Australasian Science Magazine

By Dyani Lewis

Geneticist Dr Marnie Blewitt explains how epigenetics makes us more than just our genes and how gene inactivation can be crucial to our development.

DYANI LEWIS
Hi, I'm Dyani Lewis, thanks for joining us. Our genetic make-up determines a lot about who we are. It determines whether we have blue eyes or brown, what blood group we are or whether we're a carrier of a condition like cystic fibrosis. But we're beginning to learn that we're far more than the sum of our genetic parts. Our genes only tell part of the story of who we are. Just as important as what genes we've inherited from our parents is how those genes are switched on and off throughout our lifetime. This complex system of genetic regulation has been the focus of the burgeoning field of epigenetics. To tell us all about epigenetics and its implications for our growth and development, I'm joined on Up Close today by geneticist, Dr Marnie Blewitt. Marnie heads a lab that studies epigenetics at the Walter and Eliza Hall Institute of Medical Research. Welcome to Up Close Marnie.

MARNIE BLEWITT
Thanks Dyani, it's nice to be here.

DYANI LEWIS
Marnie, epigenetics seems to be quite a new field of study in the broader field of genetics. When did people start to become aware that genes weren't the whole story?

MARNIE BLEWITT
Well it was back in the 1940s actually when they first coined the term epigenetics. The prefix epi was because they started to first realise that it was the same genetic information which was contained in all of our cells. And yet we know that we have so many different cell types so they really just used this term to say, there must be something, a layer of information on top of the genetic information. So like the prefix epi you use in epidermis, and it's just a layer of skin on top of everything else, this layer of information in addition to the genes. So it was actually a long time ago that they knew that this was true but it still is actually a recent field that's been growing and expanding, when we've been able to start to put together what's happening really down at the molecular level inside the nucleus. So what we actually mean by epigenetics rather than just talking about, I guess, phenomenon.

DYANI LEWIS
Many people have commented that the discovery of epigenetics is like a modern day revival of the ideas of the French biologist, Jean-Baptise Lamarck, who was a predecessor of Darwin in thinking about the process of evolution. Is epigenetic the same as Lamarckism?

MARNIE BLEWITT
It's a very common comment. I think the key difference is that Lamarck proposed that a giraffe might grow a neck that was longer and longer, would grow in its own lifetime a neck that was longer to be able to reach leaves that were high on a plant. But the difference with epigenetics is that it's still about Darwinian aspects I guess. So Darwin would say, actually there was selection. So you had a range of different animals, different giraffes and some had long necks and they're the ones that could survive. So epigenetics is the same, these epigenetic marks get laid down, they get set very early in life. Usually while you're still in utero, depending on which organ we're talking about, certainly for neck length while you're still in utero. And then there's selection similarly for the fact that there might be genes that are being expressed to promote additional vertebrae in the neck, for example in the case of a giraffe.
So it really works the same way as genetics in that sense and quite different to what Lamarck proposed which was that you would, throughout your lifetime, obtain some additional characteristics. There's a very common confusion but there's still actually quite a clear distinction between the two.

DYANI LEWIS
The similarity that is drawn is that throughout our lifetime various things in our environment can actually change our genetics, that would be the…

MARNIE BLEWITT
Yes.

DYANI LEWIS
…similarity. But even that doesn't really hold up as a good comparison?

MARNIE BLEWITT
I think, while there's some evidence that these epigenetic marks, this epigenetic information, the information in addition to your genes, might be able to be passed through to the offspring, this is the difference, is that genetic traits are inherited to the next generation. That's why a giraffe with a long neck will give rise to another giraffe with a long neck. But epigenetic changes that happen, and are applied during your lifetime because of the influence of the environment, are basically never inherited or very, very, very rarely. We know about a small handful of cases just in some model organisms, so we don't think yet that we can inherit whatever our parents acquired in their lifetime, that that can be passed on to the offspring. At least we don't know very much about it yet if it does occur.

DYANI LEWIS
In the introduction I described epigenetic as switches that switch genes on and off. What does this actually mean at the DNA level? How do epigenetic factors actually switch genes on or off?

