A new target for personalized medicine

Interest in epigenetics has taken off over the past few years as molecular biologists have realized the major role it has played in evolution. Indeed, growing evidence that adaptations at the epigenetic level are sometimes passed down to successive generations would explain how species quickly adapt to environmental changes and transfer this information to their offspring. More practically, though, epigenetics also offers new avenues to treat a range of diseases because it adds another layer of gene regulation ripe for therapeutic exploitation.

While the focus of animal, and particularly human, research has been chiefly on the role of epigenetics in disease, many discoveries about the role of the epigenome in adapting to environmental changes have been made in plants. These insights have significance for agriculture owing to their potential to improve yields and enhance tolerance to drought or pests. Plants are also important sources of knowledge about epigenetic adaptation in general because, as they are unable to migrate to greener pastures, they depend extensively on such mechanisms to cope with short-term, but potentially fatal, environmental changes. Research into epigenetics in both plants and animals is now contributing to our understanding of how external factors lead to the establishment of epigenetic marks that alter gene expression on highly variant timescales.

It is in humans, however, where the most progress has been made to identify genes that are modified epigenetically by life experiences, and that alter behaviour or susceptibility to disease. One of the first genes to yield evidence of the epigenetic regulation of behaviour is the glucocorticoid receptor (GR). A study by Gustavo Turecki and colleagues (McGowan et al, 2009) at McGill University in Canada identified a change in the epigenetic status and activity of the GR gene associated with early adversity in suicide victims. There was a clear distinction in expression levels of one particular variant of the GR gene in the hippocampus between those suicide victims exposed to early life adversity—defined as recognized sexual or physical abuse, or severe neglect by parents—and those that were not, with the former producing significantly less GR messenger RNA. “What we do is retrospective, looking at individuals who have died, so we have to have access to their tissues,” Turecki explained. “We go through a process of interview as well as accessing medical files from social agencies that allow us to obtain information about the earlier development of individuals whose brain samples we have obtained.” Their findings replicated earlier studies in rats and indicate a common response to parental, particularly maternal, care, according to the authors.

…epigenetics also offers new avenues to treat a range of diseases because it adds another layer of gene regulation ripe for therapeutic exploitation

The GR gene is expressed differently in various tissues, with a variety of functions, but most relevant for behaviour is its role in mediating stress and response to perceived threats. “The GR gene regulates to a significant degree the activity of the HPA (hypothalamic pituitary adrenal) axis,” Turecki said. The HPA axis controls the stress response and many other processes including digestion and energy metabolism, as well as mood and behaviour. According to Turecki, stress causes a reduction in GR gene expression, which in turn stimulates the HPA axis to increase alertness and hypervigilance. Sustained stress during early years seems to permanently inscribe these changes through an epigenetic mark, in this case DNA methylation—the other major epigenetic mark being the modification of histones. “A GR gene that's methylated produces less GR protein in the hippocampus and the downstream effects are a more hyperactive HPA axis that is constantly activated,” Turecki explained, suggesting that this could be a natural evolutionary response to a perceived hostile environment during childhood.

…drug therapy would not necessarily be an easy option because epigenetic marks are often specific to particular tissues or regions of the brain

Since his 2009 publication, Turecki has discovered that several other variants of the GR gene also appear to be methylated as a result of early life adversity. Most recently, the group has begun to uncover other genes that are epigenetically regulated during early life. “So we are now in the process of identifying specifically other genes that are epigenetically regulated in association with behavioural changes,” Turecki said.

The work begs other questions, as Turecki conceded, such as whether these epigenetic changes are proportional to the degree of adversity, being perhaps greater in people who experienced severe rather than just slight neglect in childhood. This question is hard to answer in humans because of the problem of grading levels of neglect years after it took place. “I think the best way to address that question would be through animal models,” Turecki commented.

But the big question is how permanent these epigenetic changes are and whether they can be subsequently erased, either through behavioural therapy or drugs, which would create new options for treating chronic depression. “This opens the door to our understanding that environmental or traumatic events in life do leave indelible marks,” Turecki said. “The question is to what extent these marks are indelible and to what extent they are dynamic.”

