Understanding the role of epigenetics in human variation – let’s start with honeybees?

Painting by Edvard Munch

 

When you next walk up the street, take a look at the faces of passers by.  Even a casual glance will reveal the enormous variation in facial characteristics – hair colour and how the hair grows, the set of the eyes, the size and shape of the nose.  These are just a few of the more obvious differences and they are outward manifestations of the great variation that exists in many human traits.  Each person is a unique product of their own “nature” and “nurture”.  By “nature”, I mean their genetic make up, the set of genes laid down as DNA in their chromosomes.  By “nurture” I mean everything else that influences how we are – our family environment, what we eat, how much exercise we take, whether we smoke and so on.

Identical twins by Diane Arbus

This very evident variation between almost all humans is one reason, I think, why identical twins are so fascinating.  Here we are confronted with two people who look very similar if not identical and this raises all sorts of questions about human individuality.  You will probably tell me that identical twins should look the same as they have the same or very nearly the same genetic make up.  This is indeed true as they derive from a single fertilised egg that divides with the two identical halves implanting and developing as separate foetuses.  If you look carefully, however, you see that “identical “ twins are not identical and they become less alike as they age.  How could this be, given that they have almost identical genes?

Here we enter the realm of a relatively new science called epigenetics.   In our chromosomes we find a set of genes (about 21,000), composed of DNA and each coding for a cellular protein building block.  The genes have, in the past, been likened to blueprints for building cells but if the genes are the only determinants of how we are then identical twins should be truly identical.  Epigenetics tells us that there can additionally be chemical tags added to the genes that change how they are expressed.  It is these epigenetic changes that underpin the differences between identical twins.

There are several different kinds of modifications that contribute to this chemical tagging but the one I want to focus on here is called methylation.  This is the addition of a small chemical tag to part of the DNA in a gene.  This affects the way the gene is expressed leading to changes in the structure of cells and ultimately in human characteristics.  We might liken these epigenetic changes to the pedals on a piano.  The pedals on a piano change the way a note sounds (soft, sustain) and this is a bit like the addition of an epigenetic marker changing how a gene is expressed.

Scientists want to study the mechanisms of epigenetics and how epigenetics contributes to human variation.  Humans are, however, far too complex so they look for tractable “model” systems.  The hope is that the model system will allow scientists to obtain answers to basic questions about epigenetics and the outcomes can then be tested in humans.  One excellent model system is provided by the honeybee and recent research has illuminated mechanisms of epigenetics in this social insect.

Honeybee society is both complex and well structured.  In a colony of honeybees, there will usually be one queen, distinguished by her greater size and, once mated, her ability to lay fertilised eggs.  These eggs grow in to larvae that are nurtured by “nurse” bees.  All larvae are initially fed Royal Jelly and if this continues they will develop in to new queens.  Most larvae are switched away from Royal Jelly to consume nectar and pollen and become worker bees that are smaller than the queen and perform various tasks within and without the hive.  They begin their lives cleaning, nursing etc. in the hive.  Later on they become foragers for nectar and pollen.  The nurse/forager transition occurs in the same insect with the same complement of genes, so it must depend on epigenetics.    Because there is such a profound change in behaviour that must depend on epigenetic changes, the honeybee is a very attractive model for studying the interplay between behaviour and changes in DNA in an organism.

Researchers in the US have recently reported results using the honeybee model to look at epigenetic changes in the brains of worker bees when they switch from nurse to forager and back again.  They tracked the methylation status of genes in the brains of honeybees.  First they took nurse and forager bees from a hive and analysed the DNA in their brains.  Comparison of the DNA methylation signatures of nurse and forager bees identified a hundred or so genes with differences in methylation tags.  Next they wanted to see if this was a reversible difference.  To investigate this, they removed all the nurse bees from a hive so that when foragers returned, some of them reverted to nurses.  When the foragers reverted to nurses the methylation state of just over a hundred genes changed.  About half of these genes overlapped with those identified in the nurse/forager comparison experiment.  For a group of about 50 genes, therefore, methylation state switches reversibly as nurse/forager behaviour switches.  This does not prove causality but it strongly suggests that the epigenetic (methylation) changes are linked to the behavioural change.  This is the first time that a complex behavioural pattern in bees has been related to changes in epigenetic tags.

