Week 2 Topic 3 - Exploring mental health using stem cells: What are iPSCs?

Question 1

PART 2

Answer 1

OK, in the previous session, we learned how Shinya Yamanaka and his colleagues taught us how to
make IPS cells, Induced Pluripotent Stem cells, starting really from any tissue sample. And I explained
how in our laboratory, we start from hair samples, but you can start from blood, or skin, and
probably any tissue– and how using those, we make these IPS colonies.
From the IPS colonies, we make IPS lines. So in this session, I wanted to explain what you can do with
them, particularly what you can do with them if, like me, you’re interested in brain development, and
you are interested in disorders in brain development.

Question 2

PART 3

Answer 2

So this is the final session in this series about iPS cells– induced pluripotent stem cells– and how we
can generate them and use them as cellular models.
At the start of the session, I suggested to you a thought experiment. And the thought experiment
was this– wouldn’t it be nice if we could reach back in development and study how developmental
disorders, like autism, actually arose in the brain of children in utero? Who are going to go on and
get a disorder? I suggested that we could use these cells that I’ve been telling you about– these iPS
cells– to do precisely that.

Question 3

What does iPSC stand for?

Answer 3

iPSC stands for

induced Pluripotent Stem Cells,

Question 4

What are the three stages of basic

embryology that lead to the development of PLURITPOTENT INNER MASS CELLS ?

Answer 4

To understand pluripotentiality, we really have to go back right to the start of
embryonic development in mammals, including humans.

So to remind you of some sort of basic
embryology,

1. life starts with a fertilised egg– sperm fertilising the oocyte.
2. This fertilised egg grows, divides, forms a morula, and eventually it forms the structure we call
   a blastocyst.
3. blastocyst is the first stage in development where you can see two distinct
   populations of cells.

a. So there are these outer cells here, the trophoblast cells, shown in yellow in this
diagram.

b. And these are the cells that we are interested in are the inner cell mass. And it’s these blue inner cell mass cells that we’re interested in, because these are the cells that are pluripotent.

Question 5

So what does pluripotent mean?

Answer 5

what pluripotent means is that these cells - THE INNER MASS CELLS - have the ability– the capacity– to
generate all the different cell types that make up both the foetal and then, later, the adult body.

So, they can generate the immune system, the nervous system, the heart and circulatory system,
muscles, and all the rest.

So, that’s the definition of pluripotent. These are cells that can make
everything that makes up the body.

Question 6

What is one of the most important property of pluripotent cells in embryological?

Answer 6

EPHEMERAL

Now, the other point about this idea of pluripotency– although it’s the most important concept you
can imagine, really, in embryology, the property itself is incredibly ephemeral. So, these inner cell
mass cells have this property of pluripotency, but they only have it for a few days, and then they lose
it as these cells give rise to derivatives that are going to go on and make these different lineages of
cells that make up the body.

So, it’s a very ephemeral property. It’s there for a few days, then it disappears. And as far as
we know, it never really reappears again during the entire life cycle of the organism. There’s
some dispute about that– there might be some populations of pluripotent cells crop up in other
circumstances later. But, by and large, it looks as if these inner cell mass cells, for these few days,
are the only cells that are ever pluripotent.

Now, that makes it a very interesting property, needless to say, for an embryologist. But it also makes
it a very difficult property to study, because if a property is only held by a handful of cells for a very
short period of time in development, how on earth are you going to start to understand what it
means in biological terms.

Question 7

How do we study pluripotent cells since they are so ephemeral?

Answer 7

EMBRYONIC STEM CELLS
And the real key to progress in this area has been the development of what we now call ES cells–
Embryonic Stem cells. Embryonic stem cells were first derived from mouse, but have subsequently
been derived from lots of different species, including, now, human.

And the key here is– what ES cells are is a population of stem cells derived directly from the inner
cell mass.

And the key was the generation of culture properties that assured that the ES cells
retained this pluripotency.

So, whereas for the inner cell mass cells, pluripotency is this ephemeral property that only exists
for a few hours or days, in the ES cell lines derived from the inner cell mass cells, it’s a permanent
property.

Question 8

How long to ES cells retain the property of pluripotency?

Answer 8

ES cells are permanent cell lines that have this property of pluripotency forever.

So, you
can grow up as many of these cells as you like, and keep growing them, keep referring back to them.
But these cells are pluripotent in exactly the same way as inner cell mass cells are, and that means,
of course, just like the inner cell mass cells, they can give rise to all these different cell types that
make up the body.

Question 9

So, what’s the biological basis for pluripotency? Who did the experiment that demonstrated the biological basis for pluripotency?

Answer 9

Well the answer is that until very, very recently, we
had a very poor grasp of that. But the experiment that gave us the first clue as how to go about
looking for the biological basis for pluripotency came from an experiment done by this gentleman,

JOHN GURDON
John Gurdon– then at Oxford, now in Cambridge.
And what he did was this experiment.

He was working on frogs, and he took eggs from frogs– so, an
egg being the pluripotent cell of a frog, just equivalent, as it were, to the inner cell mass that we were
just discussing.

