Slide 3
Welcome to Part Three of our lecture series on synaptic plasticity and learning and memory. In this
part, we will discuss hippocampal memory tasks. That is because we need to understand behavioural
tasks, before we can explore the question of a long term potentiation is an important learning,
memory mechanism.
As you know, the hippocampus is very important for learning memory in humans. That is thought
because surgical ablation of the hippocampus affects memory in patients– patients that have been
treated for epilepsy, and unfortunately then suffering severe memory impairment. However, such
studies lack a little bit of precision because when you do surgical removal of a brain area, you would
affect a number of different brain areas and not only the hippocampus.
So, one would need to do animal experiments to know exactly, kind of like, what the role of the
hippocampus is by just lesioning the hippocampus, for example. So the behavioural task we’re going
to discuss now are all sensitive to lesions of hippocampus in rodents. And we will primarily focus on
mice because mouse studies have been used primarily to address the issue of long term potentiation
and memory. But it applies also to rats.
Before we go into the task, we will distinguish between different memory processes. And then we will
go into behavioural tasks.
Slide 4
Memory can be distinguished also on the time scale and by the brain areas involved. So we can talk
about a short term memory, or long term memory, a working memory, which is very short lasting.
And then the brain areas involved, so memories that require the hippocampus, memories which do
not require campus, for example memories that require the cerebellum. And so the focus here will
be really on, as I said, on memories that require the hippocampus.
So if you were to lesion the hippocampus, the memory is impaired. But it’s not to say that memory
is stored only in the hippocampus. It was probably stored also outside the hippocampus. But the
hippocampus is just an essential component of memory. Then basically, we will focus primarily on
long term memory, because long term potentiation is thought to be an important mechanism to store
memory for very long periods of time.
Lecture transcript
Week 4 © King’s College London 2019 2.
Transcripts by 3Playmedia
In humans the hippocampus is important for declarative memories, as I mentioned in part one
briefly. So it’s important for knowing who the prime minister is, or who a particular sports star is, or
other memories, kind of like for example, when you remember where the cinema is. These type of
memories you are aware of require hippocampus.
Of course, in rodents we cannot really speak of so-called declarative memories because we do not
know what rodents are aware of. So, therefore, this definition of declarative memory in rodents
doesn’t really apply. However, in rodents, it has been shown that the hippocampus is particularly
important for spatial memory. So memories of to find a particular location in a space. Or contextual
memory– so these are memories to remember a particular environment.
These types of memories have resulted basically from the discovery of places in the hippocampus.
So these are cells that fire only when an animal is in a particular location. So, therefore, being
discovered by John O’Keefe at University College London. And their discovery basically suggested
that the hippocampus is important for making a spatial map of the environment. And in humans
maybe even a cognitive map, maybe in animals also cognitive map.
Slide 5
To study spatial memory, people use the so-called water maze. It’s a task that was developed by
Richard Morris at St. Andrews at the time. So what Richard Morris did is he basically designed a
swimming pool that contains a submerged platform. And the water in the swimming pool is made
opaque, milky. We use in the laboratory white nontoxic paint for children. So the animals cannot look
through the water and cannot see the platform.
What the animal has here is basically a platform it can rest on, as you can see on the right side. And
then a swimming pool, basically the animal cannot climb out. So what the animal can learn in this task,
ultimately, is to locate a platform location. Using these cues in a room, like for example, this chair or
this picture on the wall, so they can basically learn how the map of a room looks like, and where in
that map a platform is located.
Now this behaviour task is really complicated. So the animal cannot learn this in one trial. So it’s a
very difficult task. What the animal learns here is many different things. So when you put for first time
a mouse into the water maze, the animal will try to climb out of the pool. So they swim around the rim
of the pool to try to get out until it has learned that there’s no escape.
When the animal goes into the next learning phase, it will explore the environment. So in another
trial, it will basically swim around randomly, until it finally bumps accidentally into a platform they can
rest on. So the animal will learn that there is actually a platform I can rest on. So it learns to use the
platform, so by swimming around randomly.
When the animal gets more clever, and it will finally develop a strategy to locate the platform. So for
example, it might learn that the platform is located to a particular distance to the pool wall. So then
it can swim with a particular radius to locate a platform. So these are kinds of strategies that are socalled
procedure, basically. They’re relatively efficient, but they are not ultimate spatial learning.
So these procedure strategies ultimately lead to spatial learning. Every animal learns that a platform
is located in a particular location in the room. And once it knows where in the room it is, it uses the
map, the spatial map, to navigate to it, the platform.
