Thanks below for inviting me.

Thanks yeah despite the fact that I grew
up in the Midwest I do not like this

weather at all.

So this is our Parker building
just to show you that we get some

cash from that guy too so today I'm going
to try to keep things fairly general so

I know this is a very diverse audience and
so one of the main things that we're

interested in my lab is how the brain
can repair itself after damage and

of course that's related to the mechanisms
that it uses during development as well so

we study normal development as well as
plasticity and responses that the brain

has to injuries or or other insults and

of course as you probably realize we're
all quite good at damaging our heads.

So we've heard a lot of news about C.T.E.

in football players lately and by the way
that happens in soccer players too.

Soldiers little kids are always
trying to kill themselves and

then we do these crazy sports like that or
that and then even if you

don't do any of those things if you would
drive in Atlanta you're doing this.

So lots of ways we can injure our heads
the difficulty being that when you go into

the hospital with a head injury basically
what they do is they stop the bleeding and

then tell you well we'll just have
to wait and see how things turn out

don't you think that's a little primitive
in this era of modern that it's and so

we ought to be able to have
a better response to head injuries

than waiting for the brain to do most
of the work in healing itself so

we're trying to understand
how the brain does that so

that we could harness those mechanisms and
improve therapeutic approaches.

So how do we choose a model system for
traumatic brain injury has

to be a system that we can
manipulate fairly easily and

in my laboratory we actually use
three different model systems

we use the hamster when we're interested
in looking at the retina Colicchio or

projection so from the retina to
the visual mid-brain we use ferrets

when we want to do cortical physiology
because trying to do cortical physiology

in rodents is a bit of a challenge so
we like to avoid that and

then if we're forced into using
knockouts and transgenic then we.

Grudgingly use my eyes which are a lot
harder to do physiology and.

So we use these three models and

today I'm going to talk entirely
about the hamster and the mice.

Results So
how can we manipulate sensory pathways or

alter their development in a way or
injure them so that we can study these

mechanisms of repair Well we
could just put a lot of extra and

put and see if the brain knows how
to accommodate that extra And but

that's a question that's
more related to Lucian Arey.

Hypotheses about the brain but
we do sort of increase the input and

one of the approaches that
I'll show you today and

I'll explain that a little
better in a moment.

We also can decrease the amount of
input so a large proportion of the lab

right now is studying the effects
of sensory deprivation.

Or we could rewire the inputs so

we can rearrange inputs in relation to
their targets and that's work that.

I would like to talk about except I
haven't got anything new on that for

the last four or five years I haven't yet

found a student who wants to do
the thirty six hour recording sessions.

The volunteers.

The student that work is in Philadelphia
now doing her post-doc you to.

She was a beast.

So you're probably familiar with
a slide that looks something like this

showing you how sensory information
can be arranged topographically.

This is the most common figure

that you might see in a medical neuro
science textbook showing you how the.

Body's surface your skin surface and

the receptors in your skin are represented
along your son out of sensory cortex so

there is a two dimensional representation
so that the sensory information in your

brain there is also a two dimensional
representation of visual information and

that's what we're going to
be concentrating on today.

In particular the model that
I'm going to talk about was.

Coming out of work really
done by Roger Sperry

you may remember him as the split brain
patient guy he won a Nobel Prize for

that work but he did this work as
well in the one nine hundred sixty S.

where he turned the eye
of a frog upside down

cut the optic nerve and let the optic
nerve regenerate because frogs

can do stuff like that we can't we'd
like to know why not but at any rate

what Sperry noted was that the optic
nerve even though the eye was upside down

it grew back to the same location in
the mid brand where it used to be so

he used the idea of
chemical attraction and

pre specified information
to to develop this chemo

affinity hypothesis and so as a result
this was a behavioral experiment so

he didn't know it was going on.

In the brain but the frog was unable
to catch its dinner because if there

was a fly here it would always look for it
down there because the eye was upside down

but the connections hadn't changed you
know what Sperry didn't realize as perhaps

perhaps he has one of those type
A guys you know Nobel Prize winner and

he didn't have the patience to wait to see
what the long term effects of that was so

he had no idea that the frogs would
eventually correct that situation

because they want to eat so
the brain has a pretty strong motivation

to correct its conic to be based on
the information that it's getting and

its world and
it took John Schmidt who's a professor

at SUNY Albany who started
doing this work and

similar manipulation such as the ones
I've shown here to show that the system

can regulate itself so I think I probably
can't I don't know maybe I can so.

We see here the retina and we have put
supposed to be a green a red and a black.

