>>Ladies and gentlemen welcome to the
2013 Royal Society GlaxoSmithKline Prize Lecture.
I’m Jean Thomas, I’m the Biological Secretary and I have the housekeeping
duties of asking you to turn off your mobile phones, please, because the lecture
is being recorded and webcast. And also, to tell you in the that
actually, I hope, unlikely event that there’s a fire, you don’t go out through
the usual doors but you, because of the snow and whatever, you go out through
these doors instead. So, 2013 is the Royal Society Year of
Science and Industry. This is the year when the society will
showcase excellence in UK industrial science and strengthen links between the
society industry and academia. The Royal Society recognizes that world
class research and development in the UK industry is essential for transforming
innovative ideas into commercially successful products into its economic
growth and securing the science space. And it will be proactive in anticipating,
understanding, and responding to the needs of industry’s scientists.
Symposia and meetings with high industry interests have been added already to the
society’s calender which already includes longstanding initiatives in scientific
excellence, such as the Royal Society Industry Fellowship and the Brian Mercer
Awards for Innovation and Feasibility. So, the year of science, science and
industry will bring a renewed focus on engaging with the industrial sector to
develop cogent arguments that high level investment in the UK science space is
essential for international competitiveness.
Something we would all, I’m sure, sign up to.
Now, to the prize and the lecturer, the Royal Society GlaxoSmithKline Prize and
Lecture is awarded biannually for original contributions to medical and veterinary
sciences published within 10 years of the date of the award.
The prize consists of a very nice gold medal, an even nicer check for 2,500
pounds, and the recipient is called upon to deliver an evening lecture at the Royal
Society which is why we’re all here this evening.
And this is really a, a capacity audience, and the reason we’re a few minutes late
starting is that there is an overflow room, and I don’t remember that in the
last, certainly, in the last four years of chairing these evening lectures.
So, Adrian has really put in a big crowd tonight.
So, no pressure there, Adrian, at all. It was, the award was initially
established following a donation from the Wellcome Foundation.
First award was made in 1980 the centenary of the work and foundation and since 2002,
it is being supported by GlaxoSmithKline Limited.
So, this year’s recipient of the prize is Adrian Bird an old friend and colleague,
I’m delighted that he’s received this award.
Adrian has held the Buchanan Chair of Genetics at the University of Edinburgh
since 1990. And he’s a member of the Wellcome Trust
Center for Cell Biology in Edinburgh. His research focuses on the basic biology
and biomedical significance of DNA methylation and other epigenetic
processes. His laboratory identified CpG islands as
gene markers in the vertebrate genome. And he discovered proteins that read the
DNA methylation signal to influence chromatin structure.
Mutations in one of these proteins, MECP2, and I’m sure we’ll hear a lot more about
this, this evening, causes the severe neurological disorder Rett Syndrome, which
is the commonest genetic cause of mental retardation in females.
Adrian was made a Fellow of the Royal Society back in 1989.
He’s received several awards, numerous awards for his work, including notably the
Louis-Jeantet Prize for Medicine, the Charles-Leopold Mayer Prize of the French
Academy, and the Gatineau Prize. This evening, it’s the turn of GSK and the
Royal Society to give him this special GS Royal Society, GSK prize.
And has to give his lecture in order to earn that.
His lecture is entitled, as you can see, Genetics, Epigenetics, and Disease.
So, Adrian, over to you.>>Thank you very much, Jean.
Thank you very much for this award. It’s a great honor to, to be asked to give
this lecture. And thank you very much for braving the
elements to come and listen. I think, probably the title of Genetics,
Epigenetics, and Disease is broad enough that it sounds like it’s going to change
all our lives in this next 45 minutes. But in fact, I’m going to focus on a
relatively small part of it ultimately. But I’m going to start off reasonably
broad. There’s one deliberate mistake on the, the
first slide. I hope it’s the last one.
It’s the year. So let’s go back in time to the draft
sequence of the human genome because this was a, heralded as a, a time when biology
really became a, a hard science. If you like, it was seen as the, the
beginning of the end. We now knew the entire code for all, we
knew the sequence of all the genes required to make a human being.
But it’s pretty clear that it was actually the end of the beginning.
And the somewhat apocalyptic predictions that now one simply had to automate, the
discovery of all the medical innovations that would result from the genome sequence
was premature. In fact, it’s likely in my opinion that
there’s still another century of biology to be done and this will be an exciting
century of discovery converting the promise of the genome into the reality of
biomedical applications. And that, one of the issues I think that,
that, that we would really love to be able to solve, a big, a big question if you
like, is where DNA, despite being the thread of life, you can put it in a tube
and gaze, gaze at it for as long as you want and it remains utterly dead.
So the question is really what does it take to make it alive?
When Craig Venter synthesized a bacterial genome an important synthetic biology
milestone, it had to be put into a living cell before it became alive.
How can one bypass that? As the chemists say, you only really
understand something if you can make it. We can’t actually make life but it would
be good to know some of the rules required to do that.
So, some key unanswered questions about the genome that, that remain and this is
only a selection. First of all a basic fact, genes make
proteins, here is the chromosome, here is the sequence of the genes, there is the
RNA. It encodes the sequence of the amino acids
that lead to the protein that folds up to then do all the lifelike things that are
required. But how are only the right genes expressed
in a cell type? This has been a question, a long standing
question. Do we know the answer to it?
Why globin is expressed in blood cells and keratin is expressed in skin cells,
etcetera. We, we approximate knowledge about it, but
actually, there’s an enormous amount to find out.
Most of the genome is actually inaccessible.
This is this gray, it’s rather difficult to look at this picture I think because
the DNA is gray and looks although it should be in the background but this is a
nucleusome, the repeating unit of the, of the chromosome, if you like.
The fundamental repeating unit. And the DNA clings to the outside of it.
And proteins that want to make genes active, can’t actually get at the DNA
properly. So, how does the gene activation machinery
gain and how does it keep access? Again, we have some beginning answers to
this, but we don’t, by any means, have a full picture.
Protein-coding DNA sequences are only 1% of our genome.
So, if you look at a piece of the human genome, you see these vertical stripes
correspond to the bits of this gene that are separated from each other.
In fact genes are fragmented and they are a tiny minority of all the DNA.
What is the rest of it for? There is an enormous, there’s a vast
majority that is, that we can’t explain. This isn’t the case with all organisms.
This, for example, is yeast, and you can see now the genes are packed together.
It’s difficult, it used to be said casually that the rest of this DNA was
just junk. But now, it’s sort of almost politically
incorrect to call it junk. It’s particularly after the encode project
which found lots of potential regulatory sequences throughout here.
So, this other DNA is doing stuff. And perhaps, it’s doing stuff that makes
for example, humans and other mammals far more complex than yeast.
So finally, there are questions almost sociological questions.
Does the environment have any impact on gene expression?
And this is a, a question I’ll allude to in a moment.
But it’s not one that is the main subject to this, this talk.
So, I put in the title Epigenetics because I’m quite mine, our work is quite often
described as epigenetics. It literally means above or in addition to
genetics. But the definition has been controversial
and I’m just going to skim somewhat lightheartedly over some of this because
it’s, it’s at meetings to do with Epigenetics.