Now we can test for spatial learning by after training, in that we remove a platform from the
swimming pool and allow the animal to search for the missing platform. So if the animal then swims
toward the missing platform location, and spends most of that time doing the so-called memory
probe trial when the animal has remembered the spatial location. So the animal searches in the area
where the platform used to be. Contrary, the animal searches an equal amount of time in all areas of
Week 4 © King’s College London 2019 3.
Transcripts by 3Playmedia
the swimming pool, when the animal indicates a random search for a missing platform in the memory
probe trial which shows what the animal has not learned the spatial location.
Slide 6
So this is what one does in the water maze. And now I’ll show you an example. So this example is
of some work which we have done in the laboratory. What we see here is after different training
sessions. We had 15 training sessions. We trained four different types of mice. So kind of like, for
simplicity, let’s just focus on WT, so wild-type mice. So these are the open symbols. And what we
record here is the latency in seconds. So this means the time the animal needs to reach the platform.
So for example, training session one indicates that wild-type mice, the open symbols need between
60 and 70 seconds in average to reach the hidden platform. Well, already in training session two, this
latency has declined to 30 seconds or 50 seconds. And with more training sessions, basically the
latency decreases until finally the animal reaches some kind of asymptote. So within 10 or 20 seconds
it can reach the hidden platform. So the animals have improved over time with a learning curve. And
so it’s clear learning in these wild-type animals, normal mice.
In contrast, the black symbols indicate a mutant mouse. We discuss this mutant mouse later in more
detail. But this mutant mouse, basically, has not improved over time, indicating that the mutants do
not learn. So something is impaired in these mutants that prevents learning. As I pointed out earlier,
such a curve, such a training curve alone, is actually not sufficiently indicative of impaired spatial
learning in the mutants or of spatial learning in the wild-type mice.
That is because wild-type mice could use, for example, an alternative strategy to locate the platform.
So we could circle the particular radius for example. And the mutants may have some kind of like
performance abnormalities. But let’s not go into mutants now. Let’s just focus more on how a normal
animal, a wild-type animal, would learn. And so, basically, just looking at this training curve alone, is
not sufficient evidence that there’s spatial learning in the wild-type animals.
To get the evidence, what we have to do is we have to give a memory probe trial. This indicates the
different strategies again, just making the point that there’s some learning, that there’s no escape.
It’s a first phase. And the use of a platform, basically one strategy learning, so these three things can
happen when you get an improvement in wild-type animals. But to the animals use really is spatial
strategy.
Slide 7
For a spatial strategy, we do a spatial memory probe trial which is indicated in this slide here. So
if we just look at control animals without normal animals, you can basically now analyse the search
patterns. So the platform has been removed. The animal is allowed to search for the missing
platform. And we divide the pool into four quadrants. The quadrant of a platform used to be the
training quadrant and and three other quadrants.
And you can see from this control example, that most of the search time the animal spends in the
training quadrant, that’s also been quantified below. You see that the animal spends about 60% of the
search time in that quadrant and not much time in the remaining three quadrants, indicating that the
animal has a clear spatial bias. So it clearly remembers where the platform used to be. So there’s
evidence for spatial memory. So spatial learning has occurred.
On the right, this is a drug-treated animal. We’ll come back to this in a moment. Just to make the
point, this would be an animal that does not have a spatial memory because this animal searches now
equally in all four quadrants, as quantified below. You can see that basically the search time in the
quadrants is between 20% and 30%. For each quadrant, there’s no difference between the quadrants
indicating a random search. So there is no spatial memory in these animals.
Week 4 © King’s College London 2019 4.
Transcripts by 3Playmedia
Slide 8
Now a more simple task that has been devised to assess for hippocampus dependent learning, is
the so-called passive avoidance task. So this is a task that can be learned in a single training trial. So
what’s done here is a rodent is placed into a lit compartment and the lit compartment is connected to
dark compartment. So a rodent, like a rat or a mouse, likes to be in the dark. So the animal would go
into the dark, once it is placed in the lit compartment. It goes almost immediately into the dark, where
a door can be closed and a mild foot shock can be provided.
So the animal here learns when that going into the dark could be dangerous because a mild foot
shock is provided. So the animal, therefore, would be learning to avoid to go into the dark. So at the
time of testing, when the animal is placed back into the lit compartment, the animal avoids to go into
the dark. So this is a passive avoidance task because the animal does not have to move to avoid the
dangerous situation. It’s passive.
So passive avoidance requires the hippocampus. So if the animals would obtain lesions of the
hippocampus, then the animals would not remember. But in the dark there was a shock, so we would
go again into the dark compartment. And they would just have no memory of the association between
dark and foot shock.