1. But what he did was he destroyed the nucleus of those egg cells. So, he used radiation, which would
   destroy the DNA. And sure enough, he was able to generate what he called enucleated eggs. So,
   these are pluripotent egg cells missing a nucleus, now.
2. What he does then is he takes cells from fully differentiated tissue– in this case, the skin. So, he takes
   skin fibroblasts– also from frogs– and he does the reverse experiment. So, what he does now is he
   takes the nucleus from those skin cells, and then the experiment is to combine this nucleus with this
   enucleated egg.
3. So, what he’s got now is an egg with a nucleus transplanted from this fully differentiated skin cell. And
   the question, obviously, is, can this egg– this egg with a transplanted nucleus– go on and develop into
   a tadpole. In other words, is this cell– this constructed cell– is it pluripotent?
4. And the answer was that it was. So, he was able to show in a convincing number of these
   circumstances where he transplanted the nucleus into the egg– he was able to show that in fact that
   cell could go on and form tadpoles. So it was, indeed, pluripotent.

Question 10

What is a blastocyst?

Answer 10

A cell? a part? A thing? In the first stage in development where you can see two distinct populations of cells:

It possesses an inner cell mass (ICM) which subsequently forms the embryo.

The outer layer of the blastocyst consists of cells collectively called the trophoblast.

Question 11

What is a trophoblast cell?

Answer 11

The outer layer of a blastocyst, I think.

Question 12

What can we conclude about Gurdon’s frog experiment?

Answer 12

THE NUCLEUS HAS ALL THE GENES REQUIRED TO GENERATE AN EENTIRE ORGANISM
1. There was a number of different conclusions to do with the fact that it shows that this nucleus, even though it’s from a fully differentiated cell, nonetheless still has
all the genes– all the information required to generate an entire organism.So, that was actually the
primary reason he did the experiment.
But we can actually take another conclusion from this. What we can conclude is the following.

PLURIPOTENTCY RESIDES IN THE CYTOPLASM
2. If we ask, where does pluripotency reside, it’s clear that it’s coming with the enucleated egg. In other
words, the pluripotency is captured here, in the cytoplasm of this enucleated egg. And we can
conclude that because this cell has been able to instruct this nucleus– that was derived from the
differentiated cell– to act in a pluripotent manner.
So, in other words, we can conclude that there must be factors in this enucleated egg that tell this
nucleus– once the two become combined– that tell this nucleus that it’s got to start turning on the
genes, behaving in a way appropriate for a pluripotent cell. So, we can conclude there are factors in
the cytoplasm of pluripotent cells that dictate pluripotency.

That’s a very important conclusion, and it
allows us to beg the next question, which is, what are those factors?
Slide 7
So, the Gurdon experiment told us there must be factors in the cytoplasm of that pluripotent
cell that capture the essence of pluripotency.

Question 13

Who is Yamanaka and what did he do and when?

Answer 13

YAMANAKA DISCOVERED THE FOUR FACTORS IN THE CYTOPLASM THAT DICTATE PLURIPOTENCY in 2007.

He’d found the four factors that
were both necessary and sufficient to generate pluripotent cells, starting from just common-orgarden fibroblasts.

The following year, he was able to show exactly
the same process starting with human fibroblasts.