One can see various opinions expressed with varying degree, this one I believe
was in Barcelona with great vehemence. So, let me just try to sort of consolidate
this. The original epigenetics definition comes
from Conrad Waddington, who was actually my predecessor as Buchanan Chair, Chair,
Chair of Genetics in Edinburgh. And what he meant was in contrast to
pre-formationism, but the development proceeded by the gradual unfolding of the
information in the genes, to produce the whole organism.
So, for him, how information of the genes is read during embryo, during embryonic
development to give the whole organism was the essence of what epigenetics was about.
We would now call this developmental biology.
How the genotype gives rise to the phenotype.
But it’s acquired, or a sort of, a special status in epigenetics, really, because of
this iconic picture, the epigenetic landscape.
I’m not going to dwell on this either. Because quite honestly, having had it
explained to me several times, I’m never totally sure, exactly how this helps.
It’s a picture of a bull rolling down a hill.
The number of options for the bull get progressively less.
But I don’t feel that this encapsulates anything very useful.
This, however, is a fundamentally important question that remains on our
agenda. Second definition of epigenetics which is
rather different has actually different origins epistemological origins.
How characteristics are inherited across cells or organism generations without
changes in the DNA, its sequence, itself. An example of this is this cat, the
so-called tortoise shell cat, or calico cat, in, in, in the US, which has these
patches of fur. It has two x chromosomes.
One of them has a gene that gives black fur, the other one has a gene that gives
orange fur, and cells early in development, inactivate one or the other
of those chromosomes for, for reasons we don’t, which I will, I will come back to
actually, a little bit later. And you get a patch of skin because the
cell that originally inactivated the orange fur gene gave rise when it divided
to cells that did exactly the same thing. So, that was inherited.
All the gene or the, the DNA is still there in these cells, in, in the orange
ones, and the black ones, but there is difference that is inherited and that’s
epigenetic according to this definition. So, heritable traits of this kind might be
influenced by the environment. And this is sort of revitalized that an
ancient argument about nature versus nurture, where nature is genetics, the
idea that we’re, our genes are, are in control and nurture is the opposite, the
idea that our environment determines who we are.
Of course, it’s a mixture of both but epigenetics has given a, a, a new lease of
life to the nurture argument. And so, one can see articles such as this
and there are many examples I could have chosen why your DNA isn’t your destiny,
the new science of epigenetics reveals how choices you make can change your genes and
those of your kids. Now, I’m not an expert on some of the
epidemiology behind this, but the, the molecular biology, in my opinion, is far
less convincing than it is for other aspects of epigenetics.
It is, however, an extremely interesting idea, that the environment can give rise
to changes that get passed on, but it is systematically overstated in a lot of
places one finds it described. So, one has to be circumspect about the,
this kind of argument in my opinion. There are couple of excellent examples in
plants, in worms where immune, immunity is involved, but some of the more
sociological aspects, in my opinion, require further evidence.
So, I’m sticking with this as my example of heritable epigenetics.
It’s closer to the molecular biology we actually understand.
So, Epigenetics 3, biological significance of the epigenome.
Another definition, it’s risen pragmatically.
What is the epigenome? Well, here is a genome of a, of a cell.
It’s, it’s the chromosomes that were obviously designed for an experiment
because there are fluorescent pinpoints here.
Ignore those, that’s a human chromosome compliment.
If you explode those chromosomes, you see beads on a string and this is that
repeating unit I referred to earlier, the nucleosome with the DNA going round the
outside. It looks like beads on a string.
So, the epigeno, epigenome refers to markings of those beads, of that string of
beads in such a way that the region, it is regionally, regionally adapted to its
function. So, for example, there can be a region
where gene is stably ON, and there is a whole plethora of marks that appear that
reinforce that decision. And similarly stably here, a gene OFF,
such as the black-coat gene in our orange patch of fur.
And again, you get adaptation, and this is the epigenome, and the study of what the
epigenome means, is another definition of epigenetics.
So, you have DNA methylation here where these methyl groups are added to the DNA.
You can’t do much to DNA without changing its propterties, its important properties.
Almost, the only thing it seems you can do is put these methyl groups on and even
that is bad in a way. I don’t have time to go into, it causes an
increase in the frequency of mutations. But the, by far, the most elaborate way of
marking the chromosomes, is via these beads which, invisibly on any of the
structures I’ve shown you before previously, have tails.
And these tails are basically ticketing entities that you can add chemical
information to. That the cells can write information in
the form of chemical alterations. And so, you add this and, that says,
stably ON or stably OFF. Again, we have the broad outlines.
We can correlate quite a lot of these with activity and silence.
But if you were to ask exactly what each of these modifications does, we have,
still have a lot to learn. So, if you like, it’s the, epigenetics is
the structural adaptation of chromosomal regions so as to register signal, or
perpetual, perpetuate altered activity states.
And importantly, proteins that read these marks, write the marks, or erase the
marks, remove the marks are implicated in human disease and quite a lot of
excitement in pharma including GSK is devoted to finding out what these drugs
might be good for in terms of human disease.
So, epigenetics then embraces key unsolved problems in Biology, how, how the genotype
give rise to phenotype, that’s the Waddington one, how traits are inherited
across cell or organism generations without changes in the DNA sequence and
how structural adaptation of the genome facilitates gene activity programs.
As far as I’m concerned, this is not a word one needs to dwell on with sort of
almost a theological interest about what it means.
Everything it possibly means is interesting.
So, let’s get on with studying it. And I, I like to think of it as how the
genome is organized and managed to make DNA if you like, come alive.
So CG is one such signal it’s one of those marks and you’ll notice CG is not actually
a, a mark, it’s actually a sequence, it’s a 2 based pair sequence.
Dna sequences that recognize proteins are usually longer than that because they’re
rarer. If you have a sequence of one base, every
few bases you come across it, and it doesn’t have much information.
Two bases is not much better, but nevertheless, as you will see, CG is used
as a genetic signal and also as an epigenetic signal.
So, here’s a piece of DNA, flattened out so it’s no longer helical.
Those two strands are anti-parallel and CG is paired with itself.
So, CG pairs with CG. This TA is paired with itself, but it’s
nearly so interesting. And one of the things we’ll talk about
that can happen to CG is that the C can get, gets methylated.
And that, since there are two of them, that can be a symmetrical event.
And it looks like this, they sit in the major grooves.
I’ve already shown you a different picture, though with less vulgar coloring
that shows the two methyl groups sitting in the major groove and they influence
interactions between proteins and DNA. So, what are the features that adapt CG
for a genome signalling function? The first is that, as I’ve mentioned, you
can get it in, in three different chemical forms, actually there are more than three,
there are another two but that, it’s not yet clear whether these are biologically
important or just by-products, at least it’s not clear to me.
You have CG unadorned, you have CG methylated, and you have CG where the
methyl group has had an oxygen added to it, and it becomes hydroxy methylated.
So, it exists in different forms. Specific proteins are attracted or
repelled by different modified forms and we’re going to talk more about that.
Highly variable in frequency, so then, the frequency of CG despite the fact that it’s
just a two base per sequence is dramatically different going along the
genome. In the bulk of the genome, 99%, it’s quite
far apart. These lollipops represent CGs.