Sun coming here.

They're making their topographic
projection into the optic tectum.

Of for example a goldfish or
a frog similar idea so

they each have a particular domain and

then this region in
the tectum would be paying

attention to the region of visual space
that this part of the retina is looking at

what John Schmidt showed was that
if you ablate half of the tectum

and this is in an adult
because it's a frog or

fish you can do these kinds
of things with those guys.

You notice now that this green one
doesn't have its normal target.

The red ones kind of dancing on the edge
of not having any place to go but

what happens.

Is that we get a compression of
the projection a compression

of the topographic map so

nice that that happens in
the so-called lower vertebrates.

But in fact it also and
this was the coverage of my Ph D.

advisor Barbara Finley It also occurs
in mammals if we do it early enough.

And so

we use neonatal hamsters because as you
can see they're streamlet immaturity birth

although apparently they can talk this
guy's really got a lot to say here but

they're very immature It only takes about
fifteen and a half days to make a hamster.

So that's a lot faster than most
other animals also they're mammals so

the idea is that we can get access
to very early developmental stages.

Compared to what we could do
in other animals like cats and

primates which were the model systems of
the day and looking at the visual system.

And as a result then we can get
access to the visual projections

from the eye before they have made all
their connections within the brain.

So maybe in this developing system.

We could get this compression to occur and
in fact it does.

So just to give you a little
orientation here I'll try to.

Be equal opportunity
with the screens here.

So here is a general
projection from the I into.

The visual ballast lateral nucleus or

the superior to us which
is the visual mid-brain.

L.G.M. projects into visual cortex
visual cortex projects back disappear

you're collecting lists so that's just
the general anatomical picture here.

And if we look at the doors of ventral.

Nasal temporal axis of the retina those
are really represented within the tectum

to produce that topographic map.

So if we were to do on
electrophysiology experiment which is

the way we have approached this
experiment I'll tell you about we can and

that's the tightest hamster or mouse or
any other animal we should so choose but

those are the ones where using we can put
an electrode in the superior Kyllikki

lists by the way superior optic
tectum those are the same thing but

it's called superior colicky listen
mammals and not fishes frogs reptiles for

some reason that I don't understand that's
the same thing so if we were to move our

electrode along to different locations in
the Tech them we would find that we would

get a progressively different location
of the visual field that it was

representing there now
if we remove half of.

This Earth in these really mature animals.

We find the same thing the map

has now scrunched itself down
on to the smaller target.

Now originally this was supposed to
produce a situation this was the plan that

Barbara and I had that this was going to
produce a situation where we have the same

amount of information in the retina but
it's converging onto a smaller space so

any individual cell within that target
seems perfectly logical to conclude that

each cell is going to have to
represent more visual space.

And we were really interested in doing
that because there was a huge debate in

that day about whether receptive
field properties such as orientation

tuning accept or access to were built
from specified parallel pathways or

from convergence of different inputs so
that was the big debate and

we were going to answer it.

And I was very upset to find in fact that.

Even though we're representing twice
as much input on half as much target.

Individual neurons in the superior

are still looking at the same
amount of visual space.

So they have repaired themselves to keep
the size of receptive fields the same.

And of course I was really upset not
realizing that this was even cooler than

what we had planned to have
happen in the first place so

graduate students if you find that you
get the opposite of what you expected

Do not despair it can
lead to an entire career.

So if it had not.

Made that correction right the cost of
increasing receptive field sizes and

a topographically amount projection
would be to alter acuity so

that it would look more like that so
you would have a much fuzzier view of

the world than you might think well for
a hamster what difference does it make but

it could in fact make a difference and
we're in the process now

of doing some behavioral experiments
to check their perceptual ability and

it looks like they really are and
paired in their escape behavior so

they don't escape as well from a looming
stimulus coming from above which for

a hamster you know that's likely to
be a raptor predator and they need to

they need to be able to escape quickly
because those guys move pretty quickly too

OK So that led us then to ask this
question how do they do that right how

do you get a preservation of
receptive field size even though

you've got this increase in
convergence of the two populations and

so we propose that perhaps the and
I'm da receptor

which has some special properties that
I'll tell you about might be able to tell

the the cells in this the period colicky
lists how much visual information that

is how much space they were
representing and then be able to

prune down the projections enough to keep
the receptive field properties the same So

our hypothesis was that any day receptor
activity is necessary to control

the convergence at that level of the
single neuron so it's very important that

we make single unit recordings rather than
multi unit or field potential recordings.