Slide 9
So a typical example of this task is shown here in this slide. So in panel A, you can see a wild-type
animal, so normal wild-type mouse. So at the time of training and the open bar shows you that the
latency to go into the dark is below 50 seconds. It’s maybe like 20 seconds. But after the training,
after one training trial, a wild-type animal avoids to go into the dark. So it now spends more than 250
seconds in the lit compartment.
And on the right, as comparison, it’s just a mutant animal. And this mutant animal has also a very
short latency at the time of training. So it’s motivated to go into the dark, where it gets a shock,
but it will not remember that it has received a shock in the dark, basically. So again in this mutant
something, some process is impaired that impairs hippocampus dependent memory formation.
So this is the one trial learning task. So this is in contrast to the water maze that requires many
training trials. And the advantage of one trial learning tasks is that all the animals learn at the same
time, so after one trial that is. So if you want to study molecular or cellular processes that underlie
learning and memory, then this is a great task because all the animals are kind of synchronised. They
have learned at the same time.
In the water maze in comparison, some animals may learn after training session five, others after
training sessions seven, and so on. So the animals are not synchronised. So they learn at different
time points. And it’s very difficult to find out when each animal has learned. So therefore, there is
much more noise in analysing molecular and cellular mechanisms that underlie learning and memory.
The other advantage of one trial learning task would be that one can easily distinguish from short
term memory to long term memory. So that is the advantage of the synchronisation. So that because
we know exactly when the animals get trained. And then we can look for example, for short term
memory 30 minutes after training, as indicated in this cartoon here. It’s in panel B. So this is from
memory 30 minutes after training. Again the wild-type animal show an avoidance. The mutants do
not show an avoidance. So the mutants are impaired in short term memory. And we can look at long
term memory, which is usually done 24 hours after training. And we basically see that the long term
memory is impaired in the mutants but not in the wild-type animals.
So the mutants have an impairment in short term and long term memory, in this case. But because
short term memory is impaired, it’s likely that this is because of the long term memory impairment.
Week 4 © King’s College London 2019 5.
Transcripts by 3Playmedia
Slide 10
So this part introduced you into some behavioural testing of protocols for assessing hippocampus
dependent memory.
We needed to have this session basically to familiarise you with these behavioural tasks because
these tasks have been used to study whether LTP is a memory mechanism. So LTP was measured in
the hippocampus, these are hippocampus dependent memory tasks. And the questions in the next
part will be is what happens when we block the induction of long term potentiation. What happens to
learning of memory? What happens if you block the maintenance of long term potentiation, that would
erase existing hippocampus dependent memories? And so on.
Week 4 © King’s College London 2019 1.
Transcripts by 3Playmedia
Topic 1
Learning, memory and synaptic plasticity - Part 4 of 4
Week 4
Biological basis of learning, memory & cognition
Module:
Biological Foundations of Mental Health
Professor Peter Giese
Department of Basic and Clinical Neuroscience
Slide 3
Welcome to our fourth part on synaptic plasticity and memory. So in this fourth part, we will finally
discuss whether LTP in the hippocampus is important for learning and memory. So you will learn
about approaches, how LTP was studied in this context, whether LTP is actually induced by training
in a memory task. I will explain in a moment. And we will learn about methods– how LTP has been
manipulated, and whether this has impacted on learning and memory.
For LTP– LTP, as I mentioned in the second– or first part, in particular– is an interesting phenomenon,
because it’s long-lasting synaptic plasticity. It’s an enhancement in synaptic plasticity. It has input
specificity. That is also a very interesting property. It has associative properties, co-operative
properties. So all of these properties make it a very intuitive model for learning and memory. But in
addition, it also follows Hebb’s postulate.
Slide 4
In 1949, the Canadian psychologist, Donald Hebb, wrote a book The Organization of Behavior. And
in this book, he illustrated a principle in how he thought that neurons should behave when an animal
learns new information.
And he basically wrote is when an axon of neuron A excites neuron B, and repeatedly or persistently
would do so, when some changes like growth processes, or metabolic changes would take place, in
one or both cells, so that A’s efficiency to fire B is increased.
So certainly speaking, LTP follows this phenomenon. Because when you have a high frequency
stimulation, where basically, you enhance the synoptic transmission between neuron A and neuron B,
and so the likelihood that neuron A fires neuron B is increased. It is strictly speaking, however, not the
firing of what is enhanced, it’s the synaptic transmission that is enhanced. But loosely speaking, LTP
follows the Hebb postulate.
Slide 5
Up till this point, we have always talked about LTP after electrical stimulation. So LTP was induced by
electrical stimulation. And the question is, really, does such type of synaptic plasticity really exist in a
behaving brain?
Lecture transcript
Week 4 © King’s College London 2019 2.