The lollipops that are solid represent methylated ones and the open ones
represent unmethylated ones. So, 99% of the genome has not many CGs and
most of them are methylated. But then, there are these clusters where
the density is about 10 times higher and these are the so-called CG islands.
They are interesting because they sit right on top of the control regions for
genes. So, here’s a gene, it’s red in this
direction and then these blue bits are spliced together to make the messenger
RNA. And sitting right on top of the promoter
is this CG island, and this amounts to about 1% of the genome.
There’s another one there. And here’s a biological consequence of the
methylation. If you look at this CG island, it can,
under certain circumstances, this happens on the inactive x, this happens at
imprinted genes. It happens at germline genes in the soma,
it happens in cancer, abberantly. It gets methylated.
And when that happens, you shut down transcription of the gene.
And because methylation is something I haven’t gone into, is relatively stable,
it can be transmitted from one generation to another, if you like, copied.
When cells divide one cell generation to another, it’s this is quite a
stable[UNKNOWN]. So, one of the things DNA methylation
does, is it shuts down the expression of genes.
So, we’re gonna talk about specific proteins that are attracted or repelled by
modified forms of, of CG. And I’m gonna start just with a protein
that recognizes unmodified CG so it cant recognize this or this.
So, Cfp1, sorry about the acronyms, it’s a, it’s a protein that recognized, it was
discovered, in fact, by David Skalnik it binds to non-methylated CG.
I don’t know why I’ve drawn the DNA at this jaunty angle, but it, it just meant
to show that it’s interacting with it. And it also interacts with a complex of
proteins. An enormous complex, well, relatively big
complex, called set 1. And this complex does something to the
nucleusome. We’ve seen this before, this is the bead
on the string, the DNA going round the outside.
Haven’t, in, in, when you look, determine the structure of something like this, you
don’t find the tails, the things that you write on.
And so, I’ve drawn them freehand, nobody actually knows where they are because
they’re so floppy, they don’t come up in the x-ray structure.
But, amino acid lysine number 4 gets methylated and this is done by this
complex. So, we have a protein that binds to
non-methylated CG that recruits a complex that methylates this.
Now, why is that interesting? This is a mark of active genes, so if we
look where CG island are, CpG island as they’re more often called, in fact, here
are the CpG islands, I’m not going to tell you how we know they’re there.
But you’ll notice these, this gene is going this way, there’s a CpG island at
the start of it. This gene is going this way, actually
bidirectionally, there’s one going this way, one going this way, there’s the CpG
island. So, they’re all the CpG islands, there’s
the RNA polymerase, the protein, the machine that makes that starts to be
converted, copied into to messenger RNA. And it’s just at the beginning of them
because this is the, a particular form of RNA polymerase that is only at the
beginning of genes. And here is this mark, H3K4me3, which
means this purple blob on this tail, which is put on here.
So, we have the non-methylated CG cluster here and we have the mark, and this mark
is involved in gene expression. So could it be that the, the proteins
attracted by the CG brings in this and that’s what causes this mark?
If you look as, as we did where the protein is, it coincides with the CpG
islands. So, it’s in the right place.
If you take it away, the k4 trimethylation, this, these peaks here go
down and that’s consistent with the idea that this is reading the CpG island signal
but the key experiment is really the Pete Skene and John Thomson did is to insert a
piece of CpG-rich, CG-rich junk into the genome, real junk in fact, it’s not
actually quite junk, it’s the jelly fish gene that’s been optimized for expression
in humans lacking any control sequences, just inserted into the genome, so you make
a CG island like sequence with a cluster of CGs.
Now, are you creating a new H3K4 trimethylation peak?
So, here’s a, here’s a map of all the CGs. The vertical lines show where they are and
this is what we’ve inserted. And you can see the density of CGs has
gone up. And now, you can plot that density.
Now, where’s Cfp1, the protein that binds CG.
There it is. We’ve now got a new peak of it.
And what about H3K4 trimethylation? It’s there, too.
And you notice, where there’s most CG, there’s most, more, most of that
modification. Is there, have we just made a gene?
In other words, all the stuff that does geney things is there.
No. Because there’s no RNA polymerase there.
So, this is just the DNA sequence, talking to the chromatin.
And, and as one can do this with other sequences and verify that it’s the case.
So, a CG-rich piece of DNA creates a new region of H3K4 trimethylation, this active
promoter mark, even when there is no active promoter there.
So, the presence influences of the CGs influences the chromatin structure via
this link between the DNA binding proteins and the set enzyme complex and other CG
binding proteins also recruit, rcruit modifying enzymes.
In fact, for a long time, we were used to the fact that CG islands existed, but we
didn’t really know what they were for. And, and actually, one almost forgot to
ask, well, they’re always there, what are they for?
In fact, it now seems very likely that they are platforms to set up appropriate
genome structures at gene promoters. Very important function.
And there are other proteins that bind CG, that recruit other things to them, and
this is a very, a rapidly growing area. So, suddenly, we find that the CG island
is a, is a, is a, a structure of biological importance, and we’re starting
to disentangle how. So here is a CXXC protein which has its
domain. Cxxc is the name of the protein domain
that binds CG. It comes in wearing this ludicrous wig.
And creates a sunny promoter gene, gene activity friendly region of the genome.
If there was a methyl CG,it comes in and it is goes away.
It can’t bind. So now, I’m going to turn to for the rest
of my talk proteins that bind to the methyl CG mark, the one with the purple
blobs on. The purple blobs that were on the DNA, not
the purple blobs that was on, was on the histone tail.
So, what binds this form of CG? Well, a protein that we found a long time
ago is MECP2, and this binds, specifically, and I’m going to show you
some of the prehistory of this protein. First of all, a picture from the paper and
you could either take from this how prescient he was, to be able in 1992, to
find this protein that turns out to be so interesting.
Or you can think, he’s been working on that protein for 21 years and he’s still
not quite sure where it does. Let’s you can shoot, take your pic at the
end of, of the talk. So this shows how we first found it.
We run the proteins that are in a nucleus on a gel.
And then, we probe them with a piece of DNA that’s labeled and methylated.
And then, the same sequence of DNA with no methylation.
And clearly, there’s a protein that bind one that’s methylated and at about 84
kilodalton and doesn’t bind when it’s not. And we now know a structure for this in,
in an atomic detail. Here are the two methyl groups sitting in
the major group. And this is the domain of this protein
that interacts with them. So, we were happily studying this for
blue, blue skies reasons to, to try to find out a protein that read DNA
methylation and therefore a reader of DNA methylation, and find out what it did.
When Huda Zoghbi showed that the gene that causes Rett Syndrome, an autism spectrum
disorder, is almost exclusively, more than 90% MECP2.
So, this is the gene that is mutated in Rett Syndrome.
So, what is Rett Syndrome? This is a, a, a film just taken from
YouTube, not somebody I have ever met but you can see the characteristic features of
Rett Syndrome which involve this repeated hand clasping and a period of apparently
normal development saw 6-18 months, and then regression, progressive
encephalopathy, repetitive hand movements, breathing a, a, arrythmia, a, a profound
problem. But nevertheless, a life expectancy of
about 40 years on average. So, there is no effective treatment and
24-hour nursing is required. So, this was all caused not by a brain
gene, that was what was being looked for by everybody.