Some are probably more amenable to alteration than others, since some genes that are epigenetically regulated during embryonic or early development are highly stable, whereas others are plastic. It has yet to be established which of the epigenetic changes that seem to be involved in behaviour can be reversed readily. In any case, drug therapy would not necessarily be an easy option because epigenetic marks are often specific to particular tissues or regions of the brain. Turecki therefore thinks that while epigenetic-based therapies hold great potential for personalized medicine, because they could be tailored to the specific state of an individual's genome, they are still a long way off in practice.

There is another reason for caution in treating conditions through epigenetic intervention: it might be simply too late because downstream changes in gene regulation have already taken place in response to the mark. “As an organism develops from embryo to adulthood, a single small change early on in development can become amplified as the system becomes more developed and complex,” explained Dao Ho, a postdoctoral fellow studying epigenetic inheritance at the University of North Texas in Denton, USA. “If you look at things from this viewpoint, then, although epigenetic marks are ‘erased’ after a certain period of time, the effects of the changes induced by the original mark remain permanent.” This means that treating behavioural conditions induced by early life experiences should be done as early as possible. “As we grow older it does become more difficult as our brains become less plastic,” Turecki agreed.

“…although epigenetic marks are ‘erased’ after a certain period of time, the effects of the changes induced by the original mark remain permanent”

Fortunately, other conditions can be treated without potential long-term impacts on the epigenome. If, for example, a tumour cell can be induced to commit apoptosis through an epigenetic change, there is no need to worry about downstream effects, providing the intervention is well tolerated by the patient. Indeed, some epigenetic drugs have been approved that seem to be much safer than traditional drugs that block cell division. The objective is to target epigenetic changes that occur mostly or only in tumour cells. One of the most successful examples is suberoylanilide hydroxamic acid (SAHA) that belongs to a class of drugs known as histone deacetylase (HDAC) inhibitors, which reduce or block the action of the enzyme histone deacetylase. The effect of the enzyme is to condense regions of chromatin and thereby reduce the transcriptional activity; blocking it reverses this effect, turning some genes on or amplifying their expression (Marks & Breslow, 2007).

In October 2008, SAHA was approved for use against cutaneous T-cell lymphoma (CTCL; Lane & Chabner, 2009). While CTCL is a rare disease, SAHA either on its own or alongside other drugs is also being tested in clinical trials against other cancer types. “It also shows good clinical results against mesothelioma and multiple myeloma, and Merck may be considering getting approval for these conditions,” commented Ronald Breslow, Professor of Chemistry at Columbia University, New York, NY, USA. Mesothelioma is a malignant cancer of internal organ linings, caused most commonly by inhalation of asbestos.

Although the precise mechanism of HDAC inhibitors is not fully understood, the effects on the tumour cells are clear enough, as Breslow pointed out. “It induces three results in cancer cells: cytostasis (blocking cell division or growth), cytodifferentiation, and apoptosis. Cytodifferentiation is not lethal [to cells], it simply means converting a cancer cell into a normal one. Apoptosis is of course lethal, but it is well precedented that cancer cells are more easily killed by apoptosis than are normal cells. Most people believe that cancer cells do not have the redundancies that permit normal cells to avoid apoptosis.”

Many anti-cancer drugs work better in combination than on their own, and SAHA is no exception. “A cocktail of drugs hits several different pathways at once,” noted Paul Marks, Breslow's collaborator and head of the cell biology laboratory at Memorial Sloan–Kettering Cancer Center in New York, NY, USA. “In almost all cancers there are multiple defects in the cell, so if you target several at once you have a better chance of a more sustained response.”

Compounds with epigenetic action are also being evaluated for diseases that are caused by the loss of specific cells or cell functions. One such condition is type 1 diabetes, caused by autoimmune attack against the insulin-producing β-cells in the pancreas. One potential application of epigenetics lies in targeting glucagon-producing α-cells in the pancreas, which normally are not destroyed in Type 1 diabetes and often expand in number. The idea is that α-cells can be induced to express the PAX4 gene, which is an essential regulator of β-cell development (Sosa-Pineda, 2005).