The next step is to try to analyse the mechanisms behind these effects in honeybees.  The genes undergoing reversible epigenetic change are in many cases genes involved in regulation of other genes.  This tells us that we have a complex network of gene regulation leading to changes in behaviour.  It will be a major task to understand the complexity behind this.

The wider implications of the study relate to the importance of epigenetic changes in humans.  Identical twins are only one example of the effects of epigenetic changes on human behaviour.  It is thought that many human traits including the stress response, mood disorders, learning and memory involve interaction between genetics and epigenetics.

Older fathers and new genetics

[This piece appeared in the October edition of Marshwood Vale Magazine]

Pablo Picasso and children, Claude and Paloma. Their mother was Francoise Gilot.

Couples contemplating childbirth have conventionally focussed on the age of the mother as a risk factor for disorders such as Down’s syndrome.  There has been a growing feeling, however, that the age of the father, previously disregarded, could influence the occurrence of autism and schizophrenia.  A recent study sheds light on the mechanism of this effect as well as showcasing the power of the new genetics. 

The new genetics – the Genome revealed

About ten years ago, the sequence of the Human Genome was reported.  Let’s look at what this means.  Most cells in the body contain a set of instructions that allows new cells to be built.  These instructions, found in the chromosomes, are written in the form of strings of molecules of DNA (DeoxyRibonucleic Acid).  The DNA molecules are organised in to chunks called genes and each gene contains the instructions for making one protein building block.  The Human Genome is made up of about 21,000 of these genes as well as very large stretches of DNA encoding molecules that regulate expression of the genes.  Ten years ago, the sequence of all the DNA molecules (~ 3 billion) in the Genome was determined.  This was a heroic technical and intellectual effort. The Human Genome sequence was called the “blueprint for life” and was expected to lead to huge advances in human health.

The new genetics – the reality

Expansive predictions were made about the effect of the Genome sequence on clinical medicine.  Predictive genetic tests for common diseases such as cancer and heart disease would be available within ten years and new therapies would follow.  None of this has proven to be true.  It had been expected that the genetic basis of these diseases would reside in a handful of changes in the DNA sequence, or mutations as they are called, allowing predictive genetic tests to be developed.  The reality is that many mutations have been identified, each conferring only a small risk for the disease.  Many of the mutations are also in the regulatory part of the DNA, which is not well understood, although it is under intense study.   

The new genetics – whole genome sequencing

Map of Iceland

It became clear that to exploit the power of the new genetics fully, it would be necessary to sequence complete genomes from many people.  Until recently this was impossible for financial and technical reasons.  Now this is going ahead and one of the leaders in this field has been a company deCODE, based in Iceland where it is taking advantage of some of the unique features of this small country.  Iceland has a small population (about 275,000) who are genetically rather similar. This means that mutations in the Genome will stand out from the background variation more clearly than in populations with greater genetic diversity.  Iceland holds genealogical information dating back more than 1000 years allowing inheritance to be tracked between families.  It also holds comprehensive medical records on all its citizens allowing diseases to be followed in the population.  deCODE has so far sequenced the genomes of more than 2000 people and is comparing the sequences from healthy people and those suffering from certain conditions to try to identify disease-causing mutations and has already made important discoveries in stroke, schizophrenia, osteoarthritis and diabetes.

The age of the father at conception

The company has recently compared the genomes of 78 family “trios” of father/mother/child.  The genome of a child contains contributions from both the father and mother and this study examined how maternal and paternal DNA had been changed when it was incorporated in the child’s genome.  The surprising result was that the maternal DNA in the child’s genome contained about 14 new mutations independently of the age of the mother at conception.  The paternal contribution to the child’s DNA, however, contained more mutations and the number increased with increasing age of the father; 20-year old and 40-year old fathers transmitted 25 and 65 mutations respectively.  A woman receives her complement of eggs at birth so it is not surprising that there are few new mutations in maternal DNA.  Sperm, however, result from continuous division of precursor cells; this repeated division, together with environmental insults, results in mistakes (mutations) when the DNA is copied.  In older fathers the sperm precursors will have undergone more divisions with increased risk of mutations compared to younger fathers.