And those were in the days where you thought the gene would have something to
do with the, the, the tissue that was affected, but a basic housekeeping protein
that reads DNA methylation that’s expressed in every cell type.
So, why does this disorder only affect girls?
Well, you probably guessed it’s because it’s on the x chromosome.
Males are always more affected by mutations in genes on the x chromosome
than females because they only have one x and females have another one which can
compensate and males die. There is no male Rett Syndrome simply
because males don’t survive. So then you have a new mutation.
Nearly always as it happens like many of these things, paternally derived.
And then proceeds x chromosome inactivation.
I remember so, in order that females have only the same number of functional x
chromosomes as males, they shut one off. And this happens in random cells and I’ve
shown you the example of the cat and I’ll show you the cat again.
X, this gene, this cell inactive[UNKNOWN], this progeny of this cell inactivates this
x chromosome this one inactivates the other one and this then is inherited so
this is the epigenetic inheritance phenomenon it’s passed on.
And the end result is, there’s the wretched cat again but with his different
things, I will show you a different example of that in a moment.
But the, this, well, the point to be made here is that the brain and, in fact, the
other tissues of a Rett patient consist of a mosaic of a salt and pepper mixture of
cells that are functionally normal with respect to MECP2 and cells that are
functionally without normal, I mean, without MECP2.
So here, for example, if the phenomenon a new mutation arises, sometimes, that
mutation is the only MECP2 in the cell, and the other time, it’s invisible.
And you just get the wildtype express. So, this is the mosaicism.
Now, the equivalent of the cat picture in the brain, though, is rather different.
This shows the dentate gyrus of, which is a region of the hippocampus, which is part
of the brain in, in from a mouse, I haven’t talked about the mouse yet in any
detail but just to show you that you can see patches.
It’s probably better to see it here in the merge, this is MECP2 and it’s in blobs and
there are gaps. But actually, there aren’t gaps in the
nuclei staining and so there are patches of cells here that are inactivated the
functional MECP2 gene, and there are other patches here that function that
inactivated the non-functional MECP2 gene. And you see these patches, the point I’m
making here is the patches in a cat are gigantic and involves millions and
millions and millions of cells. The patches in the brain are, for reasons
we don’t quite understand, tiny and so you get a bigger mixture of functional and
nonfunctional cells in this tissue. So, the first thing we did when we found
this out was to make a mouse. We were gonna make a mouse anyway for our
blue skies reasons but now we were energized, I would say, and that
energizing has continued to the present day by the fact that we were working on a
human disorder and we’re actually in touch with a community of people who are
affected by it. So if you take a normal mouse, it lives in
this green state for a long time. But the MECP2 minus mouse, the male, the,
the equivalent of the human male that doesn’t survive, doesn’t survive.
And these colors are meant to indicate that they get symptoms, get worse and
worse, and eventually die. The female and this, this shows a, a sign
of neurological symptoms in a mouse. It does this hind limb clasping and it
does that at this blue stage here. Initially, there is no observable
phenotype. But later on, they become ill and
subsequently die. The females and these are really the true
model of Rett Syndrome because they’re heterozygous as, as geneticists say for
these mismutations. They’re fine, and that’s how you keep the
line going. They breed for several months and a mouse
at six months of age is quite an old mouse.
It’s had quite a few liters. But then, they suddenly hit a, a, a wall
and they become immobile and they develop all the other sorts of symptoms including
hind limb clasping, arrhythmic breathing lack of mobility that, that characterize
the Rett-like phenotype. And there’s a dramatic change in their
behavior but it’s stable, just as it is with humans.
So if you like though, the MECP2 deficient mouse is actually quite a good model.
Not all, not all models are, are particularly persuasive.
But it’s quite easy to persuade skeptics that this is a good model of this disorder
because a lot of the things that MECP2 seems to do in humans it also does in
mice. So, we’ve got, we’re armed with this
model. Now, how are we going to find out what
MECP2 actually does, and how that’s connected to the function of the brain?
Because that’s what’s gone wrong in Rett syndrome.
Well, the big resource you always have is in, in genetic disorders, is the mutations
that give rise to the disorder. Particularly, if, like Rett Syndrome,
they’re all new mutations. This does not run in families, the males
don’t survive and the females don’t reproduce either.
So, it doesn’t run in families everything is a new mutation.
And so, this is the sort of picture you get.
Everywhere, absolutely all over the place. But I will point out to you that these
frame shifts, these grey ones, the longest bars, everything downstream of that is
disrupted. Because the, the protein goes out of frame
when you start making junk afterwards. So, they don’t mark the spot where there’s
an important bit of this protein. They only tell you that this the boundary
between and everything downstream gets lost.
The other ones, the nonsense mutations, also stop the protein.
That’s, that’s why you put x here. They just terminate the protein.
The ones that are most informative are the missed sense mutations.
Because what’s happened there is, you’ve put an alien amino acid.
One single subunit of the protein in the wrong is wrong.
Everything before it, is fine. Everything after it is, fine.
Just that one amino acid is wrong. And so this is telling you the really
important bits. And if you’ll notice, the blue ones, which
are the missense are not randomly distributed.
So, we went into the database. Now, of course, for a lot of disorders
that look like they might be related to MECP2, and there’s more than Rett
Syndrome. I, I don’t have time to go into that.
People look at the database. And they start, sorry, they, they
sequence. And so, there’s an awful lot of
polymorphisms, a lot of, lot of changes that are not associated with disease.
The one way of being sure it’s associated with disease, is to look for mutations
that are not found in the parents. They’re only found in the offspring.
Cuz then, the probability that, that is, is a, is a function-less genetic variant
is, is vanishingly low. When you see very specific domains here,
interestingly and, I don’t have time to go into this, there are now more and more
xsomes sequences. People are sequencing genes of normal
individuals or for people who have other things.
And so, you can find all the missense mutations where there’s no obvious effect.
And what you notice is that this cluster here, for example, doesn’t have any genes
with no obviously effect. This cluster here the same.
So, you can use the, the normal polymorphisms as a way of seeing the
inverse of what you see with the mutations.
So, now we have two domains. What’s this domain?
Well, I’ve labeled it MBD. Actually, what that stands for is
methylated DNA binding domain. I showed you the x-ray structure of that
bound to methylated DNA. That’s the bit that contacts DNA and
brings this in, and many of these mutations prevent that.
So, we’re pretty clear what’s going on, I’m just going to tell you a couple of
things about that domain. The first thing that emerged when we
studied it was people, you tend to think when you find a DNA binding protein that
it goes to specific targets and then, it does stuff there and those targets are its
main function. But actually, it turns out, it turns out
that MECP2 is incredibly abundant in, specifically in neurons.
And in fact, there are 17 million molecules per nucleus in a cell.
And this is a lot it’s one every four hundred base pairs, it means there’s
enough to coat the genome and then that’s actually what it does.
Dna methylation goes up and down along a chromosome so this a very low resolution
picture and the MECP2 goes up and down in exactly the same way.
So, it’s, it’s not in special places, it’s all over the place, with somewhat
different densities and it’s very, very abundant.