Epigenetics instead seems to operate as a separate and temporary layer of adaptation above the genome that does not alter the gene sequence itself

The Juvenile Diabetes Research Foundation (JDRF), the world's leading charity devoted to study of type 1 diabetes, is funding such research. “One of our goals is to identify compounds that can induce PAX4 expression in alpha cells, which could ultimately be useful for the treatment of type I diabetes,” said Stefan Kubicek, Head of Chemical Screening at the CeMM Research Center for Molecular Medicine in Vienna, and a participant in the JDRF network. The researcher's first efforts have been published and demonstrate that α cells can be induced by small molecules to produce insulin (Fomina-Yadlin et al, 2010). “However, the amount of insulin produced is still only a fraction of that made by fully functional beta cells. Therefore, further improvement is necessary before considering therapeutic application,” Kubicek said.

The other major research area in the field of epigenetics is epigenetic inheritance, which does not yet have any impending therapeutic applications. Early on in the research, the discovery that acquired epigenetic changes can sometimes be inherited appeared to revive Lamarckism. But this is a bit misleading, as such a conclusion implies that genes themselves are altered permanently as a result of life experiences. Epigenetics instead seems to operate as a separate and temporary layer of adaptation above the genome that does not alter the gene sequence itself.

The existence of epigenetic inheritance, whereby epigenetic marks in germ cells are passed onto descendants, is now widely accepted, but the evidence is still largely empirical and circumstantial. One of the most studied environmental drivers of inherited epigenetic change in animals has been oxygen deficiency (hypoxia), which has been shown to lead to offspring that are more tolerant of hypoxia in Drosophila (Henry & Harrison, 2004) and in zebrafish. In the case of Drosophila, the adaptation was an enlarged trachea to compensate for reduced oxygen concentration.

The problem with most of these studies though, as Dao Ho conceded, has been the lack of identifiable epigenetic marks associated with the adaptations. So far, such evidence has been confined largely to plants. The Max Planck Institute for Developmental Biology in Tübingen, Germany, has just compiled an inventory of DNA methylation changes in Arabidopsis plants over 30 generations from a common ancestor (http://www.mpg.de/4426136/spontaneous_epigenetic_variation?page=1). The inventory lists a total of 30,000 epimutations observed over that span, defined as a change in the methylation status of a cytosine base. Most of these epimutations were seen to disappear after a few generations, but some persisted for longer. The study also found evidence of spontaneous epimutation, suggesting the possibility that this maintains variability within a population to protect against sudden environmental change. “This idea of ‘bet hedging’ by increasing variance in offspring phenotype in the face of environmental stressors has gained support in the last decade,” Ho said.

Another likely reason for the existence of epigenetic inheritance is in counteracting transposons, which are sections of non-coding DNA that can jump around and insert themselves elsewhere within the genome. Transposons are mutagenic and, in the absence of any control, would quickly wreak havoc in the host genome. Heritable epigenetic mechanisms operating against transposons have been studied principally in Arabidopsis, even though it has a compact genome with relatively few transposons. “The Arabidopsis genome has just 125 mega bases,” said Ian Henderson from the Department of Plant Sciences at Cambridge University in the UK. “Many other plants have genomes comparable to our own (about 25 times bigger with 3 billion bases), with a great deal of repetitive content. In those genomes it may be that epigenetic inheritance plays an even bigger role.”

Another likely reason for the existence of epigenetic inheritance is in counteracting transposons…

This raises the more important question about the extent of the interplay between epigenetic inheritance and DNA mutation in evolution, as Christian Parisod, an evolutionary botanist at the University of Lausanne in Switzerland, pointed out. “It is still debated whether epigenetic marks increase the mutation rate,” he said. “It seems that methylated cytosines might be frequently deaminated (causing point mutation) and this would foster cross-talk between epigenetic and genetic variation. Such molecular link between epialleles and traditional alleles is currently neglected and might turn into an important component of any evolutionary theory including epigenetic marks.”

Such a new unified theory of evolution that incorporates epigenetic adaptation and inheritance is in the making, but it requires much more work to identify the factors that make some epigenetic changes more stable than others. A better understanding of how epigenetic markers themselves are regulated over longer time spans or even between generations would be highly relevant for therapeutic application.