An ambitious EC-funded research initiative on epigenetics advancing towards systems biology

Artwork

Alix Cleverdon http://alecart.atspace.com

Mixing the DNA of big cats has surprising results depending on who mum and dad are. Ligers and tigons suggest that the same genes can behave quite differently depending on their origin. Some genes are passed down in a silent way. Scientists are hot on the trail of the mum genes we inherit from our parents

Brona McVittie reports

Vulcan and Sinbad. Hybrid cats have a shorter than average lifespan and are prone to illnesses like cancer. No one has directly studied the behaviour of genes in these big hybrid cats, but related research with hybrid mice suggests that genomic imprinting could be responsible, at least in part, for the size difference between ligers and tigons

Photo courtesy of T.I.G.E.R.S. Preservation Stations

Male lions are among the most sexually active of mammalian beasts. In captivity they have even been known to mount female tigers. The resulting liger offspring make their parents look like pussy cats, often exceeding twelve feet in length and doubling parental weight. In contrast, if you mate a male tiger with a female lion, the resulting tigon is considerably smaller. What makes the liger so big and tigon so small? If a lion and tiger mate, why should it matter which is mum and which is dad? Didn't you learn in biology lessons that the genetic contribution from both parents was equal? Similar effects are apparent when you cross a horse with a donkey. The mule born of a mare and a jack (male donkey) is visibly different to the hinny born of a jenny (female donkey) and a stallion. Clearly your parents can have different influences on the way your genes work.

The dogma of the genetic age is undergoing a quiet revolution. We're starting to think less about gene sequences and more about how genes behave in the context of their environment. Bryan Turner (University of Birmingham, UK) reckons that the genome is a bit like a record. You don't play it all at the same time. The controls on your hi-fi allow you to listen to different tracks, and turn up the volume as you please. As you develop, selected parts of your genome (aka genes) get played at different volumes in response to environmental cues. When we make eggs and sperm the volume knobs get reset. When sex cells fuse, and two genomes become one, a lot of reprogramming goes on in the egg. A fertilised nucleus bundles two metres of DNA into a tiny nuclear kingdom a few millionths of a metre wide.

DJ by Oleg Kuvaev – www.mult.ru

Like a DJ, life mixes the genomic records from our parents and the controls on the mixing desk get reconfigured. But mum and dad don't always agree on volume settings. Just like the way parents play different roles in your life as you grow up their genes play different roles during your development in utero. Some genes, when they come from mum, are always silent; the active copy comes from your dad. However, we get silent genes from dad too, and then mum's copy gets to play at maximum volume. A number of these genes facilitate the way that babies grow and develop. Curiously dad seems to switch off genes that would restrict our growth, whereas mum compensates by switching off genes that promote growth.

Around 80 such genes have now been described in mammals, and although they constitute less than 1% of the human genome, their effects are strongly felt. Take for example the IGF2 gene, which makes a protein involved in embryonic development. This growth-promoting factor is silenced by females when they make eggs; you always get a quiet copy from mum. Experiments with mice have demonstrated that babies are born much bigger if mum's IGF2 gene is not silenced. Very high doses of the protein kill developing embryos in the womb. In humans failure to inherit a silent copy of IGF2 causes Beckwith-Wiedemann syndrome. Babies with this disorder are over twice the normal weight. This poses a risk of miscarriage, although modern medical techniques mean that most babies can survive with the condition. Like ligers, however, these individuals have a reduced longevity. If mum's IGF2 gene escapes silencing, certain kinds of cancer may ensue. This is one of the greatest risks for individuals with Beckwith-Wiedemann syndrome.

Foods rich in folate (vitamin B9) including leafy green vegetables, citrus fruits and strawberries, are dietary sources of methyl. Vitamin B12, found in fish, meat, milk and eggs, can donate methyl groups to the metabolism.

Perhaps the most baffling observation about our parents' IGF2 genes is that they have identical DNA sequences, despite one being switched off. Closer inspection has revealed that there are subtle differences beyond the DNA sequences of silent and active genes. They differ in the way that nearby DNA is methylated. This affects which proteins can bind to the DNA and activate genes. Mum's IGF2 gene (not methylated) is silenced by a repressor protein that binds to DNA. Dad's chromosome is methylated near the IGF2 gene, which stops the repressor from binding and turning off Dad's copy. Methyl is a very simple molecule involved in many biological processes. Adrian Bird (University of Edinburgh, UK) and other scientists recognise the importance of this tiny collection of atoms in silencing DNA.