So the question is what are those FACTORS in
the cytoplasm of pluripotent cells that dictate pluripotency? The
real breakthrough in understanding those factors comes from the work of this gentleman, Shinya
Yamanaka, in work that he published in 2006 and then again in 2007. And his approach to this problem
was the following.
He thought that the factors– first off, they must be, themselves, gene products. In other words,
they must be proteins. And he looked at the literature, looking at what had been published about
pluripotent cells, and he came up with a list of 24 factors that seemed– from the research of others
and from himself– that seemed to be associated with pluripotency.
So, these are genes that seem to always be expressed, or always be around, when pluripotency was
being studied. So, he figured that probably somewhere in this list of 24 factors must be the ones– the
really important ones– that are dictating pluripotency in pluripotent cells.
The problem, of course, is how is he going to show that the 24 factors do indeed include the
important ones, and secondly, how is he going to find out which of those 24 really are important, and
which, perhaps, are less important. So, this was the experiment that he set out to do.
Slide 8
So, like all good biochemists, he realised that step number one– he had to have an assay. He had to
have a way of recognising if he’d produced pluripotency in cells that weren’t pluripotent. And just like
John Gurdon, he chose to start with fibroblasts– so, again, skin fibroblasts. And the challenge was,
could he make those skin fibroblasts become pluripotent. And if he did, his assay would have to be,
how would he recognise that had happened.
And the key for him was to use this construct. So, what this is is a gene, Fbx15. And the actual nature
of the gene doesn’t really concern as very much today. What’s important is, he recognised that
pluripotent cells always seem to have this gene active.
So, he figured that if he could switch on this gene in fibroblasts, then maybe he’d actually come
up with a strategy that made them pluripotent. In other words, he was using this Fbx15 locus as a
reporter, as we would term it, for pluripotency.
© King’s College London 2019 4.
Transcripts by 3Playmedia
Week 2
So, he made this construct whereby if this gene became active, it would turn on this reporter
construct that we call beta-geo. And as a consequence of that, if you stain the cells appropriately,
they turn blue. So the experiment, now, is can he– by putting those 24 factors into the cells, into
fibroblasts– can he turn on this beta-geo construct? Can he turn the cells blue?
So, what he next did is actually a real tour de force– so, a really impressive piece of molecular
biology, because first off, he had to make retroviral constructs of all of those 24 genes. He had
to have a way of getting all 24 genes into these fibroblasts. So he made 24 different retroviral
constructs.
He then had to engineer a situation where he could infect the cells with all 24 genes simultaneously.
And what he found was the following– that if he was able to introduce all 24 retroviruses into a
population of fibroblasts, then some of them did indeed turn blue.
And that’s shown over here. So, this is a plate of fibroblasts at very, very low power, so you can’t
see individual cells. And these are what’s called the mock-infected. So, these are the control cells.
But this is the plate of cells that were infected with all 24 retroviral vectors. And what you can
immediately see– and what Yamanaka immediately saw– was that there are individual colonies that
have started to emerge, and they’re stained– they’re stained blue.
So, in other words, by transducing in all 24 factors, he’d been able to show that some of those
transfected fibroblasts turned on the Fbx15 locus, and therefore, putatively, the idea would be that
maybe they had become pluripotent. So, just that step is an amazing step forward, because what
he’s shown is that his first bet is correct. Somewhere in that 24 factors are the important ones that
induce pluripotency in otherwise non-pluripotent fibroblasts.
Slide 9
So, so far so good. But now he’s got another problem. How is he going to work out which of the
24 are really necessary, and which are perhaps not necessary? So, what he takes on now is an
enormous experiment. So, what he’s going to do now is he’s going to repeat the experiment I’ve just
shown you.
He’s going to repeat the infection of these fibroblasts. But what he’s going to do– instead of infecting
with 24 factors, he’s going to infect with 23 factors. And he’s going to do it over and over, each time
leaving a factor out.
And the argument is that if it leaves out a factor that really isn’t necessary, then he should still see
blue colonies, whereas if he leaves out a factor that’s important, now he won’t be able to generate
the colonies. So, now what you’re looking at here is a whole series of experiments where he’s used
the 24 factors minus one. So, each one of these experiments, he’s used 23 factors– the 24 minus
one.
And what you can see, indicated by the green arrows, are circumstances where leaving out a factor
made no difference. What you’re looking at here is the number of colonies he was able to observe–
the number of blue colonies that emerged in these fibroblasts.
And you can see leaving out factor 2, you still get lots of colonies. So factor 2 clearly isn’t very
important. And similarly, these other factors, indicated with green arrows– he could leave those out
and the experiment still worked. So those weren’t necessary to induce pluripotency.
But you can also see that there are some factors that, if he leaves them out, the experiment no
longer works. So if he leaves those out, he doesn’t get any blue colonies. So, the conclusion is that
those factors are probably necessary to induce pluripotency.
© King’s College London 2019 5.
Transcripts by 3Playmedia
Week 2
Slide 10
As a consequence of that experiment, he takes– he’s able to narrow down these 24 factors down to
10 factors. These are the 10 most promising-looking factors that emerged from this experiment. So,
what does he do now? Well, he repeats the experiment that I’ve just shown you, but this time using 10
factors.
So, first off he confirms here that by adding all 10 factors he does indeed induce pluripotency– he
gets colonies. So, what he does now is he leaves out one of the factors again. So, each of these
experiments is one of the 10– is the 10 factors with one factor missing. And again, what you can see is
that in some cases he can leave the factors out, and it makes no difference.
But there were four factors– four factors that, if he left factor out, either there was no
reprogramming to blue cells, or– in the case of this one here– there was some, but it was very much
less efficient. So, a much reduced efficiency.
So, now he’s down to these four factors. And these are the four factors we’ve now come to call the
Yamanaka factors. So, Oct3/4, Klf4, Sox2, and c-Myc. And it seems like these are the four that really
matter.
And to confirm that, he does this experiment. So, now what he shows is that just those four factors
still give him lots of colonies. So, if he uses just the four, he still gets conversion of the cells into blue
cells– so, putatively, pluripotent cells. And now, if he leaves out any one of those four, it works much
less efficiently. And any two factors really don’t work at all.
So, he concludes from that that these four factors are all necessary to generate pluripotent cells,
and the four factors together are sufficient on their own. He doesn’t need any other of the 24 factors,
and nothing else is required to generate pluripotent cells.
Slide 11
So, this is a remarkable breakthrough, because he’s been able to identify four factors that seem to
carry this property of pluripotency. But hold on– all he’s shown so far is that those four factors can
turn fibroblasts blue using that reporter construct. It’s still only a hypothesis that those blue cells–
those cells that turn on the blue gene– truly are pluripotent.
To convince himself and other scientists that those cells really were pluripotent, he has to really show
that they really can do the job of generating all the different cell types that make up the body. So he
does that, first of all, by generating what we call embryoid bodies.
So, if you take pluripotent cells– for example, embryonic stem cells– and you get them to grow
in clusters, and treat them in an appropriate way, they’ll start to differentiate into little clusters of
differentiated cells. And amongst these clusters, you will find all the different lineages that make up
the body.
In particular, you’ll find derivatives from what we call each of the three major germ layers– that’s the
endoderm, the ectoderm, and the mesoderm. And this is generally taken to be a good in vitro assay
of pluripotency.
Slide 12
So the question was, then, could he take these fibroblasts that have been transduced with the
four factors, grow them as embryoid bodies, and show that each of the three germ layers were
represented within the embryoid bodies. And what you can see here, in this figure, is the evidence
© King’s College London 2019 6.
Transcripts by 3Playmedia
Week 2
that he can do exactly that.
So, these are histological sections through embryoid bodies derived entirely from those transduced
fibroblasts. And what you can see is that there are mesoderm derivatives, like muscles and smooth
muscle, and cartilage, there was epithelial tissue, there was brain tissue, there was adipose tissue. In
other words, these cells had managed to go on and make all of these different cell lineages.
So, that was remarkable. It seemed to show that those cells were indeed pluripotent. Nonetheless,
Yamanaka wanted to go one step further, because if those cells were really pluripotent, then he
should be able to use them to contribute to the cells actually in a living mouse.
So, ES cells, the other pluripotent cell type we’ve talked about– if you inject those into a growing
mouse blastocyst, they will contribute to the development of the mouse. And they will contribute
not just to the embryoid body formation you’re of seeing here, but actually to cells in a mouse as it
develops.
So the question was, could he use these induced pluripotent fibroblasts– could he use those to
contribute to mouse development in the same way? Well, what this figure shows is in fact that he
could.
So, here you’re looking at two mouse foetuses. The one on the right is mock-transfected– so, it
didn’t receive the four factors. But the one on the left is a mouse that, as a blastocyst, was injected
with those fibroblasts that had been transduced with the four factors– and also they’d been labelled
green so that we could see what happened to them.
And what you can see in the fluorescent image over here is that this mouse has had green cells
contribute to lots of the different tissues of the body. So, those fibroblasts that had been transduced
with the four factors and are putatively pluripotent can indeed contribute to lots of different cell
types if you inject them into a mouse blastocyst.
So, he was able to confirm that these green cells had indeed contributed to all the different tissues
of the body by taking histological sections of mice, like this one. And you can see the stained cells
are in lots of different tissues– so, in the neural tube, in the liver, the heart, the gastrointestinal tract,
gonads, and so on and so forth. So, this confirmed that the cells that he’d injected into the blastocyst
of this embryo had indeed contributed to all the different tissues of the body.
Now, importantly, they contribute also to the germ cells. So, you can see cells here, resident in the
gonad. And what that meant was that they contributed– the fibroblasts that had been injected into
this blastocyst contributed to the germ line such that, when mice like this were grown up to adulthood
and bred, he was able to produce mice that were entirely derived from these injected fibroblasts.
Slide 13
Let me spell that out again. The fibroblasts that have been transduced with the four factors are
introduced into the blastocyst at this very, very early stage. The embryo grows up.
Some of those blastocysts– injected fibroblasts contribute to the germ line of this animal, such
that when this animal is bred downstream, those germ cells contribute to the next generation of
mice. Such that you’re able to generate a whole line of mice derived entirely from those transduced
fibroblasts.
So, you can see that this is real evidence that those cells– those skin fibroblasts that have been
transduced with those four factors– really were pluripotent. They weren’t just turning on the
reporter gene in vitro. They weren’t just giving rise to differentiated cells in embryoid bodies. They
were truly able to contribute to the generation of an animal in exactly the same way as the inner cell
© King’s College London 2019 7.
Transcripts by 3Playmedia
Week 2
mass cells would.
So, Yamanaka had really found the biological basis for pluripotency.