The implications

We now know that the age of the father at conception is the major determinant of new mutations in his child’s DNA.  This surprising finding raises many issues.  First, does the increased number of mutations have any consequence for human health?  The answer here is possibly.  It has been shown in the Icelandic population that as the father’s age at conception increases so does the risk of schizophrenia or autistic spectrum disorder (ASD).  Older fathers transmit more new mutations to their children.   Most of these mutations are harmless, but some may confer increased risk of schizophrenia or ASD.  The risks are still very low but given that age of fathers at conception in the UK has increased from 31.5 in 1998 to 32.4 a decade later, these findings on paternal age are worthy of  discussion more widely in society.  The age of the mother is still an important factor in the occurrence of abnormalities like Down’s syndrome where extra chromosomes are present rather than new mutations.  

Some commentators have wondered if men will now be mindful of ticking biological clocks in the same way women are.  Would young men be sprinting to the sperm bank to preserve their precious non-mutated treasures?  It had always been thought that men were immune to reproductive ageing.  This may now be a myth and we might look differently on the next craggy-faced rock star fathering children in his sixties.

How to make Usain Bolt?

Usain Bolt

I have always wondered why I was such a poor runner.  Now Jonjoe McFadden, writing in the Guardian, has supplied the answer; it’s all down to my faulty gene switches.  It’s too late for me now but according to McFadden, in the future anyone wishing to rival Usain Bolt and run a sub-10 second 100 metres will just take a gene switch drug.   McFadden also tells us that diseases like diabetes, heart disease, cancer and Alzheimer’s disease are all caused by faulty gene switches.  In diabetes, he says, a liver cell may be “genetically tripped to stop absorbing blood sugar”.  Silly me, I had always thought insulin had something to do with diabetes.

McFadden’s musings were occasioned by the recent publication of the results of the ENCODE consortium.  About 10 years ago the DNA sequence of the human genome was reported.  Surprisingly, the part of the genome containing the information for building new proteins, the genes, constituted less than 2% of the sequence.  The other 98% was, at the time, of unknown function and some rather unwisely dubbed it “junk” DNA.  ENCODE set out to study this large part of the human genome that does not code for proteins.  ENCODE’s data rewrite our knowledge and show that much of this misunderstood DNA is functional.  Functional here is a rather broad term and includes several possible mechanisms that can be loosely described as regulating how the genomic DNA is expressed.  This regulation is important for determining why certain proteins are expressed only in certain cell types thus establishing the unique identity of liver cells, heart cells etc.  The regulation may go wrong, and this dysregulation may be at the core of some common diseases.

No one can doubt the importance of ENCODE’s work in rewriting our view of the human genome but it is very important to be clear about the implications.  McFadden uses the term “gene switch” to describe all the regulatory activities outlined above.  This idea in fact derives from the Press Release that accompanied the data.  I find the term gene switch to be misleading as it suggests a mechanistic understanding we do not have.  ENCODE showed, in a variety of ways and in different cell types, that there were potential regulatory functions associated with the non genomic DNA but they did not show how all of these worked.  It will require huge amounts of research to understand the regulatory mechanisms and using a term like gene switch trivialises the present findings and the task ahead. 

McFadden then goes on to build a huge edifice around the idea of gene switches.   Gene switch drugs will in time be developed to counter defects in the regulatory mechanisms that lead to diseases but these drugs will be also be used , he suggests, to manipulate “physiology, mood, intelligence, libido, anxiety, and appetite”, also  to create new Usain Bolts and to stave off the symptoms of old age.  He also states that “many scientists believe ….  that the differences between us and our closest relatives  …. are mostly due to differences in gene switching”.  The corollary of this, he says, is that a chimp might be enabled to talk by treatment with a gene-switch drug.    This wealth of speculation is entertaining but is pure science fiction as it can be neither proven nor disproven at present. 

When the Human Genome was reported, it was accompanied by claims that the information would revolutionise clinical medicine.  The Human Genome has sparked a biological revolution but it has so far had little effect on clinical medicine leading to some disappointment.  The results of the ENCODE project are important and require serious discussion.   Making exaggerated claims about the outcomes means that the real impact of the results may not be appreciated.   Those who read these predictions may end up disillusioned and disappointed when, inevitably, the predictions are not realised. This is bad for science.