So we blocked M.D.A. receptors and

compared results in normal animals and
animals with the come pression

manipulation which I will call see
what I decide to call that today.

When you know how you have lab lingo and
you have a little abbreviations for

manipulations you do and
you go talk to an audience and

have no idea what you're talking about so
we could call it lesions or

we could call it a reduced size superior
call it let's let's let's go with that.

And so that we're going to ask say
the receptive field sizes as we did before

just to give you a little bit of
background because it's such a broad

audience we can generally go when
we're looking at activities role in

shaping conic to beauty we can
generally go with the maxim that

if you use a projection
it's going to get stronger

if you don't use a projection from and
put to target it's going to get weaker and

you might lose it and so
we can build up our brain power.

With practice Similarly
the way you use your muscles

in the gym to build up are we could
have the unfortunate opposite result.

And this really happens apparently
because it happened to Homer.

So here's a little diagram to
show you what I mean here so

here's our retina and we have a visual
stimulus that's going to move

across the retina that is going to
cause firing in a group of neurons

that represent that location and different
neurons represent the next location and so

forth so we'll have a wave of activity
traveling across the retina as a stimulus

moves from one side to the other early on
development you see that any individual

target cell can get input
from many different acts on.

And then later on in
development it refines

such that it has a smaller
receptive field.

And the reason that happens is because
the target is looking for correlations and

activity so these two are firing together
in time so that leaves us two or three.

Fire together wire together that
you may have heard in your class.

And so
since these two are coincidentally active

they're going to consolidate
their connections on to this.

And likewise it would lose the connection
with that one because it's not

simultaneous but these two
are similarly active and so forth.

So the M.D.A. receptor work and
this may be a review for you but

end of day receptors work in
a number of special fashions so

there glutamate receptor is you don't
have an M D A in your body because that's

a drug that's just the thing that we
use to recognise and day receptors but

it is a type of glutamate receptor like
the Amber receptor which passes sodium but

an M.B.A.
receptors pass sodium and calcium and

calcium feeds into SECOND MESSENGER
processes that can influence the stability

and the strength of that's enough.

The other special thing about an M.B.A.
receptor is is that they are not

able to open their poor at their resting
potential because there's the magnesium

I am that's stuck in there stuck in
the poor and so we can't get sodium and.

Calcium into that channel
rusting potential.

However if we can excite this

post in after terminal some
other way to deep polarized the.

Neuron a little bit now we have some of
these polarized nation the magnesium ion

is spelled now we can pass
calcium sodium through the M.D.A.

receptor the calcium leads us
the second messenger pathways that can

influence the stability of the soon
to make a long story short but

the idea is that this is a coincidence
detector because you have to

first want the polarisation and
then shortly thereafter another one.

So just like classical conditioning
when you need to associate

an unconditioned stimulus on
a condition stimulus this one.

Receptor allows the post an uptick
neuron to distinguish the level

of Cofidis of different inputs that
are competing for its attention.

OK So again we're going to
manipulate those and M D A.

And test the hypothesis that
they're necessary for refining and

receptive fields after that manipulation
most of what I'm going to show

today is not original data because I
want to give you a lot of information so

it's going to be lotsa graphs and stuff so
I'll try to not make it too boring but

so here's the process so
we're going to a large receptive field so

we're going to refine them down in the
separate collect as the hamster or a mouse

the receptive fields of individual neurons
are fairly large much much larger than

they would be in visual cortex
about ten degrees in diameter and

as I told you when we do this manipulation
of partially lesion ing the tact.

Lab lingo forward is P.T.

for partial tectum then the receptive
field size is maintained.

If we block the M.D.A. receptor during
development with a slow release polymer

that's deadly leaking this
antagonist of an M D A receptors.

Then we find that that blocks the normal
refinements of the receptive field so

our receptive fields are bigger
by about fifty percent.

If we come by in that with the lesion so
we lesion them at birth and we block

an N.D.A. receptor as we find of course
that it blocks the normal refinement but

it also blocks the capacity of
the system to compensate for

the loss of target to shoot when they
were set to fields get even larger

because of that increase convergence
from the retina in the reduced.

And then you know you always do
the control study last right because those

are the boring studies so this is
the inactive isomer of the drug and

thank goodness it didn't show anything.

So those Then important control with.

P.V. The Levo form of
the drug isn't active.

So a little ball and stick diagram
here to illustrate what we've done

this is my idea of a computational
nerve science model.