He’d found the four factors that
were both necessary and sufficient to generate pluripotent cells, starting from just common-orgarden fibroblasts.
****

Slide 14
So, Yamanaka reported the work that I’ve just told you about in 2006. In 2007, he published a paper
showing that you could do essentially exactly the same thing, starting with human cells. So, the
original publication had been with mouse fibroblasts. The following year, he was able to show exactly
the same process starting with human fibroblasts.

Question 14

What are the changes to the Yamanaka protocol since 2007?

Answer 14

1. It’s been made slightly more sophisticated. We can deliver the genes in slightly different, slightly better
   ways. But, in essence, the procedure that Yamanaka demonstrated to us has now been broadly
   adopted, really, worldwide.
2. So, many, many labs across the world now are generating these induced
   pluripotent stem cells– these iPS cells, as Yamanaka christened them– by using essentially his
   protocol.
3. So, one of the various things we’ve learned during that period is that Yamanaka started with skin
   fibroblasts. In fact, you can start with essentially any cell type in the body. So, for example, several
   groups now have shown that you can make iPS cells starting from blood cells.
   A couple of colleagues that I know about have made iPS cells starting simply with the cells that you’re
   able to centrifuge out of urine– so, sloughed off bladder cells that can be isolated from urine.
   In our lab and a number of other labs, we’ve started from a slightly different sample. So, we start
   with hair. And the reason for that is we’re quite interested– as I’m going to go on to tell you about a
   little later– in disorders of childhood. So, we’re interested to be able to collect biological samples
   from children. And we weren’t really keen on trying to take skin biopsies, or even blood, from autistic
   children.
   But what we can take really quite easily are hair samples. So, we pluck a hair, or a small number of
   scalp hairs, just from the head of a child. And from the bulb at the end of the hair, we can grow a
   population of cells– of so-called hair keratinocytes.
   We can then use Yamanaka’s four factors– engineered now into a different kind of vector, but
   fundamentally the same as the way he did it– and

HAIR CELLS
4. from these keratinocytes, we can grow colonies
of iPS cells. We can grow these up, expand them, freeze them down, and generate really enormous
quantities of these iPS cells. And they’re the substrate for the experiments that I’m going to go on
and tell you about in the next section

Question 15

We’re obviously interested in nerve cells. So the first job for us is to see if we can take these
pluripotent cells, which by definition, as you’ve already heard, can make all the different cell types
that make up the body. Can we turn them into neural cells?