However, silent genes are not a simple matter. There are a variety of means by which the volume knobs of our DNA might be twiddled. Considerable cross-talk occurs between three main players in silencing: DNA methylation, nucleosomes and RNA (see What Neil says). These epigenetic features facilitate a dialogue between the environment and our genetic hard-wiring. The extent to which epigenetic features are heritable is not yet clear, but we are getting a clearer picture of how they are set up. Taking DNA methylation as an example, such décor can result from what your mother ate when she was pregnant. Experiments with agouti mice have shown that feeding methyl-supplements to pregnant mothers can affect the volume settings of genes in their offspring.

The demonstration that nutrients can directly affect DNA is relatively recent news. Although we don't yet know how much our environment shapes gene silencing, there is increasing evidence to show that messing up DNA methylation during development can cause a range of health problems from cancer to schizophrenia. Without doubt, the greatest implication of such heritable epigenetic features is the influence your diet might have on the genes of your children and grandchildren.

Arabidopsis flower

Photo courtesy of Juergen Berger, Max Planck Institute for Developmental Biology, Germany

A closer look at our distant leafy relatives reveals that plants have deployed a similar means of regulating embryonic growth. Ueli Grossniklaus (University of Zurich, Switzerland) discovered a silent gene in a variety of cress, and has named it after the Greek sorceress Medea. In the mythical tragedy Medea killed her children after being betrayed by Jason of the Argonauts, whom she assisted in his mission to steal the Golden Fleece. The Medea gene in Ueli's cress plants makes a repressor protein that silences another gene involved in embryonic growth. Ueli christened this gene after one of Medea's murdered children, Pheres.

The Medea gene is silenced when inherited from the male plant. The active copy from mum serves to keep the Pheres gene quiet in her developing offspring. Without her active Medea gene, seed development goes awry generating small fat embryos reminiscent of mice embryos when Igf2 gene silencing fails. These embryos die when the seeds dry out before dormancy. However, these chubby mutants can be saved by growing them in vitro; adults resulting from such tissue cultures are normal and healthy. The growth moderating effects of the Greek sorceress gene appear to be necessary only during embryonic development. Despite similarities between animal and plant systems in this respect, we don't know whether DNA methylation keeps Medea quiet.

Why have plants and animals employed this habit of silencing genes? The truth is that no-one really knows, although various explanations have been proposed. The inheritance of silent genes, dubbed genomic imprinting, could be related to reproductive habits. In mammals, imprinting is restricted to species that house their developing young in utero. Egg-laying mammals like the duck-billed platypus, are not known to imprint growth-related genes in this way. For nine months your mother provided all the essential nutrients for your development from her bloodstream through the placenta. Rather than laying an egg and leaving you to fend for yourself on the contents of a yolk sac, she patiently responded to your demands for food during your stay in the womb.

 

Seeds develop similarly within the bodies of their parents, but they are fed via an endosperm rather than a placenta. Normal imprinting in both placenta and endosperm are critical to healthy embryonic development. In both species these organs provide the interface for maternal-foetal dialogue and are thus key players in regulating embryonic growth. Wolf Reik (Babraham Institute, Cambridge, UK) has studied Igf2 behaviour in mice. He notes that imprinting of Igf2 is critical for normal placental growth. Knock-out the gene in the placenta and foetal growth is restricted.

Some scientists believe that imprinting evolved in species with a placental-habit to cope with greedy infants that might take more than their fair share of the resources. This school of thought is based on the 'parental conflict theory' proposed by David Haig (Harvard, USA). Mum needs to make sure she saves some strength for herself and your siblings, so her genes have evolved to temper foetal growth. So why do father's genes want you to be bigger? Well, in general babies that are slightly bigger than average have better survival prospects. Of course, it is not really a case of your genes having their own will, but merely that any genes that promote survival tend to get passed on.