So early in development we've got
a projection from disappear collect

this right if we record from this
is where I need a fishing pole and

not a pointer if we were cord from the
second neuron and super going to see it

gets three different inputs right so it's
receptive field diameter is going to be

three different diameters of retinal
neurons normally during development

that would refine the thing we end up
with two inputs instead of three so

that will be the receptive field size
of the individual neurons sharper and

it acuity what we found in the in the case
where we reduced the tectum by about half.

So we went from a four cell
to cell Tekton in my diagram

we get overlap as we would initially.

But the individual axons reduce
the size of their Arbor.

And selectively project
alternative targets so

that they have a chance
to find target space so

it's kind of like a game of musical chairs
with axons if they don't find it's here

they die it's a more severe penalty
then at the birthday parties.

And you know as this has to happen a very
specific way it still has to select for

two adjacent inputs it can take
that one and that one about one and

that one because then the receptive
field size will be larger so

there is some way to filter that through
and we propose that it was due to that and

I'm David Souter so we block the M.D.A.
receptors in a normal animal full size.

We maintain that early state forever.

If we combine that with
a partial ablation of the tact

we went from a four cell tectum
to a to sell tectum in this.

Pathetic model here then you see
that we get an even further increase

in receptive field size because we block
the normally Fineman as well as blocking

the compensation for the damage.

So our results then support
the hypothesis that and

it receptor activity is necessary for not
only the normal refining process which was

already known a number of other systems
but also for compensating for that lesion.

So receptive field size is one aspect

of the properties of the target cells and
superior colicky list but those

neurons also have other special properties
that allow the animal to conduct its life.

For a frog of course that includes
catching moving objects that was what

Sperry's frogs had trouble with so they
need to know something about the objects

that they're looking at they need to have
object recognition ability and so for

a frog that can often just be something
as simple as knowing how big and

how quickly the target is moving to
identify it as a particular kind of

of prey item that might be tasty and
the same is true of

rodents who will often
identify moving objects so

even though we have maintained
a receptive field size maybe we can go

back to our original question and address
whether convergence from the retina

is playing a role in the size and
velocity tuning of these neurons.

So remember that this is our manipulation.

And we're going to ask Was that an M.B.A.
receptor blockade combined with

the partial lesion of the tech to affect
stimulus size and velocity tuning we

thought certainly it would because we were
thinking that target cells it's not just

one target cell that defines direction and
velocity because the stimulus is moving.

And so you have to compare from one cell
to the next and from one cell to the next

there's going to be more information from
the retina represented so we thought

certainly this is going to change velocity
tuning and it would likely change size

tuning too because of the difference in
representation of adjacent target cells.

How.
Ever to make that story quite short it had

absolutely no effect so not only.

Was stimulus.

Was receptive field size
maintained after the lesion.

Although not after the end of the a
receptor blockade but stimulus size and

stimulus philosophy tuning was
resistant not only to the lesion but

it was also resistant to the M.T.A.
receptor block

so that anything we could
throw at this system it just

fixed itself everything just stays
the same so keeps its acuity it

keeps its size tuning it keeps its
philosophy tuning we were getting a little

frustrated that there's just nothing
we can do to purge this system it is so

robust to any kind of manipulation and

on the one hand that's very interesting
but it wasn't what we were trying to do.

So we never did answer that
convergence question we got into this

idea of how inhibitory.

Neurons can often be
recruited to compensate for

changes in excited Torrie convergence so
we've got all this extra information

coming in from the right to a given area

of the target these this is a little bit
hard to explain so bear with me about if

if I were to pass a light stimulus
through the center of a receptive field.

Here is our center right here of
a receptive field I get a really strong

response if I move that stimulus
a little to the side I would eventually

not get a response anymore because I
would be outside of the receptive field

well we wouldn't want to probe
inhibitory connections and

because we're recording extracellular
Lee We don't know anything about

inhibition except how it affects
excitation So if we have excitation and

inhibition going on simultaneously we
should see a decrease in excitation

from our electrode that's
outside of that neuron.

So what we do is we take the stimulus
that's in the center and

we add another stimulus.

And assaye what's called
surround inhibition so

superior Colicchio cells unlike cells in.

The valley miss like a really
small receptive fields and

they have a lot of inhibition throughout
their receptive field so we as.

Another stimulus then there it's going to
inhibit the response to the first one and

as we move that second.

Stimulus away from the center of
the receptive field the influence of

the inhibition on this response is going
to decrease and in fact in the normal

animals the black lines here if
falls off pretty rapidly but then it

plateaus out what we found in the animals
that had been treated with the M.D.A.

receptor blockade is that
the boundaries of inhibition were

extended perhaps to correct for
that excess of excitation and

we're proposing and I won't go into it
today but we have built some models.