MARNIE BLEWITT
Essentially we don't have all of the answers yet but the epigenetic switches are really about how accessible that DNA is to the proteins that need to be able to get in there and make that DNA all the way through into the protein product. So the DNA needs to be accessible and you need to have proteins. RNA preliminaries come along and make an RNA, so a messenger molecule, and then a protein can be made, say for example an enzyme that's required by the cell. So some epigenetic marks correlate with when that chromatin or the DNA is very densely packed and therefore it's less likely to be accessed by this machinery that makes the proteins in the end. Or it's very open and sparse, there's a lot of space around the DNA for these proteins to actually get in there and do the job of making that messenger that I said, the messenger RNA.
So there's a correlation, how one leads to the other, how these marks actually lead to the chromatin being open or closed, is still an open question. We don't really know whether these epigenetic marks lead to the chromatin being open or closed or whether the closed nature or the open nature of the chromatin leads to the epigenetic marks. It's a bit of chicken and egg question that still exists in the field.

DYANI LEWIS
But in any case we can't really think of the genome as just being a long string of letters that's equally accessible to be read at any given time.

MARNIE BLEWITT
No definitely not. There are definitely regions that are there to be used and are really active all the time and openly accessible. Other regions which are densely packaged down and are not being used, if you like, are not being made in to their protein products.

DYANI LEWIS
How are these different tags or marks actually put onto the DNA or the chromatin?

MARNIE BLEWITT
So they're added by a set of proteins that we call epigenetic modifiers. That really is just to describe the way that some proteins which lay down these particular epigenetic marks, they're enzymes. There can be others that remove them and there can be others again that read those marks, if you like, and interpret the information for the genome. So they might read the information and then lead to some other secondary set of epigenetic marks or move the chromatin around, move the chromosomes around even. And so those epigenetic modifiers are the ones doing this work of laying down those marks or removing the marks.

DYANI LEWIS
Is there just one type of mark?

MARNIE BLEWITT
No so there are probably more than we even know about, hundreds of marks. There are probably only a handful that we've studied very much. So we know that there's methylation to the DNA itself, so that's the addition of a carbon and three hydrogens. But there's also methylation of the proteins that the DNA wraps around, the histone proteins. We can think of many, many different chemical moieties and these all happen. The reason I say there are hundreds, I could just list maybe a dozen of different chemical moieties, is because within the actual proteins, the histone proteins that package the DNA, they have an inordinate number of residues that can actually be modified. So they're not just that it can be one form that exists there, but rather maybe about 20 or 30 different places on each of those molecules that could be methylated or phosphorylated or ubiquitinated or some other small chemical moiety which is added on to the protein.

DYANI LEWIS
In terms of these different enzymes that are doing this modification work of the DNA or of the packaging material…

MARNIE BLEWITT
Mm hm.

DYANI LEWIS
…for the DNA, do these enzymes target specific sites in the genome, like specific genes, to switch off target genes? Or are there other factors that control which bits of the genome are silenced or activated?

MARNIE BLEWITT
So again I think this is a bit of a chicken and egg answer unfortunately Sue. So it depends, you're not going to have one enzyme which is acting throughout the whole genome. There will be ones that are more specialised for particular regions. But what makes them specialised for that particular region? How is that they know to get to just a particular set of genes for example? That's where we don't really know yet the answer. We think sometimes it's through the action of a different class of molecules called transcription factors. These seem to bind to particular regions, particular base sequences within the DNA. Other times we think that might be the role of RNA molecules, so a non-traditional functional molecule in the cell.
It seems that in that case they seem to guide these epigenetic modifiers to particular sites. But there certainly isn't one methyltransferase for the genome for example. There are many different - so there might be 20 different methyltransferases just for one particular residue on one of those packaging proteins. So they've got many different roles. They'll have different roles at different regions of the genome but probably different roles in different stages of development, different cells within the body. So some of them might be important in blood cells and others important in the brain for example.

DYANI LEWIS
If we weren't able to regulate our genes in this complex regulatory way, what sort of process would be disrupted?

MARNIE BLEWITT
Essentially everything. As far as we know at this point, every single process that we biologically know about has an epigenetic component to it. Because as a multicellular organism we need to be able to switch genes on and switch genes off. If you can't do that, if you have expression of genes that shouldn't be around at that time, actually that's synonymous with cancer. We know that if you have the expression of a gene that promotes cell growth, that's one of the hallmarks of cancer, just as one example. So every aspect that you can think about of normal biology is disrupted, from fertility to placentation in the embryo, to development of the embryo, to development of every single cell type that they've come across. Of course when it goes wrong it results in tumours as well.