It doesn’t behave like a transcription factor which goes to specific target
genes, it binds globally. So, that’s that domain, let’s now talk
about this domain and this is more interesting to us.
What’s more interesting to us, because we had no idea what it might be.
So, hypothesis was that this region binds to DNA.
And then, this region binds to some sort of partner that it brings in, and that’s
its job. And you can’t mutate that because it fails
to do that. And Matt Lyst really led this aspect of
the project. What he, we did, was we made a mouse with
a green fluorescent protein tag on the MECP2 and then we pull down that tag from
the brain, an extract of the brain of the mice that had it.
And then, we ask what came down with it, those of the partners?
And by mass spectrometry, we found these, these proteins.
This is the list of the top 8. Interestingly well, MECP2 came down.
That’s a relief. You expect, if it didn’t, you’d have a
real problem. Then two proteins that transport it into
the nucleus. But then these 5 subunits and more
acronyms, I’m afraid, of a complex that’s well-known.
This is a huge complex, more than a million daltons complex, which contains
which contains a histone deacetylase 3. So, what a histone deacetylases do is,
they remove a mark on one of the tails, that mark is associated with activity.
If you remove that mark, you work against gene activity.
In fact, you silence gene expression. So, this is a complex that reinforces the
silence of gene expression. Shuts the genes down by removing this
methyl group. So, here it is, there’s the methyl group,
PowerPoint extravaganza goes. So this, it’s well-known to buying nuclear
receptors. And it also, now we find that it binds to
MECP2. Now, where does it bind to MECP2?
Well, it binds it, you won’t be surprised to hear, exactly in this second domain.
And all of those mutations that cause Rett Syndrome in this second domain, abolish
the interaction with this, this complex. So, this mutant protein can’t bind DNA.
This mutant protein can’t bind NCoR SMRT. This what, which is the unfortunate name
for this complex. And also, you lose the ability to shut
down transcription. So we, we now have this fairly persuasive
model I think that MECP2 is a bridge. It’s a bridge between DNA, there’s a
methyl group MECP2 is attached to it. It’s brought in this complex which is a
gene silencing complex and if you have mutations in the DNA binding domain and it
can’t bring it in and if you have mutations in the complex interaction, you
can’t quite bring it in either. So, MECP2 then and other proteins that
bind metal CG, and there are others about which we know something.
They come in, bind. And instead of creating the sunny
atmosphere, they create a foggy trans, transcription-hostile environment.
I was going to say like, Edinburgh in January but it’s not really
transcription-hostile in Edinburgh. We express our genes perfectly well.
So then, the big question is, have we got any further now?
We know that, it’s likely to be a repressor.
And to you, it see, it probably seems likely that was always going to be the
case. Dna methylation represses transcription,
this binds DNA methylation, what more natural than it represses.
Actually, it’s a very controversial area, as to whether or not it does repress
transcription. And my feel is, is, is that the important
advance. So the question is what transcription does
it repress because it’s not obvious. When you look in the brains of the mice,
histone acetylation is up, histone H1 is up.
The epigenome is disorganized, expression of some genes is up, other genes is down,
other genes are down. Or and some are unchanged and these
effects aren’t very big. So, very, very briefly, I’m going to say
you could be controlling the activity of specific genes, you could be controlling
transcription in response to neuronal activity only when neurons fire, something
happens, this protein actually gets phosphate groups added to it, maybe that’s
something to do with it or dampening of transcriptional noise.
And this is a, a boring sounding possibility, it just kind of sits on the
genome and keeps everything down. But there’s some evidence for that, we
know that the transposons, which are selfish elements in the genome that like
to jump around when there is, normally, they jump around, represented by these
yellow dots, a very small amount. When you don’t have MECP2, they jump
around an awful lot more and so, in other words, MECP2 is preventing the expression
of the RNA that allows these things to move, and this is no function for the
organism. It’s actually something it would prefer to
keep quiet. So, that’s noise dampening, so this
question is unresolved. So now, in the last part of my talk, I, I,
I left you there with, that’s as far as we got with the Molecular Biology I’m afraid.
But I think we’re now making progress. Now, we know we have this bridge model.
Let’s now talk about the pathology and the trying to get all the way from the
Molecular Biology up to the patients, and our, surrogate for the patients which is
the mouse model. So what we really want to do is have a
molecular description of the legion in MECP2 and the count for all steps to, the
brain of the patient, so that we can understand.
And this requires we know an awful lot more than we do now, for example, how
brains work. So we’re trying to bridge this gap.
Now our involvement in this is really to do with one specific question and that is
this one. Can the symptoms be reversed?
In the pathology, as observed in post mortem brains and as seen in the mouse, is
that neurons are slightly simpler. So, if you put a, a bulls-eye over the
center of a neuron then its arms are more complex and branchy in a normal animal
compared to what they are in an animal that doesn’t have MECP2.
That’s about it for pathology. There’s no cell death so it’s not a
neurodegenerative disorder. It’s not like Parkinson’s or Alzheimer’s
or Huntington’s, where nerve cells die. It’s just a kind of shrinkage they become
underpowered neurons. So, the question then arises, if their not
dead, if we put MECP2 back, can it be reversed?
And I’m gonna tell you about that and then our attempts to do some therapy based on
that. So how do you, how do you have the, how do
you do this experiment? Well, what you do, is you take the MECP2
gene, you put a stop in it, which is just a chunk of DNA that is poisonous for
transcription and then you flank it with sequences that mean, that when you want to
you, and so that then causes transcription to stop.
You’re let the animal grow up, it has no MECP2 and it becomes ill as a result and
then, at your chosen moment, you remove that stop and start transcription again
and you can do that in ways that I, that have been published and I can tell you
about if, if you want to know afterwards but this works and that’s the first
surprise that actually works. And the reason why it works is because
Jacky Guy, who’s an unbelievably talented person in the lab is took charge of these
experiments. So, I’ve shown you the mouse that’s
wildtype, the mouse that’s male. What we’re going to do now, is look at a
mouse where it’s male, it’s on the, it’s in the death zone, if you like.
It’s and we interject it with tamoxifen, which is the way we trigger the deletion
of the stop cassette. Does it work?
Well, this is the, this is MECP2 in a normal mouse brain.
This is in a stop mouse brain. So, the stop works.
This, this looks like it’s cell that haven’t been stopped.
But actually, it’s blood cells that autofluoresce .
And then, you treat with tamoxifen and back comes the MECP2.
So, that works. And then, he is a mouse on, on the day we
started the experiment. So, it is grown up with no MECP2.
It have the classic symptoms of the MECP2 null mouse it has this tremor, it has
arrhythmic breathing which you may be able to see in the flanks.
They breathe and then it stops. They breathes and stops.
And it doesn’t move. And the film is much longer than this and
it still doesn’t move. And then when you humanely suspend it by
the tail, it, it does this hind limb clasping.
So, then, then the question is what does tamoxifen do for that?
And then, this is the same mouse a month later.
Under our animal license, this mouse would not be able to survive for more than a
week or two at the most. And here, it is a month later, remarkably
healthy. And it went on to live I wouldn’t say, a
natural life, but you know, a quite a long life.
So this is an unexpected finding, we didn’t expect it and it turns out nobody
else did either. And, and for that reason it was, it, it,
it’s turned out to be quite important in the field there.