When I had an engineering
post stuck in the lab.

Over the OP who is now
a tenured professor at U.C.

Riverside made some models of how
this increased inhibition could

fix the signal the size of a lawsuit
tuning and it worked out really nicely so

we're thinking that this change and have
a Tory connectivity is a general means

through which plasticity can occur
that was previously unappreciated and

most people think about neurons being
excited to worry toward each other rather

than and have a Tory about inhibition
shapes responses and it seems to be very

important later on in postnatal
development when excited Tory connections

have already been established there seems
to be some indication from our results and

others that the inhibitory plasticity is
a little more flexible later in life.

And so our results were that the
inhibition is changing its configuration

and that may compensate for
the increased receptive field size and

be maintaining size and velocity tuning.

So the system can be very flexible in

accommodating changes to its inputs and
outputs and

that's one way that it preserves function
in addition to changing the arbor size.

OK but that didn't tell us you know
it told us how resistant the system

was to manipulation but
it didn't tell us how the.

Compensation process at the very
beginning the compensation for

the lesion was actually happening.

And in the meantime after we
did this earlier work and

I started my own and
we started getting more into the system so

I did a post-doc completely
out of this area MIT

talking about that work today
the Fair Work started there.

But we wanted to address the the molecular
basis of this situation so

we know that there are two basic steps in

refining the projection from
the retina and to its target structures

one of those steps is is directed
by molecular gradients so

the initial formation the gross
topography of the projection is set up

based on molecular cues so that when we
started this work we didn't know what they

were but in the mid ninety's they
were identified as repulsive

acts on guidance factors called
Efron's and their receptor is the F.

receptor.

Which doesn't start with an F.

it's ph.

And so
we were reasoning that if you need this

gradient of repulsion to set
up the a national projection

may be it's the case that when we injure.

The guidance molecules change
their configuration and

there are what tells the map to
compress on that smaller fragment

and so we wanted to combine.

The ablation of the partial ablation
with the map compression and

asked say in a descriptive study for

what happens to those Efron's and
their receptor.

This work was done by a couple
post-docs in my lab may do and

Chang and by a very talented
student who was a masters student

at the time pretty much directed
this project as a talk to.

Now let me give you some
introduction here so

if we have our Efren a to simplify
the story they're going to be a lot

again today although they can actually act
as both which is really irritating but

there's just amazing molecules very
well studied in cancer research so

there are guiding invasion of for
example blood vessels into tumors so

the eye for an A is are located in
the target there from a lot and here are.

Two high post here your grading gradient.

There receptor those which are located
in the retina and on the X.

on terminals are in the opposite
configuration because it's a repulsive

interaction so the high a receptor
expressing region and temporal retina.

Doesn't like the back of the tact because
it's very inhibited area can detect how

inhibitory that is because the receptor
level is high so it prefers to be here

where there is not so much about why
again whereas where there receptor is low

those are tolerant of
the higher law levels and

then intervening steps you can imagine
to set up the entire projection.

So that was what we knew going in.

We did express in studies and
hamsters they'd already been done in

mice because we wanted to do hamster
physiology and most physiology it

was easier we found the same thing in
the hamsters that we did in the mice and

this is a lateral view of the brain so
this is cortex here that's

up down nose tail cerebellum
there you may recognize

this is the tectum from there to
there that's in fair click Elice.

Early in development you see that these
yellow areas are showing the M.R.I.

and they expression pattern of the friend
one of the friends called a five

it's very high and coddle it's a period
color Phillis later in development

a it starts to decline but
still higher caudal and then by

two three weeks of age it's kind of going
away because it's not really needed there

anymore so perhaps it's not produced for
some reason it's downright delayed.

That's an interesting question and
itself but we don't know the answer.

So to set up the hypothesis
then what we're testing.

With this descriptive study
we're going to look at

how the expression pattern of
the Efron's in the target changes

as a result of removing part
of the target structure and

there's two possibilities right if we just
cut off the gradient right there then

we would get this situation we go from
yellow to pink but not all the way to Red.

If the gradient compresses which
was our hypothesis because we're

thinking the compression of this gradient
would cause the projection from the retina

to compress on the smaller fragment then
we would see the slope of the gradient

gets steeper so it go all the way from
yellow to read but in a smaller distance.