DYANI LEWIS
This is Up Close, I'm Dyani Lewis and in this episode we're talking about epigenetics with geneticist, Dr Marnie Blewitt. Marnie, one of the most stunning illustrations I think of epigenetics is a process called X-inactivation which occurs in baby girls. What is X-inactivation and why does it happen?

MARNIE BLEWITT
X-inactivation is - yes it's definitely probably the best characterised epigenetic process that we know about in mammals. It happens because female mammals have two X chromosomes and males have one X chromosome and one Y chromosome. So if this was left unchecked, if you like, then all those genes that are encoded on the X chromosome, females would have twice the dose of them. So they'd have two chromosomes making all of those copies of proteins. But this isn't what happens actually and, if it does, it results in embryonic lethality. So we know that can't occur. Instead what happens is that females, just after implantation when there are maybe a couple of hundred cells, a choice is made about one X chromosome in each cell that shall be silenced, either the one that was inherited from the dad or the one that was inherited from the mum.
That X chromosome is epigenetically silenced. So it's actually packaged up very tightly in the nucleus and literally put to the edge of the nucleus. So it's pushed out to the edge and essentially forgotten about, in terms of gene expression anyway. That X that's inactivated now leaves just one X chromosome that's active and that's just like a male cell. So now we have just a single X chromosome that is being used in a female mammal and so we don't have an overdose of those X-linked genes.

DYANI LEWIS
Is it always the same X chromosome from the mother or the father that's inactivated?

MARNIE BLEWITT
No, so when the choice happens, when you've got maybe 100 or 200 cells, it does happen at random. So you'll end up with say 50 cells that inactivated the X from the dad and 50 cells that inactivated the X from the mum. One of these really important features of epigenetic control is that choice and that epigenetic silencing is then heritable for every cell division afterwards. Women can live up to 100 years or so, so for 100 years, for all of the cell divisions that happen in that time, that's faithfully maintained. That means that then if you could actually look at women, we're mosaics, we have patches where we are only expressing the X chromosome we inherited from our mum and other patches where we're only expressing the X chromosome we inherited from our dad. So if we could see that in terms of, I don't know, the colour of our skin, you'd have black and white patches, around a hundred of them, spread across your body.

DYANI LEWIS
So what scale do these mosaic patterns happen? I mean would we have, for example, an entire organ that just has our paternal X chromosome active and other patches, other entire organs, that have only the maternal? Or would it be within an organ you have patchy bits?

MARNIE BLEWITT
So it can be both of those things. It really depends because, while it's a random choice that's originally made, because it is random it depends on how gastrulation occurs, which is when the embryo folds in on itself and starts to form organs and then later organogenesis. So it really just depends on which cells ended up forming which organ. Certainly there are instances where you can see a whole organ is completely derived from a case where you've silenced the paternal X or the other way round. But you also have organs, more commonly, when there's actually a mixture of the two and it really depends on how those organs are first defined from their cellular source I guess.

DYANI LEWIS
There are lots of so called X-linked conditions, such as colour blindness where you've got a gene involved that's located on the X chromosome. These conditions are usually vastly more common in males who have only one X chromosome, but females are usually okay because the idea is that you inherit one faulty gene but then you've always got the back up from the other parent. So how does this work if one of those X chromosomes is being shut down?

MARNIE BLEWITT
Yes so it's really interesting. It depends, I think, on the different scenarios, how that phenotype type or how that disease might be formed. If it were a disease where it was about a blood clotting factor for example - haemophilia would be the best example - so then because those clotting factors are made by cells, in the bone marrow actually, and then they're released and they need to be going out and actually transitioned around the body in the blood, if only half of the cells are able to make the normal wild type or normal functional copy of the clotting factor, it's okay because the cells that can do it, they distribute it throughout the body. In which case therefore, perhaps half as much good clotting factor is enough. So you don't really see females with haemophilia.
But in other cases, that cell may not be able to benefit from the fact that the neighbouring cell can make normal whatever it might be. That's why you end up with some X-linked disorders where females are completely unaffected and some where actually the females are affected but they're just less so than the males because one of the reasons is half of the cells will be fine. So it just depends on the disorder and how that particular organ that's affected is made up and, I guess, the physiology of that organ.

DYANI LEWIS
Your interest in X-inactivation came from work on a particular gene that you found in mice called smchd1.

MARNIE BLEWITT
Yes.