This somewhat unedifying image of a mouse I will leave you and go, go on to the, the
females. Because, you know, those mice are young.
They’re only 6 or 8 weeks old, and so, it could be that they’re young and plastic,
and, and reversal therefore, works better at that age.
Also, they’re not real model of Rett Syndrome.
This is the real model of Rett Syndrome, and these animals are no longer young and
plastic. So, we want to do this experiment.
Inject with tamoxifen when the animals are 6 months old or so.
And then this just shows that this also works.
That’s a reversed animal, that’s a, a wild type animal, a normal animal and you can
see they’re indistinguishable. And this is an animal that was unable to
respond to tamoxifen for the, for the, because we genetically made it that way.
So, it’s still is obese, which is a characteristic of the females on this
genetic background. Immobile, and it does the hind limb
clasping and breathing arrhythmia and all that sort of stuff.
So, you can this, this mouse looked like that mouse when the treatment started.
So, the implications of this reversibility is that, obviously, Rett Syndrome is
potentially a curable condition. You have to use the potentially word there
because these are mice, not humans. But nevertheless, it’s encouraging.
It also means, Rett Syndrome, like most of the autism diseases, have been called
neurodevelopmental disorders. And the implication is that something goes
wrong during development in terms and that you can never recover from that.
And, and I, I think that when one thinks of brain diseases, brain disorders one
tends to think of them as irrevocable. And in, in actually, there’s no, the
experimental evidence to support that is, is not strong.
And this questions that, and there’s work with fragile X Syndrome as well, and other
so-called neurodevelopment diseases disorders that suggest that actually
they’re not neurodevelopmental at all. And, in fact, if you take away MECP2 in
adults, adults die. Certainly, not only required during
development, and so this reversibility may be more widespread and true than was
previously thought, and that can only be good in terms of exploring therapeutic
options. Everything we’ve done suggests that
actually what MECP2 does is sustain neuronal function.
These are cells that are never going to divide.
They take ages deciding who they’re going to be connected to.
And in an elaborate dance of synaptogenesis and culling of excess
neurons and then they never get to refresh themselves and so maintenance is probably
a vitally important function and I think MECP2 may be one of the proteins that does
that. So, that’s just a hypothesis at this
point. So, prospects for therapy, you could do
all sorts of things. And for time reasons, I’m not going to go
through this. I’m just going to talk about our attempts
to do gene therapy. The dose of this protein is critical.
So, gene therapy doesn’t sound very promising.
But I’m just going to show you some results.
Because I think gene therapy or more likely, gene editing, is the logical end
point of the genomics revolution. Having found all these genetic variability
associated with disease what better than to be able to fix it.
And I feel that where, this is what’s going to take awhile of basically
engineering to find out exactly how we should do that.
So, this by comparison with that aspiration, is very primitive.
This is a Adeno-associated virus. And the experiments here were done in
collaboration with Gail Mandel of Howard Hughes Medical Institute in Oregon and her
laboratory and[UNKNOWN] Helene Cheval did the experiments as well.
So we take this, this MECP2 promoter. We drive a, a truncated MECP2 gene, not
much fits into these viruses, and they don’t replicate.
You then, in put them in and there are 2 ways.
And you can go directly into the brain through 6 bore holes in the brain which is
very laborious. And this doesn’t actually work very well.
But what Gail Mandel’s lab did was to use this virus in an unexpected way, namely,
to put it in the systemic circulation system.
So, put it in through the facial vein or through the tail vein and then, in females
that are 7 to 12 months old. So, this is the real rats model and ask
what happens and their, their improvements are quite dramatic.
This is rotter rod, it’s a wheel you put a mouse on, turn it round slowly and they
fall off. But it takes them a while to fall off.
And actually, the first day they fall off more quickly than the second day when
they’ve learned a bit, and then the third day, they’re a bit better.
The animals, without MECP2 we’ve stopped MECP2 acquired sorry, without MECP2
acquired bad at this. But then, if you put in this virus, you’re
getting a significant improvement. This is the so-called inverted grid test,
which is simply taking the lid off the cage, and turning it upside down.
So, that’s rather a fancy name for that. And you see how long before they fall off.
And you see that the red the red bar is the way they are when you without MECP2.
And these, these are the rescued mice, they’ve improved a lot.
And the third test, and there are other tests I could show you, is the nesting
test. You weigh a certain amount of nesting
materials, you plonk it in the cage and then the next day you come back and see
how much of it they’ve used to make a nest.
So it gives you a number. And the not much of it is gone with the
mutant mice. A lot is gone with the normal mice, and
the rescued mice are vastly better. So I’ll finish up by showing you as the
final thing, a couple of mice. This is not as good as an experiment that
I showed you before. There, I showed you one mouse before and
after. This is, for obvious reasons a mouse
that’s had the virus, treated with an empty virus so it has the virus but it
didn’t have anything in it. And this is the, this is the way, it, the
symptoms looked in a females, that almost you can imagine the high, the, the, the,
the limb clasping very immobile and clearly not that healthy.
And then, this is an example of a mouse that whirls like that.
But now as, as a result, it’s a different mouse as a result of receiving the virus,
it’s vastly improved. This is the data of Saurabh Garg in the
Mandel Laboratory with whom we are collaborating.
So this, I would hesitate to say that this is a therapy that necessarily can be
adapted to humans rapidly because the viral load would be colossal.
The amount that gets into the brain is relatively small and 10% of humans have
antibodies against these viruses anyway. But it’s a basis.
So, research into the causes of Rett Syndrome is currently a hot area in
biomedical science. Epigenetics, yes, the epigenomes disturbed
in these mice and brain autism, these are fascinating areas that are combined in
this in this field. Findings over the past decades, actually
changed our perception of this disorder and by implication of others.
And by that, I mean, reversibility, because it was thought to be impossible.
The search for potential therapies is going on a pace.
I showed you what we’re doing. We are far from a learn, there’s lots
going on, and no approach is yet proven to work clinically.
There is a way to go yet, unfortunately, it’s frustrating that one feels one is
quite close if only one could engineer something that would do it.
And the goal remains to discover robust treatments that either reduce or even
eliminate the burden of Rett Syndrome. So, just to summarize then the whole
thing, we’ve ranged over quite a lot. Epigenetics is more or less how the genome
of living things is organized and managed. It’s a high level word.
There’s no worry about exactly what it means.
Every definition is, encompasses fascinating biology.
Cg is a genome signaling module, module. It’s very short, it sounds too short to be
useful but I hope you’re persuaded that actually it is used as a way of, say,
adapting regions, adapting regions of the genome to their function.
Proteins that read different chemical forms of CG, unmethylated, methylated,
lead to contrasting biological outcomes. And mismanaged, disorganized epigenomes
are involved in disease. And the extent to which they are involved
in disease, is actually profoundly unknown.
And, for that reason, epigenome manipulation, for example,
pharmacologically, may have therapeutic value in diversed human disorders.
Epigenetic drugs, so-called, are already in the clinic, the histone deacetylase is
its inhibitors, etc. And there are far more in the pipeline.
I would say that we don’t really know what they’re gonna be good for.