This is kind of an ugly figure but
that's what we found so

here's a normal superior colicky
this you see the high F.

or label there this one is I had
a different mood that day so

it was blue and
green instead of red and yellow.

And here we have our smaller tact.

Superior this was leavin on
the day of birth these are now.

Five animals I believe and
we see that the high F.

or an expression is still there but that
peak has moved forward still at the back

of this a period but the spirit Coco says
smaller in these damaged tectum animals

and you maybe see from the arrows here
the front and the back of the tectum So

it's roughly trying to make roughly fifty
percent lesions you can actually remove

about seventy percent of the Superior to
look listen you wouldn't know it from

the way the animals behave and in fact the
system compensates with receptive field

size up until about that point you go
beyond that and it starts to lose it but

then it loses it in a really intelligent
way it starts to represent only things

here and sacrifices what's there
you remember that these are your

eyes have their eyes on the side of their
head so their visual field is like this.

So gradients are steeper just to give
you an idea again about the hypothesis.

So if this is our normal map
see looking at the levels of F.

or in a five and this is just a cartoon
and it's not meant to be real but

opposed to us see locations of Pyrrhic
a little suffocation from the front

to the back is about two
millimeters in length.

In the normal animal So

if this is our slope of the gradient
if we had just cut off the gradient and

reduced the size of the tectum by half
from two millimeters to one million the.

And nothing change we
would expect the gradient

to look like that it would
never get up to that high level

if I have process this was correct then we
would expect the gradient to get steeper.

So it would go all the way
up to the same level but

in one millimeter instead of two.

And that is in fact what we found so
here are a plot of the slopes

across development normal animals and
black damaged and

white and this is the slope so
the change and

are an expression over the map
location we get deeper each age.

In the leaves and the animals so
that supports the hypothesis

that the changing
configuration of the molecular

signals can direct the compression.

On to the smaller target but
it's only a correlation.

So this is where I got dragged kicking and
screaming into using mites and

when I say I I mean the Royal I.

So my very talented team.

Was able to develop the mouse recording
we had to make a number of modifications

because they're a little bit smaller.

At all stages of life but
it wasn't too bad I'm.

Told you about so just to remind you
of the results of the Corella to

the experiment was that the gradient goes

from yellow to RAD but in the smaller
distance and we also found I'm not going

to show you the data because then I want
to keep it here forever but the gradient

of receptor seems to
compress in response so

this happens a day later than
the compression of the law again so

we see the receptor gradient changing and
I say about a day later we only

looked at twenty four hour period so
for all I know it only takes an hour

that we need to run that down that's
one of the things we're doing now.

OK so now we want to know.

Whether those friends are causal
from compression and so

we have to knock them out so
if change after an expression

is necessary for
a map compression then if we knock

out the friends it should be impossible
for the gradient to compress.

So it worked in knockout animals
that's why there's little mouse at

the top there now.

And we did the same
electrophysiology experiment.

And the normal animals of course
they had the data already we took

although not in mice so

we had to repeat the entire thing in my
normal life we then took the normal my.

Ablated the back half of the tectum
same as we had done in the stairs and

indeed this was pretty fortunate
they also compress their mouth so

it's not just a hamster thing and that's
important because mice take longer to

develop so even though we're looking at
an animal that takes more than fifteen and

a half days to hatch we still can't
get that compression so it's still

early enough in the in the development
of that optic nerve growth to do that.

And so we're hypothesising is that
in the effort in a knockout animals

that instead of getting this compression
they're going to have any effort and

so that's why this is a yellow or
pink anymore

we're not going to get the compression and
we'll still only map half of the tech.

So here we can go a millimeter
with our recording electrode and

see the entire projection onto
that target we're proposing

that in the knockout animals we're just
basically whacking off the map and

half we're only representing
half the visual field.

So here's where we get and
this really complicated

pictures that I've never presented to
an audience before so see how this goes.

We are working with rather incredibly
famous competition on earth

scientists by the name of
David Wilshaw in Scotland

we have all these computational people
in town including many of my own and

Satish and
nobody's interested in sensory systems.

No I don't mind going to
Scotland it's a lovely place.

But the thing about David as he has
been fascinated with this question of

all of his life so
I called him up like you don't know me but

I think you could really help
me with this he's like yes.

So we actually met at the European
our science meeting in Copenhagen and

he developed a way for us to
statistically analyze to visualize the.

Form of the map so
let me orient you here so

these are just normal
animals wild type mice.

I think this on you can see pretty well so

as we move our electrode across
the tectum from front to back and

from medial to lateral we get nice even
progressions a visual field location so

we're see how in this map
of the visual world the.