DYANI LEWIS
Or “smoochdee1” is that how you call it?

MARNIE BLEWITT
Exactly, we get a bit lazy so we like to call it “smoochdee1” for short, yep.

DYANI LEWIS
So could you tell us a bit about smchd1 and what happens to female mice that don't have a functional smchd1 gene?

MARNIE BLEWITT
Yes I first found it, we were doing a genetic screen where we made mutations randomly throughout the genome. Then we happened to find this one through a particular phenotype we were looking for. Then what we found is that when we had mutations in this gene, smchd1, and when we made embryos that carried two copies of this mutation, while the male embryos were fine, we didn't have any female embryos surviving. So we went back to look about when they were dying and they were dying around mid-gestation. They failed to make a proper placenta and the embryo itself was also malformed, although it's very difficult to tease out why they're malformed. Is it because they don't have a functional placenta or is it because the embryo also has issues?
But yes around mid-gestation a female embryo with no smchd1 will have a defective placenta and it will also, if you check at the expression of all of those X-linked genes, which should be only being expressed from one of the two X chromosomes as we just talked about, then they'll have double the dose of all of those X-linked genes. So an overdose of a lot of X-linked genes and we think this is what's really responsible for the death of those embryos.

DYANI LEWIS
So there's no X-inactivation taking place?

MARNIE BLEWITT
No. We think what happens with smchd1 is that it's not involved in initiating or not involved in setting up the choice of which X chromosome is silenced or indeed making that silencing choice, the epigenetic silencing in the first place. But we think what it's more involved in is the maintenance of that silenced state. So, as I said, X-inactivation, that choice is made at about the 100 cell stage and then it's faithfully inherited all the way for decades and decades to come. But within about two days of this choice and the silencing, this epigenetic silencing being set up, if you don't have smchd1 you can't maintain it. So they have normal X-inactivation for a very, very short period in time and then it goes terribly wrong because they can't maintain it through every cell division. They can't keep remembering to keep that X chromosome off.

DYANI LEWIS
Marnie, smchd1 is involved in this X-inactivation, does that mean that it's only working on the X chromosome?

MARNIE BLEWITT
So for a long time we didn't know what else it might do actually. But we've recently been able to work out other things it's involved in. And so it's not just involved on the X chromosome in females but actually it also has a role in males. So we know this partly because if you have males that lack smchd1, some of them can survive, but others of them also die around birth. So clearly it's doing something. What we've been able to find now is that it's also involved in silencing at other loci, not on the X chromosome but on autosomes, so just the regular chromosomes that everybody has, has two copies of. These are involved at the imprinted genes and these are genes that are also highly unusual. There are about 100 of them in our genome and we express them only from one copy, a bit like the X chromosome in girls, but only from one copy. The one that we express from is determined by whether or not we inherited it from the mum or the dad.
For example, you always express it from your maternal copy and never from the paternal copy. So in that way it's unlike X-inactivation because it's always one way round, these imprinted genes. So smchd1 is involved in silencing at some of those cases as well.

DYANI LEWIS
You said we've got about 100 genes so it's only a small proportion of the genome that's imprinted like this.

MARNIE BLEWITT
Yes that's true. They keep on defining more of them but it's by no means the whole genome. We think more areas of the genome actually undergo a similar sort of mechanism but at these other regions it doesn't seem to be an all or none sort of a thing, like the imprinted genes. But rather a preference for expression from one allele or one [parental] allele or the other.

DYANI LEWIS
By alleles you're referring to just different gene variants that people might have.

MARNIE BLEWITT
Yeah but in this case just 100 or up to 130 or so genes that are expressed in this strictly imprinted manner.

DYANI LEWIS
It still does very much change the traditional thinking of how genes work in that you've got two copies, one from your mother, one from your father and they contribute equally. It seems quite a different paradigm to that.

MARNIE BLEWITT
It is completely different. It's very difficult actually to get your head around when you first start to think about the basis behind this. And there are some very unusual diseases in people where they're imprinting disorders. That is where they have one copy of these genes messed up but it happens to be the copy that they were supposed to be expressing. So they might have got a normal copy from the other parent but that's irrelevant because that's the copy that's always silent or always turned off epigenetically.

DYANI LEWIS
smchd1 you've discovered is also involved in a genetic condition. What's this condition that you've found a link to?