Answer 15

And, in fact, we can.

So what we’re able to do by inducing the cells to adopt a neural fate is get them to make neural
progenitor cells. And from neural progenitor cells, to go on and make neurons.

Question 16

What is the process of taking iPSCs and making them neural progenitor cells and then neurons?

iPSC -> NPC - > Neurons

Answer 16

So let’s look at this process in just a little bit more detail.

1. So we start with this IPS cell line. We’ve
   made many IPS lines, and many are available now worldwide.
2. And the first job is to induce them to
   make neuroepithelium, rather than anything else. We’ve learned to do that by inhibiting what are
   called the Smad signalling pathways.
   And the advantage of that is that we’re able to make use of a lot of embryology that we’ve
   understood now for really quite a few years. So we know from studies principally of mice, but also
   of other organisms, precisely what it is at the early stages of normal development that push the
   pluripotent cells, those inner cell mast cells, to go on and give different tissue types.
   And we know that if we inhibit the Smad pathways that are normally driven by a class of morphogens
   called BMPs, Bone Morphogenic Proteins, then we can induce neuroepithelium. And sure enough,
   that’s what happens. So the first step in our procedure is to add what we call Smad inhibitors to the
   culture of these induced pluripotent cells.
3. And if we culture those cells, we then start to see these neural progenitors. And the first thing to
   notice about them that makes them really quite distinct from neural stem cells that we’ve come
   across in other contexts is that they really can do tissue histogenesis. That’s something I’m going to
   have more to say about in a couple of minutes.
   But notice how they make proper polarised neuroepithelium. What these cells are trying to do here
   is essentially trying to make a neural tube in two dimensions. So this is a two-dimensional culture,
   so a monolayer culture, if you will. But nonetheless, the cells are rounding up to make these what
   we call neural rosettes, these sort of flower-shaped structures, where they’ve got an actual apical
   centre just as the apical centre is in the neural tube surrounded by the basal processes of these
   neuroepithelium cells.
4. So these neural progenitor cells will go on and make neurons.
5. And if you treat them appropriately,
   you can get them to form post-mitotic neurons. And these post-mitotic neurons will become
   increasingly mature.
6. And over days and weeks, they will give rise to mature, physiologically-active
   neurons.

Question 17

What are Smad signalling pathways?

Answer 17

inhibiting what are

called the Smad signalling pathways induces the IPSC to make neuroepithelium which leads to neural progenitor cells

Question 18

How long does it take to make a neuron from a IPSC?

Answer 18

35 - 50 DAYS for human neurons.

Now, one of the things to point out at this juncture is that this process is really quite slow, which in
one sense, is a bit of a nuisance, of course. From an experimental point of view, it takes a long time
to start from the immature IPS cells, pluripotent IPS cells, and get all the way out to even halfway
mature neurons.
So this is the sort of timing of this progression in days from day one, for example, to day 35 or 50 out
here. But of course, that’s reassuring, because our human development is slow compared to the kind
of experimental animals we’re normally used to dealing with such as rodents. So this system really
does reproduce the timing and the differentiation processes that seem to underly human neural
development.

Question 19

After inhibiting the Smad signalling pathways of an IPSC and creating some neurons, what are you left with?

Answer 19

And this just gives you an example of what you end up with. So this is a plate of, in this case, cerebral
cortical neurons. And it’s really remarkable. So there we are. We’ve got to plate of, I don’t know, a
billion cortical neurons. Just think how difficult it would be normally to get hold of a plate of a billion
human cortical neurons on which you could do experimentation.
But we can get a billion today, and then I can get a billion tomorrow exactly the same. So it really is
an enabling technology that finally allows you to look at neural development in a culture dish– proper
human neural development.

Question 20

What does histogenesis mean?

Answer 20

series of organized, integrated processes by which cells of the primary germ layers of an embryo differentiate and assume the characteristics of the tissues into which they will develop into

Question 21

So, what makes the process of creating a “being” from iPSCs cease?

Answer 21

Now, I do want to pick up on this point about histogenesis. So I’ve pointed out that these cells really
do, when they neuralise, take on a neurepithelial structure. And they really do make a proper
polarised neuroepithelium. And if you allowed that to develop, the cells really do try and undergo
proper histogenesis. So you can see that down here if we look at this lower picture down here.
So here’s one of those neural rosettes that I was telling you about a little bit earlier– these polarised
neuroepithelial cells with their apical surface towards the centre and their basal surface towards the
outside. And these cells have been making neurons for a couple of days. And the neurons are the
cells stained red.
And you can immediately see in this example that the neurons have been generated by the
neuroepithelial cells. And just as in vivo, the cells have migrated out to come alive outside the
neuroepithelium. And this is an example of a little bit more advanced. And you can see there the
apical surface of the neuroepithelium, here are the neuroepithelial cells.

out here the neurons starting to form a lamina in what would be the basal surface of the
neuroepithelium.

And you can let this go further and further, and this is a much more advanced
structure. And you can see the collection of neurons out here just as you would normally see them in
vivo in a developing cortical plate.
So the cells really have a capacity for histogenesis. That other population of neural stem cells that
you might have come across in a different context other such populations simply do not have.

And you can push this really a long way. So this is a picture taken from this publication of Madeline
Lancaster’s in 2013. What she’s shown is that if you grow the cells appropriately in aggregate cultures,
you can actually make them make sort of mini brains– cerebral organoids, as she calls them.