Tactile locations are very ordered right
there's a very neat a graphic projection.

We can also represent.

And we know that the same thing happens
in the animals with ablations are we

even though the Tech them is half
as big we compress the map so

we still get entire visual field
represented on the smaller space so

these graphs just show our recording
locations so we made three.

Series of penetrations
along the poster axis.

Of the severe.

And these maps are how David represented
actual visual field location and

you'll see that these maps are mostly
blue with occasional little red lines

the red lines mean that there are some
topographic messiness there so

in the normal maps they look pretty good
also the time they have very little red in

there it's a very you know it's not linear
but it's biology why would it be linear so

it's a life representation of the visual
orbital expanded in the front and the.

Reduced.

As I said we know that
the map compresses so

we would expect it to look roughly
the same except that our electrode

positions are going to end where
the lesions started so this one was.

Not quite half one in
a little bit millimeter and

we try to shoot for one but
it's difficult to get it exactly right but

again you see that these
maps look quite normal.

If anything even less disorder and
that map which is interesting.

Now we had a number of
options to pick from.

In terms of the mice that we could
use how much time do I have.

OK.

That's good so we initially got
these super duper everything

knocked out from David Feld Haim at U.C.
Santa Cruz and just you know for

the students edification I was asking for
these knockout mice that David

felt time developed and at the time
his house was being bombed by animal

rights activists and they had to throw
their child out the window to escape.

And he was injured his wife and
child were not but

that slowed down the mouse
shipment a little bit.

So eventually they got these mice and
they're shipped.

There several different Efren A's so

they're shipped is eight to a three
knocked out a three as far as

we know isn't involved in the two is
a little bit involved in this poverty

a five is the major one and
they're heterozygous for a five.

And so if we cross them we can get wild
type a five heterozygous a five or

a triple knockout.

And.

The wild type a five then would be double
knock out the other eight to a three

knockout but
a five is still supposedly OK.

And you'll see that the results
look quite different.

So it looks like we're not able
to project very far with our

recording electrode even though
we're going the same distance.

We're not able to represent all the visual
field locations now you're familiar with

these drawings you see that on there
you see that they're pretty messed up.

Even though a five is
wild type in these guys.

And it looks like they're certainly
having trouble mapping the back half

of the tectum or even the front half in
some cases there's a lot of variability so

we don't quite know what to make about.

And these are just the normal guys.

Right so what about the reduced
tectum monthly C M particular

that they're not matching for

the most part they're there they get
a little bit of representation and.

But mostly they're only
making half amount so

we expected them to be massed up into
poverty that was already known because

these animals were used for topographic
map ing that's why they were created but

no one of them had overstudied the role
of friends in the come Prussian

topographic map city so these double
knock outs are kind of interesting So

then we got another strain where
it was just a five knocked out.

Hoping to get a more consistent
picture and we got these from.

Birth who wasn't firebombed and
so we got them right away.

Here's the normal I mean not lesion
animals with no for an A five to to PA

graphy can be a disaster or it can be
fairly normal interesting variability.

If we reduce the size of the tack them so
we leave it

in the back half of the tech and the birth
these are recordings that see if they will

get to go as far because there's not
as much tech done before run off and.

With these guys we occasionally get
mapping the back of the tectum and

sometimes we don't but
the topography is quite variable and

Hamas stuff it is so
this is why I didn't want to work on my.

Now the triple knockouts.

Have to be equal opportunity so
the triple knock outs as we expected.

They're topographic maps
are a disaster I mean you

can put your electrode at the front
of the tectum you get a field here

you move it just a tiny bit two hundred
microns in your back there are already

moving a little bit more now you're
here right so it's just really crazy so

if you look at David Plotz here you see
how the topography is a complete disaster.

And we're really on able to even know what
either we get a compression of the map or

not because there's no math so
those animals weren't especially useful.

And so we were unable to calculate them
out slope with the triple knockouts

because they were just too crazy there
really was no map and we're thinking well

the axons because they don't know where
they are they have no idea that there is

some part of the tucked in that's missing
so they just go wherever and the entire.

Retina might go to the target but
there's no way for

it to distinguish which part of the target
it should end up in because it doesn't

have the signals that tell it to get there
what David is most interested is that you

can get little bits of orderliness in
these projections and he's wondering is

that Are those the axons that are special
maybe there's a special kind because of

course there's lots of different accents
or maybe there's different being than

cells that are more successful in
surviving this sort of manipulation but so

with the absence of the T.K.O.
where we could not map the slope.