MARNIE BLEWITT
Yeah so our collaborators actually, they were able to find it. So it's a type of muscular dystrophy, it's a very special type of muscular dystrophy, it's not the ones that we normally know about, like Duchenne muscular dystrophy and others. It's called facioscapulohumeral dystrophy type 2. It's just because it has particular muscle groups that are involved and obviously, based on its name, you can tell that it's the ones that are involved in the upper body and in the face. In this case the patients have no smchd1, they don't have a deletion in both copies like you would in a recessive disease but rather they have a mutation just in one copy of smchd1. We think this is probably why they're, in general, alive and able to make it to adulthood. But this is a progressive sort of muscular dystrophy that gets worse with age and tends to have an onset on the teenage years and then you progressively get worse and worse, unfortunately for the patients.

DYANI LEWIS
I'm Dyani Lewis and my guest today is molecular biologist, Dr Marnie Blewitt. We're talking about epigenetics and X-inactivation, here on Up Close. Marnie, when you're looking at mice, how much can the work that you do in mice around epigenetics be translated to what's happening in humans?

MARNIE BLEWITT
That's a very good question. So it's something that we need to constantly ask ourselves. A mouse is clearly not a human, how much of it is relevant? And interestingly for epigenetics we find actually the vast majority is. So while we can't go and test what we find in a whole animal study in mice, straight away in humans, because it's not ethical and not possible, what we can often do is test whether the same sort of paradigms are true from what we are finding in mouse studies, in human cell lines for example. Or tissues that we can obtain and test whether or not the same sorts of things happen. And so while there are certainly species differences, these species differences are more pronounced in particular fields. So you can imagine if you're studying behaviour there will be large differences there. But when we're talking about really basic biology, like epigenetics, and we're thinking about these primary processes, like X-inactivation, much of it can be almost directly transferred into the human work. So we know smchd1 is also important in human X-inactivation as an example.

DYANI LEWIS
Now there's one gene that you've been looking at it particular. It has effects on regulating so many different genes. There's all of the genes on the X chromosome for example but then all of these other genes on the non sex chromosomes, the autosomes. Given how broad its effects are, will we ever be able to design a drug, for example, that corrects this mis-regulation if it goes wrong?

MARNIE BLEWITT
We hope that we will be able to design a drug that, for example, can boost the expression or boost the function of smchd1, the good copy of smchd1, in those patients that have one copy that's been abrogated. That's certainly what we're trying to work towards, it's actually an international collaboration between lots of labs trying to do this. In the case of smchd1 the reason we're most hopeful about being able to do that is because in those patients we find evidence that all of smchd1's functions are disrupted. So if we had just disrupted smchd1's function at one particular locker, so only one genomic location, but everything else was left unchecked and everything else was fine, then it would be slightly more difficult. Although, having said that, there are many people who are working to do just that, just modify one particular region. But because in these patients we think that smchd1's function is abrogated and it affects all of its genomic loci where it plays a role, then actually we're in a better position to be able to try and activate smchd1's function.

DYANI LEWIS
Are there many genetic conditions like this where epigenetics plays such a crucial role?

MARNIE BLEWITT
I think almost every condition, epigenetics plays some role. But certainly there aren't many where epigenetics is so critical as this one. In fact, not just this one, but there are other disorders called repeat disorders, people might know about Huntington's disease as an example. So what I mean by a repeat disorder is a disease that's caused by having a small tract of DNA, a small number of base pairs, that's repeated maybe 30 or 40 times or a few hundred times in one region. That disrupts the neighbouring gene or the gene within which that is found. What seems to be coming to light now is that those repeats, when they are either expanded or contracted, can either be activated or silenced accordingly.
So the repeat disorders seem to be more about the epigenetic changes that those repeats cause, when they're either expanded or contracted, than they necessarily do about other types of controls. So our understanding of repeat disorders in general, of which this one that I mentioned, facioscapulohumeral dystrophy type 2 and type 2, are part of, we think epigenetics is really critical to understanding the disease pathology in these cases.

DYANI LEWIS
These conditions are quite rare. Aren't there more common examples of when epigenetics goes awry, that they can affect our own health?