And you can see that these organoids have got real structure. It’s a cortical structure out here
developed entirely from these induced pluripotent stem cells growing in a culture dish. So these cells
have a capacity for histogenesis that really is remarkable.

They really do try and build the nervous
system just in a culture.

**AT SOME POINT, THESE PRECURSORS TO A CNS BECOME DEPENDENT ON A BLOOD SUPPLY**
Now, eventually, this process ceases, and it can only go so far, because obviously, a brain can only get
so large before it really does become dependent on a blood supply. And there’s no blood supply in
these cultures. So there are enormous challenges to carrying this forward infinitely.
Nonetheless, these cells have a capacity for histogenesis we’ve really never seen before in other
neural developing systems in vitro. And the potential use of that for regenerative medicine is one that
I’m not going to talk anymore about today, but it’s fairly obvious.

Question 22

What is a neuroepithelium?

Answer 22

the part of the embryonic ectoderm that gives rise to the nervous system

Question 23

What is a neuroepithelium?

Answer 23

the part of the embryonic ectoderm that gives rise to the nervous system

[the ectoderm is the outermost of the three primary germ layers of the embryo; from it are derived the epidermis and epidermic tissues such as nails, hair, and glands of the skin; the nervous system; external sense organs such as the eye and ear; and the mucous membranes of the mouth and anus]

Question 24

What are THREE ways that
we use these cellular models/culture dishes full of iPSC neurons to study the aetiology of a disorder like autism?

or other neurodevelopmental disorders?

Answer 24

FIRSTLY

The first thing we can do is fairly obvious. We can take– we can make
iPS lines from lots of different individuals– both from cohorts of patients that have got particular
disabilities or particular neurovariants and from controls– so from neurotypical individuals.

We can
grow them in the way I’ve described to you– making neurons.

And we can compare.

And we can ask–
can we see any differences in the way the patient minds grow from the way– the way the control
minds grow?

So that’s one way we can approach disease aetiology using iPS cells.

SECONDLY
A second way would be to induce mutations in the iPS cells. So you’ll be aware there are a number
of ways now of inducing genetic variation– so genome editing as it’s usually called– into cells in vitro.
So the CRISPR-cas9 system is probably the most popular at the present time. But there are others,
like ZNFs and TALENs, that allow you to do similar things. So we could do that. And we know that
a number of different genetic variants are associated with different disorders. So we can induce
precisely those variants into the iPS cells and ask what difference does it make now to how they
develop into the type of neurons they give rise to?

THIRDLY
A third thing we could do would be to study environmental risk factors. So, for example, we know that
in autism, there’s a big increase in risk of a mum giving rise to a child who goes on to get autism if she
suffers from influenza during her first trimester. And we think we understand that that’s induced by
cytokines– pro-inflammatory cytokines– produced by the mother in response to the viral infection.
And those cytokines are able to exert an influence across the placenta on the development of the
foetus. Well, in principle, we can do that– we can study that. We can ask if we expose these iPS cells
to similar cytokines, what difference does it make to their development? And can we see anything in
that difference that we recognise as being possibly part of the aetiology of the disease?

Question 25

Summarize the development of iPSC to neurons.

Answer 25

1. We start with iPS cells.
2. We neuralise them to turn them into neural progenitor cells, and
3. then we slowly, but surely, differentiate them to young neurons and,
4. later, mature neurons.

Question 26

What does aetiology mean?

Answer 26

the cause, set of causes, or manner of causation of a disease or condition

Question 27

What kind of phenotypes might you expect to see in an experiment where you are messing with the environment of a iPSC?

In other words, what kind of assays are we
going to run on these cells to detect differences?

Answer 27

ONE

So one obvious thing we can do is look at gene
expression. So we can take lines from patients or controls, or we can induce mutations, or we can
expose the cells to environmental risk factors, and then ask– what difference does that make to the
expression of genes as the cells develop, both at the earlier stages of neuro development– like these
neural rosettes– or at the later stages of young neurons and more mature neurons?

TWO
We can also obviously look at physiology. So I’ve told you that these cells eventually become
physiologically active– so electrophysiologically active. They develop the kind of channels and the kind
of receptors that you would normally expect to see in human neurons. Well, we can ask– do any– in
any of these types of studies, do we see any difference in the types of electrophysiological properties
that the cells develop?

THREE
And a third difference we might look for might be what we might call morphogenetic difference. In
other words, I talked about the histogenesis that happens through this developmental profile. Well,
is that histogenesis altered in cells from patients versus controls? Do the cells grow a same size and
shape as they should? Do they form neurons in the same way as they should? Do those neurons start
to wire up and form the appropriate structures as they would in normal circumstances? So there’s
an example of some of the studies that we can do and some of the possible phenotypes we might
see.

Question 28

What are phenotypes?

Answer 28

the set of observable characteristics of an individual resulting from the interaction of its genotype with the environment

Question 29

What are assays?

Answer 29

the act of analyzing, or of conducting that test

Question 30

Describe the paper where they used iPS-derived

neurons to study the pathophysiology of a disorder called Timothy syndrome. [PASCA and Collegues]

Answer 30

Now, I don’t want to
talk about Timothy syndrome in any great detail, except to say that Timothy syndrome is known to
be caused by a particular mutation in a particular gene.