Please see the.

Slope of the map So
here's our regular guys right so

the slope of the map is about here.

And the wild type animals that have
the ablation the slope has to get steeper

to get all the retina represented there
but if we do the statistics we see

that in every one of the knockouts
excepting the T.K.O.

where we can't calculate the slope of
the match because there is no match.

They are having a problem some
more than others with compressing.

And increasing the slope
of the projection and

I haven't done the statistics yet for the.

Meeting so
I couldn't quite get it done for

you guys but it's going to be
interesting so if we actually look at.

The slopes.

Here's our normal animals
with all of the data and

there is our lesion animals wild so
much steeper

there they are here and the different.

Types are mapping somewhere in the middle
there we thought well that could

be due to the fact that we didn't leave
them all the same but the groups don't

distinguishably differently in size so

they're all roughly half OK So
that supports the causal

connection then that we do interfere
with compression of the map if we don't

have the friends it's definitely not clean
data so we're going to have fun trying to

sort out exactly what's going on in that
suggests a lot of further experiments but

just to summarize then I've
told you how activity dependent

mechanisms through an M D A receptor is
can compensate and preserve stimulus.

Tuning so response properties.

The compression requires
at least effort and

maybe some other factor that help
gets some normal topography but

if we knock out those genes then
we don't get the compression so

we have to find the number of
different factors that allow us to

understand what makin those are used by
this very simple projection to preserve

function in the face of rather challenging
damages removing half of the structure.

So just acknowledge all the different
people that were involved

then I think I can reach that high but
we have our

many collaborators here I didn't get to
talk about it but Josh since it's just.

Transgenic mice that have different
populations of ganglion cells

labelled with G.F.P. so we can start
to look at differences between types

of gangland cells and
how well they survive this manipulation.

To the correlation study you
telling me was the bees to

not only to ferret recordings but
pioneer the map recordings as well to

Vollmer was working with her he's
now a post-doc and Larry trusts.

You David not as my current student and
a number of different postdocs and my

very first student not Dylan across go who
was not involved in these projects but.

We now have funding from.

And us stuff we did have funding from and
I.

And a number of other organizations the
brains and behavior program and the Center

for behavioral neuroscience Georgia State
provides money for us to do this work when

we can't get any funding but now we
have some funding so we get to proceed.

And with that I'll finish and
take any questions that you have

THANK YOU THANK YOU.

Thank you.

So there has been one group that
looked at their production from cortex

back to attack them in these knockouts but
not in a lesion scenario.

We didn't record from the road.

So we don't know but
that's a really interesting question and

it's a little challenging to record from
the written amount you can routinely do

it in frogs for example but
difficult in a mammal sorry.

We don't know exactly it
seems to be expressed

not only in the tech the target
cells themselves but

in the extracellular matrix and that may
be because it's created but there hasn't

been a lot of biology done on that in
the visual system there's been a lot of

work done on it cancer systems but I don't
there imagine that it would be the same.

Yes.

They were all in us the ties with love to
find a student who wants to do the work to

develop the way behaving properly.

So the way that we're kind
of lucky compared to.

People who are working in monkeys or
carnivores and that in the rodents we can

use an anesthetic called your thing which
we think of as a plastic wrap right but

it's also an anesthetic and it suppresses
all the different neurotransmitter systems

fairly evenly so we don't have the same
problems like we would with ketamine

selecting selectively depressing and
I'm da receptors but

still it would be much better to
do this and in a way preparation.

Thanks for that question others.

I I have no idea because we have
not recorded from the inhibitory

neurons directly would have to do that and
slices.

There's been plenty of recording from the
very large inhibitory neurons in serval

cortex as I'm sure you know but
in Superior I don't know that there's

been very many people that in slices
have recorded from the inhibitory And

that's something that I really would like
to do because I'd like to see me even

just to look at them morphologically
how they're changing in the scenario

they should be changing
morphologically and

that would be something that's fairly
easy to address we just haven't done it.

Thanks that's a good question.

Yeah.

So.

Well could depends what you compare
it to yeah they're much bigger

than visual cortex but they're small
enough that we call them bug detectors.

Well they won and
they get to that base mall or so

they go into the dorsal
stream they would get larger.

So I mean.

This.

So we don't know

because we haven't recorded there but
judging just what happens in the tact

you can break the system by removing
about three quarters of the target

not before so presumably all that
compressed information is being

piped up through the dorsal stream and
along the ventral stream to operate in

stimulus localization then an object
recognition such as it is in a hamster.