Human-Specific Genes and Neocortex Expansion in Development and Evolution


(uplifting piano music) – [Announcer] We are
the paradoxical eight. Bipedal, naked, large-brained, long the master of fire,
tools and language, but still trying to understand ourselves. Aware that death is inevitable, yet filled with optimism. We grow up slowly. We hand down knowledge. We empathize and deceive. We shape the future from
our shared understanding of the past. Carta brings together experts
from diverse disciplines to exchange insights on who we are and how we got here. And exploration made possible
by the generosity of humans like you. (digital music) – Okay, the question that
we have been interested in over the past 25 or so years, is how do brains in evolution get bigger? And to put the question in
a bit more specific terms, what underlies the evolutionary
increase in the number of neurons that are generated
in cortical development? And as you heard from
the previous speakers, this reflects the activity
of cortical stem and progenitor cells. And the answer, how you
get more neurons formed, is simple: all you need is
an increased and prolonged proliferation of these
cortical stem-cells. But the key question is, what
underlies the differences across species? Why do we have more neurons
than, say the macaque, or the chimp? Let me give you a very
brief introduction to these stem-cells. What you see here is a
developing mouse brain and the neocortex forms
from a structure called the dorsolateral telencephalon,
that is shown here in the box, and shown here at greater magnification. And what has fascinated
me, as a cell biologist over the years is that this cortical wall, this developing cortical
wall, has polarity. It has an apical side, a you
heard from previous speakers, and a basal side. And the process of making
neurons is a strict apical to basal process in which
progenitors that form the primary germinal layer,
the ventricular zone, progenitors that we collectively
call apical progenitors but notably the apical or
ventricular radial glia, these progenitors do not
make many neurons but rather, generate a secondary class
of progenitors that form the secondary layer of the
germinal zone called the subventricular zone. We call these collectively
basal progenitors and there are two major types: the
basal or outer radial glia and basal intermediate progenitors. And in a important paper
from Coleta colleagues it was shown that in more
highly developed brains, the outer subventricular
zone is very characteristic and it is very highly populated by basal radial glial cells. So these are the one implicated
in cortical expansion. And there is actually a cell
biological reason why these progenitors have advantages
to expand the cortex. The apical progenitors, because
of their cell biological nature, can divide only at the ventricle, and that’s a very limited space. In contrast, the basal
progenitors can divide anywhere in the subventricular zone. There they can thicken
the subventricular zone, and this is a huge advantage
if one wants to maximize the number of cell divisions. And this is the reason why
cortical expansion is linked to basal progenitor
abundance and proliferation. So a few years ago, we decided
then to embark on a search for human specific genes
that would do two things: increase basal progenitor generation, and also induce or promote
basal progenitor proliferation. And when I say we, I talk
about an outstanding PhD student in the lab: Marta Florio. What Marta has developed is
a cell biological approach to isolate these various,
these two major classes of stem-cells. This is a cartoon of the
developing cortical wall, and you can label the apical
side with an antibody against a protein called Prominin-1, and the basal side where the
membrane die that diffuses into plasma membrane called DiI. And with this approach, you
get a dual labeling of the apical radial glia and a
single labeling of the basal radial glia such that the
aRGs would be blue and red and the bRGs in the
mouse would only be red. Now the neurons would also be red, but in the mouse you have a
transgenic mouse line that, with the neurons in addition
green, and we can sort them away. So we get relatively pure populations. In humans this is more
complicated, and so we focused our attention to those cells
that had duplicated the DNA which turns them into
progenitors, so the aRGs are blue plus red, the bRGs are red
only, but please note that the bRGs in the G1 phase and of
the cell cycle so before they duplicate the DNA, would be
recovered in the neuron fraction shown here, and in fact we know
that about 10% of the cells in the neuron fraction are these cells. And this will become important
as you will see in a minute. What Marta then did was, she
isolated the cell populations and determined the transcriptome
and went for those genes that are highly expressed
in the stem-cells, and found a total of about 400 such genes
that were highly expressed in the basal and apical
radial glia but very low in the neurons. She then eliminated, from
these roughly 400 genes, those which were also expressed in
the mouse stem-cells which brought the number down
to 266, and then of these, she again eliminated those
that were also expressed in the human neurons or human cortical
plate, which brought the number down to 263. These 263 genes come in
