24 Jun 2010

Epigenetics, the study of how modifications to DNA and chromatin regulate the activity of genes, is the hot new thing in bioscience. Emerging research indicates that epigenetic regulation may play a role in many complex human diseases, including Alzheimer’s (AD). The AD field is just beginning to grapple with the question of how to handle epigenetic data, and how to integrate it with existing genetic data. To facilitate this process, the National Institute on Aging convened a workshop of about 60 scientists in Bethesda, Maryland, on 7-8 June 2010. Suzana Petanceska, Marilyn Miller, and Tony Phelps of the NIA organized the gathering, called “An Integrated Epigenetic-Genetic Approach to AD.” It brought together AD scientists funded through the NIH Roadmap Epigenomics Program with other researchers working in genetics and gene regulation to share their data, discuss challenges, and promote collaborations.

The five-year NIH Roadmap Epigenomics Program began in 2008 as an “incubator space” for new topics, according to program coordinator John Satterlee of the National Institute on Drug Abuse in Rockville, Maryland. This $175 million program has five components, including an initiative to create reference epigenomes of a number of human cell types, a data analysis and coordination center, projects to encourage new technology development, and the discovery of novel epigenetic marks. The centerpiece of the program, Satterlee said, is the effort to enhance our understanding of human health and disease. To this end, NIH funded 22 research projects examining the role of epigenetics in disease, including four projects specific for AD and cognitive decline. Representatives from all four AD studies spoke at the workshop.

The workshop provided a portrait of a burgeoning field still in its infancy. Epigenetics research is exploding, said Randy Jirtle of Duke University in Durham, North Carolina. Jirtle showed data indicating the field has entered a stage of exponential growth in the last five years, with the number of epigenetics research papers roughly doubling every two years. “Epigenetics will become synonymous with biological research,” Jirtle predicted. Speakers outlined ambitious agendas for extensive mapping and discovery research, but most projects are in their first year, with few data to show yet. One of the main issues that engaged the community was how to manage the flood of data that will be generated, and how to standardize their methods to allow meaningful comparisons among studies. The group also discussed how to integrate epigenetic data into genetics databases, as epigenetics must be viewed in the context of the underlying genome.

The Spotlight Shifts From Genetics to Epigenetics
For many years ApoE4 was the only gene proven to be linked to the common form of non-autosomal-dominant AD, but recently, genomewide association studies and meta-analyses of many association and linkage studies have begun to uncover new AD-associated genes, with several new genes reported in 2009 (see AlzGene for a comprehensive display and Top 10 list). Allen Roses, Duke University, described his work on a variable polyT repeat in the TOMM40 gene, which he found to be related to the age of onset of AD (see ARF related news story on Roses et al., 2009). The effect of most of the newly discovered genes is quite small, however, and some of them do not improve the ability of a model to predict AD, according to Sudha Seshadri, Boston University, which means these genes at present are not useful in diagnosis (see ARF related news story on Seshadri et al., 2010). Geneticists continue to mine for new interactions, using both huge studies, such as the Alzheimer’s Disease Genetics Consortium reported on by Gerard Schellenberg, University of Pennsylvania in Philadelphia, or ADNI, as well as small family studies. For example, Margaret Pericak-Vance, University of Miami, Florida, outlined an approach to find rare genetic variants with strong effects by studying families in which several members develop late-onset AD. Nonetheless, one main realization arising out of the field’s intensive genetics efforts over the past decade is that genetics alone may not provide a full picture of AD heritability.

Into this impasse, enter epigenetics. These are heritable changes in DNA structure, such as the addition of methyl and acetyl groups, that affect gene expression but do not change the underlying DNA sequence. Jirtle compared epigenetics to the “software” that allows cells to access and interpret the information stored in the DNA “hardware”: in effect, a programmable computer within each cell. Because epigenetic modifications control what genes are active in any cell, epigenetic regulation is the primary means of cell differentiation. Each tissue type has a distinct set of epigenetic marks, or pattern of chemical modifications, meaning each person contains over 200 different epigenomes. A particularly intriguing feature of epigenetic regulation is that it acts as an interface between genes and the environment. Epigenetic marks can change over a person’s lifespan, either as part of normal aging, or due to environmental factors such as diet, drugs, pesticides, and disease. Epigenetic dysregulation may affect human ailments as diverse as cancer, psychiatric disorders, addiction, autoimmune diseases, asthma, glaucoma, and dementias, Satterlee said.

Another theme Satterlee and other speakers touched on is the potential of epigenetics to identify better biomarkers for the diagnosis or prognosis of AD. Several presenters discussed plans to hunt for AD biomarkers by correlating changes in AD brains with epigenetic markers in CSF or blood.

Although epigenetics is generally believed to hold promise for the development of novel therapeutic strategies, only one presenter discussed a particular therapeutic application. There was also little use of animal models. As epigenetic marks can vary greatly among species, results from animal models may be difficult to extrapolate to humans, Jirtle said. Most of the research reported at the workshop consisted of broad exploratory studies, seeking to pinpoint epigenetic changes between AD brains and normally aged brains. The attendees noted that they still lack definitive proof that epigenetics plays a significant role in AD, and their first job is to firm up the preliminary evidence. Most of the reported research involved the best-studied epigenetic marks—DNA methylation and histone acetylation—while a couple of speakers talked about a potential role in AD for a different form of genetic regulation, that of non-coding RNAs.

Gene Regulation by Non-coding RNAs
Non-coding RNAs make up the majority of the genome and come from what used to be called “junk DNA.” They play a major role in the regulation of gene expression. Claes Wahlestedt, of The Scripps Research Institute in Jupiter, Florida, described a large exploratory study that is examining changes in non-coding RNAs in the CSF, entorhinal cortex, and hippocampus of AD brains. Initial results indicate that hundreds of non-coding RNAs are increased or decreased in AD brains compared to normally aged brains. Wahlestedt held out the possibility that some of these CSF transcripts could become biomarkers for AD. He also suggested non-coding RNAs might have value as novel therapeutic targets, if scientists find a way to harness their ability to regulate the levels of harmful proteins such as β-secretase (Faghihi et al., 2008 and Faghihi et al., 2010).

MicroRNAs are small, non-coding transcripts that regulate translation of target genes’ mRNA, according to Peter Nelson, of the University of Kentucky in Lexington. Many scientists have observed altered patterns of microRNA expression in AD brains, Nelson said. He described a method to directly identify the target transcripts of some of these altered microRNAs to discover what genes are being up- or downregulated in AD brains. This approach fingered a neurodegenerative disease risk factor gene, progranulin, which is targeted potently—and unexpectedly—by a microRNA (miR-107) downregulated in AD (Wang et al., 2010).—Madolyn Bowman Rogers.

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