Beyond the special case of imprinting, silencing by means of DNA methylation represents the evolution of a primitive immune system. Around half of our genome is silent, made up of nonsensical repeats that don't contain the instructions to make protein. Some of them can hop around in the genome, a behaviour that can have fairly disastrous consequences for the function of neighbouring genes. How did they get there? There are no definitive answers but some look and behave like viruses, well-known stowaways in the genome. Cells seem to mark repeat sequences with methyl to shut them up, which serves to protect our genes. As Denise Barlow (University of Vienna, Austria) has suggested imprinted genes may have arisen accidentally, the result of DNA methylation spreading to other regions of the genome. She points out that this accident must have had some evolutionary advantage to have caught on, although the matter is still open for debate.

Throughout our lifetime and across evolutionary time, silencing DNA has allowed both our cells and us to evolve different habits. Consider for a moment the riotous din of 30 000 human genes were a large proportion not switched off. We don't need to use all 2m of our DNA within every cell. Furthermore, you wouldn't want the cells in your eyeballs to make fingernails. Ligers and tigons show us that the same DNA can have very different consequences depending on how it gets packaged up by their parents. DNA sequences don't paint the whole picture. Their revelation is only the beginning. The blueprint alone holds the potential for life, but life is everything that happens beyond DNA.

Neil Brockdorf
Photo by Brona McVittie

Epigenetic marks regulate the 'open' or 'closed' state for regions of the genome and thereby control the 'on' and 'off' status of genes. A key point is that epigenetic marks can be heritable and provide a means to transmit the 'on' or 'off' state through the process of cell division. Currently we know of three major players: RNA, the nucleosome and DNA methylation, the three pillars of epigenetics. These players talk with one another extensively providing a coordinated orchestration of gene switching essential for making a complex organism.

RNA
Best known for its role as the messenger, transferring genetic information from the DNA to protein manufacturing factories located outside the cell nucleus, RNA is increasingly recognised as a key player in the epigenetic story. At present we know about two types of 'epigenetic' RNA, very small RNAs, called small interfering (si) RNAs, and very big non-coding (nc) RNAs. The siRNAs are involved in establishing the 'closed' configuration at certain sites, notably in DNA repeats found at centromeres and elsewhere in the genome. As for the ncRNAs, some are involved in establishing an 'open' configuration in specific regions of the genome, whilst others function in establishing the 'closed' configuration either in specific regions or even over a whole chromosome. There are examples where transmitting the memory of the 'open' or 'closed' configuration through cell division requires the continuous production of one of these RNAs and in this respect the RNAs can be regarded as epigenetic marks.

The nucleosome
There are four core histone proteins that form the nucleosome, a structure which is used to package the DNA in the nucleus. Histone proteins can be modified at a number of different sites by adding or taking away either small chemical groups, termed acetyl-, methyl-, and phosphate-, or larger 'protein' attachments, termed ubiquityl-. The effect of these modifications is to change the nature of the nucleosome in a manner that affects amongst other things how 'open' or 'closed' the chromatin is. There is evidence suggesting that specific combinations of histone modification can be read like a code, determining for example whether the associated gene should be on or off. This is thought to involve a set of factors that recognise and bind to a given modification present at a specific position on a specific histone. In addition to histone modifications there are a number of 'variant' histones, related to one of the four core histones but with specific properties, for example helping to make a nucleosome more 'open' or 'closed'. Finally there is the linker histone, termed H1, that has an important role in regulating how tightly nucleosomes are packaged. Histone modifications and histone variants are central players in epigenetic processes in all organisms.

DNA methylation
DNA is made up of four different bases that represent the four letters of the genetic code, adenine, cytosine, guanine, and thymine. Sometimes the small chemical group termed methyl- is added to a base, conferring an extra level of information. In higher organisms (i.e not bacteria) methylation is largely confined to the base cytosine. Methylated cytosine is associated with formation of 'closed' chromatin and therefore with switching genes 'off'. This is thought to involve a set of factors that recognise and bind to the modified base. Cytosine methylation is thought to have first arisen as a 'defence' against invading DNA elements, termed transposons. It has since been co-opted as a mechanism for epigenetic gene regulation. An important feature of DNA methylation is that it can be faithfully copied during the process of DNA replication, i.e. when cells double their chromosomes in readiness for cell division. This provides a nice example of how epigenetic information is transmitted from one cell generation to the next. DNA methylation occurs in many, but not all, higher organisms."