two classes: 207 have an rtholog in the mouse
genome, but the gene is not expressed in the mouse
stem-cells, and 56 have no ortholog in the mouse
genome but exist, of course, in the human genome. These were for us the
more interesting ones. And now remember that we had,
we noticed by the approach of our way of isolating
these cells, that 10% of the cells in the neuron
fraction would actually be the stem-cells, the basal
radial glia, so we introduced another filter and asked which
of these 56 genes is at least tenfold more highly expressed
in the stem-cells than in the neurons, and amazingly
that reduced the number down to one. This one gene has this
pattern of expression: it is highly expressed in the
apical and basal radial glia, and this is the contamination
if you wish, of the bRGs in G1 and the neuron fraction. This gene is called
ARHGAP11B, and as I said it is specifically expressed in
human but not mouse apical and basal radial glia, but not neurons. This gene is interesting
for several reasons: one is what it, how it arose. It arose as both first
shown by Evan Eichler who’s actually here in the audience
as a product of a partial duplication of a ubiquitous
gene called ARHGAP11A, which encodes a specific
enzyme-like protein called a Rho GTPase-activating-protein. But it has, in contrast to the
mother gene, it does not have the GAP domain in full,
but it has a novel sequence here at the C-terminal end
of 47 amino acids, which are actually also specific to humans. Now this sequence comes about
because 55 nuclear tides shown here in purple, which are
present in 11A, are actually not present in 11B, and this
leads to a reading frame shift and this new green protein
sequence shown here. And the other reason why
this gene is interesting, is when it arose in
evolution, as both shown by Evan Eichler, but also by
my friend and colleague Svanta Paabo who cannot be
here today, unfortunately. It arose about 5,000,000
years ago, after the lineage that leads to the chimpanzee
segregated from the lineage that leads to us, but before
the lineage that leads to the Neanderthals segregated from
the lineage that leads to us. And as you know Neanderthal brains were at least as big as ours. So we have here a gene that
is expressed in the right stem-cell at the right time
of development, and that is only found in those
hominids with 1.3 liter to 1.4 liter brains. So the obvious question
was with that gene, increased brain size or increased
or increased neuro number in a model system, the mouse. And this is what Marta
investigated, by expressing the gene by a technique that we
call in utero electroporation and to make a long story
short, what she found was that when you express this
human-specific gene in the mouse developing neocortex,
it triples the number of mitotically active basal progenitors, so the progenitors implicated
in neocortex expansion. So it increases these progenitors,
but there are two ways how it can do that: either you make more of these
progenitors from the apical radial glia or, you increase
the proliferation of these basal progenitors once formed. And we solved this, or we
studied this issue and made use of a method that Elena Taverna developed. Elena, like Marta, is also
from Milan, and you can believe me these are such outstanding
people that whenever I get an application from Milan, I accept it. So what Elena developed
was a technique where she microinjects, an organotypic
slices into the apical radial glia. So here you see the slice
and here you see the pipette, and basically she injects
blind, but the pipette is filled with fluorescent Dextran
and so you can identify the injected cell and very accurately
trace what happens to the daughter cells and
granddaughter cells of this cell after one cell cycle or
two cell cycles, and what Elena could show, and I’ll
just summarize this data is the following: she showed
that when you stick in this human-specific gene, you will
change the mode of division of the aRG cell to a style
where it immediately gives rise to two basal progenitors. That is good but it’s
also bad unless these basal progenitors would keep
on dividing, and that is in fact, the second thing
that ARHGAP11B does. Normally in the mouse, one
of these basal progenitors will make two neurons, but
in the presence of 11B, it will now keep on proliferating. So this is what we found in the mouse: Nereo Kalebic very recently expressed this human-specific gene, not in
the mouse, but in the ferret a carnivore with a folded brain. And when he does that, he
found that also in the ferret, 11B massively increases
the basal progenitors, but interestingly, it very
dramatically increases the basal radial glia which is not
what happened in the mouse, but in the ferret it
increases the relevant basal progenitor type. And importantly, these basal
radial glial cells have markers that are very
characteristic of the human state rather than the mouse
state or the ferret state. So the other question we ask
is whether 11B can induce folding of the mouse
neocortex which normally is an unfolded brain. So when we elect operate we
have always a control side which is smooth as shown here,
and then we can stick it in the gene and we see where
it is expressed by the green color, and when we look five
days after sticking this human-specific gene into a