And the gene in question is this gene we–
C-A-C-N-A-1-C. We tend to refer to it as CACNA1C. And this gene encodes a calcium channel– this
particular calcium channel. And the gene encodes the alpha-one subunit of this calcium channel. And
we know from a whole series of other studies that calcium channels are very, very important in signal
transduction in neurons, but also in other cells, in fact.

So in this paper, the authors generated iPS cells from patients with Timothy syndrome and also from
other individuals– control individuals– who didn’t have Timothy syndrome– didn’t have this mutation
in the CACNA1C gene. And then they differentiated the iPS cells into neurons using a procedure
quite similar to the one that I’ve described to you. And then they asked the question that I posed
earlier– namely,

can we see any differences in the neurons that have been derived from the Timothy
syndrome patients versus those derived from controlled individuals?

And the first thing to look for,
obviously, was can they see any difference in the behaviour of the calcium channel? And they found
that they could.
So what you’re looking at here now is calcium flux in either control cells or cells derived from the
Timothy syndrome individuals. And you can immediately see there’s a difference associated with this
calcium channel mutation. And it clearly is the calcium channel because if you treat with this inhibitor
of calcium flux– inhibitor of the calcium channel you rectify it.
So this is very, very reassuring in the sense that these iPS cells show a phenotype that’s precisely
what you would predict, given that these cells carry this mutation of the calcium channel that we
know is associated with the disease. So that’s good and that’s very reassuring. But at the same
sense, it’s slightly predictable. We’ve taken cells here that have got a mutation in the calcium channel–
and, lo and behold, if you look at that calcium channel, it behaves in a mutated fashion. So you could
say it is no surprise here– this is just telling you precisely what you would have predicted.

But in this paper, they go and use these cells to do something a bit further. So what they look at here
is histogenesis– this process of building tissue that, as I’ve said, the iPS cells are really very good at.
So what they tried to generate was cortical neurons– cerebral cortical neurons.

FEATURES OF CORTICAL HISTOGENESIS
So let me just use this histological section over here to remind you of some features of cortical
histogenesis. So this is a piece– a section through a human cerebral cortex. This is the pial surface
out here. And this would be the ventricular zone. And here is the white matter and here is the grey
matter. What you can immediately see from this image– what, of course, you already know– is that
the cerebral cortex is a very strongly laminated structure. You can see the layers very, very clearly
within this piece of cortical tissue.
So in– these iPS cells– these differentiated iPS cells– they– these authors use different markers to
try and ask have they got the normal distribution of either upper-layer cortical neurons, like these
layer two-three cells here, compared to what are called infragranular cortical neurons– these
deeper layer five-six neurons.

And using different markers that label different neurons of different types, they are able to show that,
in the Timothy syndrome cells– here shown in red– there were fewer cells labelled with the markers
for lower neurons. And more of the cells labelled with markers indicating they were supragranular–
so the upper-layer neurons. So in other words, the neurons derived from the Timothy syndrome
iPSCs had a greater propensity to make upper-layer neurons, and a reduced propensity to make
lower-layer neurons.
Now when the authors looked at this in a bit more detail, what they discovered is, within this lowerlayer neuronal population, there was another disturbance.

FINDING 1
Namely, that a smaller proportion of the
lower-layer cells showed expression of gene SATB2. Now that rung bells. That seemed important,
because we know that these SATB2-positive cells, amongst these lower-layer neurons, are a
particular type of neuron.
So by and large, these lower-layer cortical neurons can take on either one of two fates. They either
become subcortically projecting neurons. Or they become what we call callosal projecting neurons.
Now the subcortical projecting neurons project, as the name suggests, to other regions of the brain–
the so-called subcortical regions. And those could be the thalamus, the striatum, the cerebellum, the
spinal cord.

FINDING 2
But the SATB2-positive cells belong to the other population. They belong to the callosal population.
And the callosal population get their name from the fact that they project across the corpus callosum
to the cerebral cortex on the other hemisphere. So what this observation says is that neurons from
the Timothy syndrome patients have a lower proportion of the SATB2-positive cells– that is, a lower
proportion of the callosal projecting neurons.

Now what the authors were also able to do is look at transgenic mice. And these were transgenic
mice that had been engineered to carry precisely the mutation that is found in Timothy syndrome.
So these are mice with the same Timothy syndrome mutation. And when they look at the cortical
structure in these mice, what they find is exactly the same thing as is being observed here. Namely,
that the number of SATB2-positive cells in the lower layer of the cortex of the mouse is fewer in the
Timothy syndrome mutated mouse than in the control mice. They see exactly the same histogenic
phenotype in the mice as they see in the cells that carry the calcium channel mutation.

So what do we make of that? Let’s summarise what we’ve just seen. So what we’re looking at here
is cells that carry a mutation in this CACNA1C gene and comparing them with cells that don’t have
that mutation. What we see is that the cells that have the CACNA1C mutation have a calcium channel
deficit. Now that’s reassuring, but it’s not surprising– the CACNA1C is, after all, a calcium channel
protein. But what we’ve seen beyond that is that those cells have an altered expression in the deep
cortical neuron population of this sacB2 gene. And that has led in the mice to an altered callosal
projection. In other words, an altered histogenic phenotype. So the development of the cortex seems
to be quite different in these cells. So this is an example of where we’ve been able to use iPS cells
derived from patients and ask some very fundamental questions about the impact that the mutation
associated with their disease on brain development.

Question 31

What are the pros and cons to using the iPSC cellular models to study neurodevelopmental disorders?

