, which will produce epigenome maps of human cell types of interest in disease. Initially, these centers are working to produce high-resolution maps for just five cell types. These maps will include genomewide methylation data, the acetylation state of 53 histone sites, RNA sequencing data, and DNAseI hypersensitive sites. The centers will also produce less detailed maps for more than 100 human cell types. Tissues examined include breast stem cells, blood primary cells, pancreatic islets, and various cell lines. The mapping center led by Joseph Costello at the University of California, San Francisco, plans to create brain-related reference epigenomes using cells from selected regions of fetal and adult brains, said Ting Wang, Washington University, St. Louis.

Data from the epigenome mapping centers will be coordinated and analyzed at the Epigenomics Data Analysis and Coordination Center, led by Aleksandar Milosavljevic at Baylor College of Medicine, Houston, Texas. As a start, the center released the first version of the Human Epigenome Atlas on May 14, said Cristian Coarfa, also at Baylor College of Medicine. Successive releases will provide more detailed data.

The challenge of cataloguing epigenomic data goes beyond the epigenome maps, however. Researchers would like to integrate epigenomic data with data generated by the Encyclopedia of DNA Elements (ENCODE), an NIH project that seeks to identify every functional element in the human genome. In other words, ENCODE plans to discover the purpose of the vast majority of human DNA code once labeled “junk.” Since all of this regulatory DNA may contain epigenetic marks, the merging of these databases would be powerful.

Importantly, data must be presented in a format that is meaningful to biologists. Mere lists of numbers are not helpful, a computational biologist at the workshop pointed out. In the group discussion, the University of California Santa Cruz Human Genome Browser repeatedly came up as an example of the kind of searchable database that researchers would like to see for epigenomics. The current genome browser is not adequate for epigenetics questions, however. In particular, a database for AD research would need to include pathological and clinical data. Several participants suggested they would like to be able to visualize the 3D structure of networks, as well. Wang, who works on the browser, suggested that these features might be added.

Participants also discussed the issue of how to share and compare data meaningfully. Since every experiment uses different methodology, comparisons are problematic. Participants agreed that the field is too new to standardize methods, as it’s not yet clear which methods are best. One member suggested creating standard controls for certain experiments to permit comparisons across studies. The group also discussed the feasibility of forming a consortium for AD epigenetics research. In theory, this would facilitate communication and the sharing of data. The consensus seemed to be, however, that it’s too early for that.

Workshop participants agreed that a concerted effort must go into building a computational foundation that can support not only the new field of epigenomics, but also the integration of epigenomic data with pathological data. To quote one workshop attendee, “The human disease side [of AD] is going to blow this whole thing open in terms of complexity.”—Madolyn Bowman Rogers.

This is Part 3 of a three-part series. See also Part 1 and Part 2. View a PDF of the entire series.

Comments

No Available Comments

Make a Comment

To make a comment you must login or register.

References

News Citations

  1. The Rodent in Old Age
  2. Overworked HDACs Leave Transcriptional Posts to Push DNA Repair
  3. Bethesda: Dawn of the Epigenetics Era
  4. Bethesda: The Methylated Brain

Paper Citations

  1. . Gene microarrays in hippocampal aging: statistical profiling identifies novel processes correlated with cognitive impairment. J Neurosci. 2003 May 1;23(9):3807-19. PubMed.
  2. . Hippocampal expression analyses reveal selective association of immediate-early, neuroenergetic, and myelinogenic pathways with cognitive impairment in aged rats. J Neurosci. 2007 Mar 21;27(12):3098-110. PubMed.
  3. . Hippocampal and cognitive aging across the lifespan: a bioenergetic shift precedes and increased cholesterol trafficking parallels memory impairment. J Neurosci. 2009 Feb 11;29(6):1805-16. PubMed.
  4. . Aging-related gene expression in hippocampus proper compared with dentate gyrus is selectively associated with metabolic syndrome variables in rhesus monkeys. J Neurosci. 2010 Apr 28;30(17):6058-71. PubMed.
  5. . Deregulation of HDAC1 by p25/Cdk5 in neurotoxicity. Neuron. 2008 Dec 10;60(5):803-17. PubMed.

Other Citations

  1. View a PDF of the entire series

External Citations

  1. epigenome mapping centers
  2. Human Epigenome Atlas
  3. Encyclopedia of DNA Elements (ENCODE)
  4. Human Genome Browser

Further Reading