developing mouse brain, we actually see that folds can
arise in about half of the embryos that are expressing this gene. And here is another case where
you have the electoporated area and you see these folds
in the mouse developing cortex which are somewhat
reminiscent of the folds that happen in a fetal human brain shown here. Question is how does 11B
achieve these effects? Is it like the gene from
which is arose, 11A, a RhoGAP question mark? And so a postdoc fantastic
postdoc from Japan, Takashi Namba, took on that question, made various constructs, the mother gene 11A, a full
GAP domain of 11A, a truncated GAP domain of 11B, and a modern 11B form, and asked which of these
exhibits RhoGAP activity, and the answer is these
upper two ones do, and these bottom ones including 11B, do not. And we actually know that
the ability of 11B to amplify basal progenitors, is tightly
linked to the existence of this new, green, human-specific sequence. Question then was, can the
evolutionary increase in brain size in us be
explained by ARHGAP11B? And there we have a
problem, because as I said, Evan Eichler and Svanta
showed that this gene arose about 5,000,000 years ago,
but 5,000,000 years ago our brains were small. The big increase in brain
size happens later, but we now have at least an
interesting possibility of explaining this enigma. And this has to do with
the question, how this loss of the 55 nuclear tides that
gives rise to the reading frame shift and the new
sequence, actually comes about. When we started this work,
we thought that as the gene duplicated, these 55 nucear
tides got lost at the level of the genomic DNA, but when we looked more closely
we actually noticed that the genomic DNA of 11B contains
these 55 nuclear tides. What’s happening is something
much more interesting. What you see here is in the
top, the sequence of 11A, and this is the spliced donor
side, the side where one coding sequence is moved
and so spliced onto the next one. Now in 11B, there is a
point mutation; a C to a G, which creates a new spliced donor side, and when this is taken, the
55 nuclear tides here are removed and that induces the
shift in the reading frame because it truncates the coding
region of EXON 5 and then when that’s spliced together
to a messenger RNA then you get a new protein sequence. So this human-specific
sequence is caused by a single nuclear tide splice mutation. This raised a very interesting
explanation, or at least the possibility, and that
is that ARHGAP11B arose as a with a full GAP domain
and having this activity 5,000,000 years ago, and
then later, but before the Neanderthals branched off from
our lineage, later as point mutation takes place, which
gives this protein a new function or this gene a new function. So Marta investigated this
possibility by creating what we call an ancestral ARHGAP11B. So it is a gene that is
essentially or a CDNA that is identical to modern 11B, except
that it doesn’t have the G, but the C, so it cannot
make the green sequence but makes this kind of a protein. And we then asked the question,
will this ancestral 11B exhibit RhoGAP activity? And so Takashi again made
various constructs, and tested their possible activity in
an essay in which a decrease in phosphorylation is
indicative of RhoGAP activity. And as you can see, the
modern 11B does not exhibit RhoGAP activity, but the
ancestral 11B, like the 11A, exhibits RhoGAP activity. Next question then was, would
this ARHGAP11B increase basal progenitors, and again by in
utero elect operation into mice in contrast to the modern
11B which increases basal progenitors, the ancestral
11B is unable to do so. So we have a rather interesting
situation as follows: we have a single C to G base substitution which creates a new
splice donor site in 11B. That leads to the loss of
the 55 nuclear tides as the messenger RNA is made, that causes a reading-frame
shift, that leads to a human-specific c-terminal
sequence that is key for basal progenitor
amplification, we believe. So if ARHGAP11B should have
contributed to the increase in the evolutionary increase
in brain size in humans and Neanderthals, it
would have done so by a single point mutation. And with that I would just
like to acknowledge my collaborators, Marta, who
is now a postdoc in Boston, with significant help from
Mareike Albert, Takashi Namba, and Elena Taverna and Nereo
Kalebic who did the ferret work that I briefly mentioned. This is Dresden at night,
and these are people in my lab I didn’t have time to mention. These are people who have left the lab. This is the famous Semperoper,
I’m not all the time yet I notice. And we have the good
fortune of collaborating with Svante Paabo, very
beautiful collaboration with him. The fetal tissue is given
to us by Robert Lachmann and Pauline Wimberger, and we
have a lot of bi-informatic support from Michael Hiller,
and these are present collaborators outside
of Dresden in Leipzig. And this is the reconstructed
church in Dresden, and I don’t mention this
people’s work on prominent but I would like to
acknowledge our support by the Max Planck Society, the
German Research Foundation and the European Research
Council, and with that I thank you for your attention
and I’m curious which questions you will write down. (applause)

5 Comments

Add a Comment

Your email address will not be published. Required fields are marked *