MARNIE BLEWITT
Absolutely. Cancer is probably the best studied example. So because there are hundreds of different forms of cancer, what's very interesting is every single cancer tissue or cancer sample that's ever been studied, has global changes, so really widespread changes in epigenetic marks. While we have traditionally thought of cancer as a genetic disease - in other words like an accumulation of mistakes that are made throughout your lifetime and then if you happen to have enough mistakes in the one cell then this cell can then grow in an uncontrolled manner and you can't kill it easily, so it's a cancer cell or it becomes a cancer - we actually now know that there's much more to it than that. In those cells that have lots and lots of genetic mistakes, they also have many, many epigenetic mistakes. This is really interesting because, as I mentioned earlier, epigenetic marks are mitotically heritable, so through cell division they keep on being maintained. So just like a mutation of course keeps being maintained, epigenetic changes can be as well and so they perpetuate through the tumour. So if there's selection for some particular epigenetic mistake because it happens to give that cell a growth advantage, that will be perpetuated.

DYANI LEWIS
So people are looking at drugs that correct these again are they?

MARNIE BLEWITT
Yes absolutely. So there are four that are already approved by the FDA in America and are in use in clinics around the world. But there are hundreds more, literally hundreds more, that are currently in clinical trials with pharmaceutical companies worldwide. So these are trying to target particular classes of those epigenetic modifiers we spoke about earlier. So some of them target the methyltransferases, others of them target different types of these epigenetic modifiers. They're hopefully going to be very powerful because they might be able to help undo those epigenetic mistakes that are in cancer. So if they're used, either by themselves or in combination with standard chemotherapy, the evidence to date suggests that we have a much better chance at fighting cancer and being able to remove the cancer cells as best we can.

DYANI LEWIS
So there are some drugs that can alter epigenetic markers. What about things like lifestyle factors?

MARNIE BLEWITT
Yeah so basically I guess you can consider that almost anything that you can put into your body might have the capacity to change your epigenetic makeup. This is really just a growing field of people who are starting to study it at the moment. The sorts of things they're looking at are diet, so your dietary make-up, but also there's been some recent interesting work done on alcohol. So what happens when you consume alcohol – to your epigenetic make-up? The particular focus has been on consumption of alcohol during pregnancy. So we know too much alcohol during pregnancy can cause foetal alcohol syndrome which is really a horrible disorder which affects a lot of organs but particularly the brain. But what they've also been able to show is that alcohol consumption during pregnancy can alter the epigenetic make-up of the offspring. This is quite interesting because something that we've always avoided, or for a long time the advice has been to avoid alcohol during pregnancy, we now know a little bit more at the molecular level of how it might be influencing those babies.

DYANI LEWIS
Is that period in utero during early embryo development, foetal development, is that the most important part of our life in terms of getting these epigenetic markers right? Or are there are other things after we're born that are equally as important?

MARNIE BLEWITT
I think it's definitely the most critical period, is in utero development. That's because essentially every location throughout the genome, we are laying down new epigenetic marks. So it's when the tissues are being patterned and so you're setting up particular epigenetic states that are required for those different cellular systems. The reason you have different cell types is because of the different epigenetic marks and these are laid down originally when those cell types are first being developed. This laying down period, or when epigenetic marks are being removed, they're the most sensitive to the environment. So that's the most critical period but that doesn't mean that it's not important again later. There are other periods but they'll be different for each tissue. A nice example is for the brain.
So the brain of course does a lot of developing in utero but again it does a lot of developing in infancy and childhood when there's a lot of learning that's happening and that promotes particular brain development. There's been a study, at least in rats, that shows that the development of the brain early after the birth really depends on mothering style and this is epigenetic. So if you have a good mother you end up having a rat that is not very stressed and that's because of epigenetic changes in their brain. If you have a bad mother - and a bad rat mother means they don't lick their pups very often - so if you have a bad rat mother that doesn't lick their pups very often, they end up not having these same epigenetic changes in their brain, so not the appropriate ones. Then they end up having too much stress hormone for the rest of their lives.

DYANI LEWIS
Is there anything that the pups can do to overcome that?

MARNIE BLEWITT
They've tried these epigenetic drugs to see if they can modulate that in any way. The answer is not really, not to the same extent as those early changes. The reason for that is because again those epigenetic marks are being established during that period. So this time when those epigenetic marks are first being laid down is when it's most critical to get it right.

DYANI LEWIS
It's a really intriguing area of research. Marnie, thanks for being our guest today on Up Close.

MARNIE BLEWITT
Thanks Dyani.

DYANI LEWIS
Dr Marnie Blewitt is from the Walter and Eliza Hall Institute of Medical Research where she runs an epigenetics lab.