Because there could be all sorts of disorders where global epigenome mech,
manipulation has a consequence. And so, I think we’re in for exciting
times. This is my lab.
I, this is not my whole lab. This is just, it’s people who were in my
lab, whose data I’ve shown. People who are in my lab, whose, some of
whose data I’ve shown. Gail Mandel, who I’m grateful to for
allowing me to use our the, her data the data that was generated in her lab on the
gene therapy. Brian Kaspar, who made the viruses for her
and, and us. And Stuart Cobb and colleagues from
Glasgow University who are our nearest, tell us about neuroscience of which we’re
ignorant. And finally, all this plethora of funding
agencies from the very small to the very large who have made our work possible over
the years. And this is my lab.
I wanted to go to north of Scotland for our retreat.
They wanted to go somewhere hot and they won.
So, this is Barcelona again. Thank you very much.
>>Thank you, Adrian for a splendid lecture, and, and Adrian will now take
questions for ten minutes or so. There’s one over there.
If you if you have a question, if you’d like to ask one, put your hand up in
advance, there are two roving mics so we can send one near you so that we can keep
the questions coming.>>Hi there you recently published an
article in The New Scientist, about epigenetics I wonder if you’ve got time
just to mention a little bit about the post mortem suicide data that you came
across. You weren’t able to talk about it at
length in your article, but I wondered if you could briefly mention a point about
that?>>It’s not my data, it’s, it’s a report
that people who had committed suicide claiming to have been abused as children
were, had a different degree of methylation of a stress hormone receptor
promoter in their brains. And I don’t feel, and I’m not sure the
authors would feel despite their elevated status of the publication, that they have
achieved statistical significance with just 12 and also leaving out, as it says
in the method section, outliers, that presumably didn’t fall within their
average. So, I feel if you know, the way you treat
your children becomes hardwired in, into their lives at, through this epigenomic
mechanism I feel before one announces that to the world, one has to be pretty sure.
And I don’t feel, in this case, they could be.
>>At the back.>>If there’s a guard, guardian of the
genome, is there a guardian of the epigenome?
>>Sorry, I missed the very first phrase.>>If there’s a guardian of a genome, is
there a guardian of the epigenome?>>A guard, is there a guardian of the
epigenome, I don’t know, this is a phrase and people, I mean, guardians of the
genome are proteins like P53 that takes steps when the genome is damaged.
It’s not really clear if damaging the epigenome.
The epigenome is quite in a state of flux. There is no one epigenome.
Now, that’s really the problem with epigenome analysis, epigenomic analysis.
Every cell type will have a different one and actually, do , so it depends on your
stance, some people believe that the epigenome is somewhat fixed and the genes
operate within this rather inflexible set of rules.
I, from what I have seen, find the epigenome does what its told by the genes
and their transcription factors. In a way, the epigenome adapts the genome
to its function as determined by other proteins.
Now, that’s, there’s a question of degree between those 2 extremes.
But I don’t see the epigenome as something that gets fixed, and then is transmitted
forever, and you can’t do anything about it and even your grandchildren can’t do
anything about it, two generations later. That, there may be some of that going on.
But I feel it’s likely to be far less than is sometimes suggested.
It’s dried up the questions.>>Another one, 4 rows from the back.
Oh, there’s one here. Okay.
Speaker:[cough] Why does it take so long for clinical symptoms to become manifest,
both in the mice and in the humans?>>Let, that’s a very good question.
We actually haven’t the faintest idea, I mean, you could argue so in the case of
humans, the time when they get it, 18 months of age is a time of great activity
in the brain. And so, you, that sort of fuels the idea
that this is a neurodevelopmental disorder and you only really start getting the
problems when the brain is going through particular types of dance of the neurons
in particular synaptogenesis . But actually, in the, in the mouse, if one
looks at that, these mice are 6 months old, they’re, they’re not going through
any developmental processes at all as far as we know, they’re, they’re just
gradually aging like, like the rest of us. So in that case, it doesn’t quite fit.
And I think that the alternative hypothesis is, that without MECP2, the
functional half life of your neurons, not their lives but their functional half life
is reduced. And they, you, you’ve crossed some
threshold at which the brain stops to, stops working properly.
But actually, I don’t have a satisfactory answer for your question.
So, it’s, it’s one of the key questions.>>Hi, my questions is about gene therapy
and one of the limitations of this is that you can’t cause large enough change
throughout the entire tissue to correct the fault.
And I was wondering whether or not you thought that gene therapy had greater
therapeutic impications for epigenetic disease rather than genetic disease.
>>Well, I’m not quite sure about the link between the first bit of your, I mean, I
agree with the, the, the reservation that you have to hit a high, a high percentage
of cells. Actually, these adeno-associated viruses
do that. They’ve selected stereotypes.
They’re all naturally occurring, none of them have been engineered in any way and
so there’s scope to improve them. But they spread throughout the brain so
the systematic injection if you look in the brain, it’s everywhere.
So, it, it does get everywhere and if you do the injection into the brain, and these
are, there are clinical trials now for[UNKNOWN] disease, lysosomal storage
disease where there are 6 bore holes and they’re putting it in and they’re getting
a big spread through the brain of the, this virus.
They don’t, they don’t divide, and they keep churning out the protein for a very
long time. So, I think that sort of thing can be
solved. I wouldn’t say that gene therapy, I mean,
gene therapy, this is a blunder buster approach.
You’ve got, you’re shoving in an uncontrolled number of genomes into cells
different numbers into different cells. It’s, it’s the primitive end of what
hopefully ultimately will be a rather sophisticated therapy.
But I wouldn’t say it’s for epigenetic, rather than genetic.
I haven’t seen anything that suggests to me that it prefers one or the other.
>>Well, I, I was also going to ask, whether or not it was the case of being
the opposite when you said that the, the epigenome was constantly in flux.
So, does that mean using gene therapy to try and to correct the epigenome is
actually more difficult than gene therapy?>>What you’ve got to remember is that
Rett Syndrome is a genetic disorder. It’s, it’s it affects the epigenome.
There, there aren’t epigenetic disorders and genetic disorders necessarily.
This, the, the, the, the mutation is a standard mutation in a gene, and it’s
inherited in a Mandelian manner if you, you know, in those very rare families
where it’s transmitted. So it’s a genetic disorder that effects
the epigenome. It’s not an epigenetic or genetic
disorder.>>So, so, Adrian, do, do the virus
treated mice, do they stay cured?>>Well.
>>Or do they get well again when they get older?
>>That’s a good question and we don’t know because actually those pictures were
only taken within the last 4 weeks. They’ve lived for four weeks beyond that.
>>One at the very back.>>But actually, the model experiments say
that these things are expressed for a very long time.
They don’t ever get integrated, and for that reason, it seems they don’t seem so
susceptible to being shut down by the epigenetic mechanisms that are scouting
for strange things in the genome.>>Hello.
What’s your feeling about the big psychiatric disorders like schizophrenia
and bipolar affective disorder and so on? Do you think your research in these
epigenomatic processes, approaches are going to be important for those?
>>I don’t know, there’s a, you know people argue about autism.
As to whether or not it’s more environmental than genetic.
And there’s now really quite strong evidence that it’s genetic.