Answer 31

So what I’ve told you is that we can use these cellular
models to study neurodevelopmental disorders. What I want to suggest to you is that these models
that I’ve described have some real advantages compared to other ways that we might study such
disease. But I also want to tell you that they’ve got some disadvantages. So this is a balance.

PROS
So, amongst the pros– amongst the advantages– the first thing to say is obviously

1. these are human.
   So we’re used to having to study neural development in animal models. But what these cells are are
   actually true human cells. And I showed you that plate earlier, filled with cortical neurons, and pointed
   out that this was an enormous advantage that we’ve never had before– the ability to actually study
   human neurons developing in a culture dish.
   So one advantage is they’re human.
2. The second is they have what we call good construct validity. And
   that was– a good example was the Timothy syndrome paper I cited to you. These cells have precisely
   the mutation that we find in patients. Of course they do– they’re actually derived from those patients.
   And so the mutation that drives the disease– precisely that mutation– is there available in the iPS
   cells. So this was what we would describe as good construct validity for the disease. And that’s one
   of the assets of this model.
3. Another point I would make is that we’ve got good controls. So we’re able to compare the
   development of the cells with the mutation with cells derived from individuals that don’t have the
   mutation. So we’re very well controlled for that mutation in the analysis. I should say that we can go
   beyond that. I pointed out that we can engineer these cells– so we can either introduce mutations
   into cells or, indeed, we can use CASPR/Cas9-type genome editing to remove mutations from cells.
   And that way we can actually use same cells with the same genetic background with or without
   particular genetic variants.
4. The system, I would argue, is very tractable. So these are not the easiest cells to grow in culture. But,
   nonetheless, they are cells that one can readily culture. And that makes this a very approachable
   system that can be used– as it currently is being used worldwide– to study these developmental
   processes.
5. There is an enormous interest in the pharmaceutical industry in these models, because they seem
   like they would be amenable to high throughput screening. So we’ll be able to put these cells into
   drug discovery assays and use them to discover novel drug targets, novel therapeutics for some
   of the disorders that we’ve been discussing. And the cells that are genetically and phenotypically
   manipulable in the way that I’ve described to you.

So those are the advantages of this cellular system. Nonetheless, it’s important to point out there are
some problems that we’ve yet to totally overcome– some limitations to the system.

CONS

1. The first I point
   is the variability. So I’ve pointed out that we can take cells from different individuals, and compare
   their performance and the kind of assays that we’ve been talking about. But, of course, one of the
   problems is that no two individuals are the same, either genetically or epigenetically. And so, if we
   just took cells from maybe half a dozen different control individuals– so neurotypical individuals who
   didn’t have any neurodevelopmental diagnosis at all– we’d nonetheless find some degree of variability
   between them. And that variability would be more than we’d expect to see in similar cells taken from,
   say, mice. Where we’re able to control the genome and we know that every mouse in a colony is
   genetically identical– something that obviously isn’t true about individual humans.
2. The second problem is a more pervasive one. And that is that the disorders that we’re studying
   generally are going to be disorders of the system properties of the brain. What do I mean by that?
   What I mean is that a disorder, like schizophrenia or autism or ADHD, isn’t just going to be a property
   of a single population of neurons– let alone a single set of molecules within a single population of
   neurons. Rather they’re going to be disorders of properties that emerge from the brain working as
   a whole. And those properties are currently going to be inaccessible to us. So one of the challenges
   with the iPS system is to try and build models that will actually have system properties inbuilt. I
   showed you the beautiful picture of Madeline Lancaster’s where she’s able to grow the cerebral
   organoids. And I think we’re going to have to do more to try and push those forward– to try and
   get us a whole brain, if you will, or as close to whole brain as we can to start to understand how
   phenotypic differences exist in that sort of gross level.
3. I pointed out that development of iPS cells into neurons is reassuringly slow. It’s good that it’s slow,
   because human development is slow. But it’s infuriating that it’s slow if you want to get on and do
   experiments. There’s no question that the fact that you have to wait so long for your cells to develop
   is a disadvantage in a practical logistic sense.
4. And then the final point is really an important one. Most of the disorders– in fact, all the disorders
   I’ve been telling you about are really characterised by altered behaviour. These are behavioural
   phenotypes. An autistic child is defined as being autistic purely on the basis of clinically observed
   behavioural differences. Now our problem is these cells don’t behave– or at least they don’t show
   the kind of behaviour that we’d be interested in getting from patients. And so it’s going to remain an
   article of faith the extent to which the cellular and molecular phenotypes that we start to observe–
   and that I’ve told you about– actually relate to actual human behaviour.

So we’re going to continue
to require good clinical phenotyping, good clinical data. We’re going to continue to require animal
studies. Mice might not be human, but they do at least have observable behaviour that you can try
and relate to human behaviour.
So although this is a new model of neurodevelopmental disorders, I think it’s got tremendous utility.
It’s going to be a long time before it completely supersedes the other behavioral-based assays and
behavioral-based systems that we’re going to need access to. So thank you for listening. That’s my
presentation– or series of presentations– on cellular models of neurodevelopment. I hope it was
enjoyable.

Question 32

What does good construct validity mean?

Answer 32

Construct validity is “the degree to which a test measures what it claims, or purports, to be measuring.”

Question 33

What does tractable mean?

Answer 33

capable of being easily led, taught, or controlled