But it’s not one genetic disorder. It’s hundreds of genetic disorders,
actually, literally. There are very many genes that are
contributing to autism. That’s why it’s been so difficult to pin
down genetically. So, I would say autism, this work is
related to you know, relevant to, it’s difficult to know psychiatric disorders.
You know, they tend to, autism, despite being caused by large numbers of different
genes, where almost no two patients have the same set of genetic lesions,
nevertheless have common presentations, features that are in common among them.
So, it’s almost as though when there are problems with the brain, it gravitates
towards certain types of behavior. So, either it doesn’t work at all, in
which case, survival is, is in question or it gravitates toward certain types of
presentation. So, in other words, it says more about the
way in which the brain can cope than about the function of the proteins.
Now, I, I would say schizophrenia is a very interesting question.
They’re having meetings about schizophrenia and bipolar, always been
groups that go off and talk about the possibility that’s it’s pure, called
what’s purely by epigenetics and the environment.
But I think, increasingly, as more and more sequencing gets done, my bias would
be that they will find, we will find genetic causes for these disorders, and
then all of the[UNKNOWN] of ways of dealing with genetic disorders will be
drawn upon to try and fix that. To me, one surprising thing of that, if I
may,[UNKNOWN], the brain, you would have thought, is the most inaccessible place to
ever do stuff like gene therapy. But actually, it’s quite a good place
because the cells don’t divide, there’s not there is immune response in there but
it’s nowhere near as virulent as it is in other places, so you can do more stuff in
the brain and things will spread through the brain, so maybe that will apply to the
disorders you are talking about, but it’s really a long way away, I think.
>>Is anyone working on Friedreich’s ataxia?
That’s where they.>>Lots of people are working on, on
Friedreich’s ataxia but you’re not talking to the right person to ask about it
unfortunately. I mean, I, I, I would need a quick
reminder about exactly what the lesion is, is it DNA repair?
I can’t remember, can anyone remember? Dna repair, yes, that would be very
different if it’s, if the lesions due to DNA damage then, then that’s not really
quite in the league of epigenetics as we’ve been discussing it.
But rest assured, Friedreich’s ataxia is a very active area of research.
>>Okay. There’s one there.
Then, one in the green cardigan and then, one against the wall and we might have to
wrap it up that, that point.>>Yeah.
Fantastic talk, Adrian. It’s interesting that you observed obesity
in the, the mice. Were any aspects of feeding or weights
changed with the adeno virus treatment?>>Well, unfortunately, we’ve never really
investigated that. I think they don’t eat more, I think it’s
probably but, but we don’t really know. All I can tell you when you do the
reversal, it’s the nicest thing to watch, you just simply watch them come down over
a period of three weeks, their weight goes from being obese to being normal.
So, but on the other hand, we’re, we’re activating MECP2 in every cell in the
body. And that actually is an area we’re
interested in now because are there also peripheral, neuroscientists call
everything that isn’t the brain, the periphery, for some reason but are there
peripheral phenotypes that we’ve been missing, perhaps.
>>Okay, I hope these are two quick questions.
>>Hello, I was just, just to ask about that.
So you, can you see more[UNKNOWN] cells in other organisms, or organs, and does they
have>>We’ve never looked.>>And some genotype.
And, and, another more thing, so if it’s not like that, it means that in some
cases, some cells will be dying because they don’t express it, and so they are
kind of selected and because in the brain they cannot.
>>No, but that’s, that doesn’t happen, you see.
>>Okay.>>You, you, you can, the cells without it
don’t die. That’s why reversal is such an interesting
possibility, and in fact, with respect to[UNKNOWN] and other tissues, there are
so, there’s quite a variation in the severity of Rett Syndrome.
And quite often, it’s due to skewed x chromosome inactivation in the population.
There’s quite a variation, it isn’t always 50-50, half of the paternal, half of the
maternal. Quite often, it’s very skewed, like one in
10, one in 20 one in one and two in 10, it’s very skewed.
When the skewing is against the mutated version, the symptoms are much milder and
so you get what’s called a speech preserved version where, you know, there,
there is a spectrum to normality. Now, you did measure that by looking at
the blood. So, presume, whatever you find in the
blood looks as though, quite often, it reflects what’s going on in the brain.
>>Okay, thanks. Is there one last one?
>>Thank you. Do all animals have epigenomes?
Or is it just a characteristic of the higher animals?
>>Even, even yeast has marks on its genome.
Actually, yeast doesn’t have DNA methylation.
C elegans, this worm, doesn’t have DNA methylation.
Drosophila melanogaster has virtually no DNA methylation.
So, the models that are quite often worked on, don’t.
That’s chosen actually quite often because they’re, they have a small genome rapid
generation time. But those animals, nevertheless, have what
one would call an epigenome, because they have histone tails covered in marks.
So, these are quite conserved to, throughout all, all organisms that one
would call eukaryotes. That’s animals, plants, fungi.
>>Okay on that affinitive note, the I think we could have kept on going for much
longer with the questions but Adrian has another rest of his program to get
through, so I was asked to bring this to a halt the floor, the floor now actually.
I think you’ve done rather well out of it. As expected, Adrian has given us a
wonderful lecture, wide ranging, quite challenging and he has covered a, a very
large area so thank you very much for that.
And now, I want, going to ask Patrick Vallance, who is the President for
Pharmaceutical R&D at GSK to come and present Adrian with his medal.
Patrick joined GSK as Head of Drug Discovery, I think in 2000?
2000 and 2006. And he’s now President of R&D
Pharmaceuticals. What’s that?
Oh, yes he’s the man with the monies. He’s the one actually giving the check.
>>Yeah, I’m the man with the check. And you’re giving the medal.
Adrian, thank you very much, it was absolutely terrific, and it’s a real
privilege, for GSK to be able to fund this lecture and prize.
And it’s been a privilege for GSK, or its precursor companies, for 32 years, as Jean
said, that we started with the Wellcome Foundation in 1980, and then,
GlaxoWellcome, and now, GSK. And some things have stayed constant over
those times and some things have changed. The company is much, much bigger.
The company is totally global. And, of course, the whole nature of drug
discovery and development has changed. But some things have stayed constant.
And one of the things that stayed constant is our base in the UK and our commitment
to the UK and just about 50% of our R&D activity is in the UK.
And one of the reasons that we’re here is excellence of the science base and I think
that’s been admirably demonstrated today, Adrian, by what you’ve said and, of
course, is embedded in the values in of this institution.
The second thing that stay constant, perhaps not constant, but in a way,
started at the beginning, and is very, very important to us now, is a very close
working relationship with academics. I think it was absolutely the hallmark of
the Wellcome Foundation. And I think it’s absolutely the hallmark
of what we’re doing now. And it’s perhaps no surprise, that I think
it was 5 or 6 years ago when we decided that we needed to understand what the
opportunities were in epigenetics for drug discovery.
We reached out to the very leading academics in this field to find out what
we should be thinking about. And Adrian, it was you that came in to
talk to some of our team to help us to get started on that.
So, it’s terrific to hear this today. Unbelievably impressive and Jean gives the
lasting thing, which is the medal and I give the transient thing, which is the
money. But I hope it brings some pleasure and
thank you very much indeed.>>Thank you very much.
Thank you very much, thank you.