Transcripts by 3Playmedia
So because when we use electric stimulations, we just assume that these electrical activity patterns
really exist. To address this point, researchers have asked the question whether training in a
hippocampus dependent memory task can actually induce LTP. So to this end, we have used this
passive avoidance task, which I explained in part three.
So the passive avoidance task has restraint. But it’s a one trial memory paradigm. It’s hippocampus
dependent.
So just to briefly repeat, we put an animal into a lid compartment– a mouse or rat– in this case, rats
were used. And then the rat goes into the dark; door closes. A shock is provided– a mild foot shock.
And so the animal learns that the dark– it associates with a mild foot shock. So at the time of testing,
the animal is put back into a lid compartment. And the animal would avoid to go into it dark.
Slide 6
Animals have been trained in this task. And before training, multiple electrodes were implanted
into area C1 of the hippocampus, as shown in panel A, here. So you see basically eight recording
electrodes– one to eight– and one stimulation electrode.
The reason for having so many different recording electrodes is that the idea is that behaviour is
used to induce LTP. Maybe only a small set of synapses will undergo LTP. And many other synapses
will not be affected. Whereas, if all of the synapses would undergo LTP when the animal could only get
one memory, and that’s the end of it.
So, obviously, the animal should produce many, many different memories. So therefore, only a small
set of synapses should undergo LTP. But then you should use multiple recording electrodes to detect
the LTP. So they used eight recording electrodes.
And so we trained the animal in the inhibitory avoidance task. And in panel B you can see where
the trained animals, in average, basically– are shown in orange– show after the training and
enhancement of synaptic transmission, and where enhancement seems to be long lasting. So it’s
measured here up to four hours after the training.
Panels G and I show this in more detail. So in this case, the field, EPSP, is a measure to plot it, versus
basically the number of synapses or the percent of the electrodes– and for four different groups of
animals.
So the animals that have been trained, this is the red circles. And the animals that are naive– let’s
just focus on these two groups first. The naives are blue. And so what you see here is the population
of synapses are such that for blue, that the average is around 100%. And when basically, after the
training, the red ones you can see that they have shifted to the right. You see some synapses that
show more EPSP slope. So some synapses show 150% EPSP slope, which you never observe for
naive animals.
So some synapses are severely potentiated. That is all true for 30 minutes, 1 hour, and 2 hours after
conditioning. So these synapses have produced LTP.
Slide 7
There is also other behavioural control. So for example, animals that have received the shock only–
in other words, the animals are placed immediately into a dark compartment, get a shock, and so it’s
not– so the animals really don’t know, actually, what the task is about. They just experience a shock,
but they cannot really make an association between the shock and something else.
Or alternatively, an animal is just placed into a lit compartment, allowed to go into the dark, but
Week 4 © King’s College London 2019 3.
Transcripts by 3Playmedia
doesn’t receive any shock. So again, there’s no association between darkness and shock. And again,
these two groups, which are just performance controls– no LTP was induced. So only LTP was
induced when animals were trained in the inhibitory avoidance task.
So this shows that behavioural training can induce LTP. And in this paper– please, have a read of the
paper, if you’re interested. This paper will show that it was actually NMDA receptor-dependent. So
LTP really occurs during memory formation.
Slide 8
The next set of experiments, which were done to probe whether LTP is important for memory, is to
manipulate LTP and then to test the impact on water maze behaviour. And so as I explained in part
three, with Morris water maze task– so the water maze task assesses for whether there is spacial
memory or not.
In particular, after training, when a memory probe trial is given, during the memory probe trial the
platform is removed from the swimming pool. And the animal is allowed to search for the missing
platform.
Slide 9
So here the animals– some rats were treated with an NMDA receptor blocker, AP5– and versus
controls. And you have seen this graph before. It was a graph for part three.
It indicates, basically, in a memory probe trial, the AP5 treated animals have a random surge. So
there is no spatial memory. Whereas, for control animals, which are just treated with some kind of
saline, they have a spatial memory, because they surge most of the time in the training quadrant.
Now this experiment then indicates that blocking NMDA receptor impairs spatial learning. So since
NMDA receptors are important for the induction of LTP, which suggests that LTP is important for
learning. However, as every experiment, there is a catch.
The catch is that the drug dose in these experiments was relatively high. So the animals had some
performance abnormalities. They fell off a platform, for example.
Furthermore– and blocking NMDA receptors does not only block the induction of LTP, it also blocks
long-term depression, for example. And so in principle, one cannot exclude the possibility that
impairment in long-term depression has resulted in the impaired spatial learning.
Slide 10
Another way of blocking LTP is to generate mutant mice. So this is because only a limited number
of drugs is available to block particular molecules, like the NMDA receptor. Whereas genetics, with
mouse genetics, one can manipulate any gene of interest.
The Nobel Prize committee awarded the Nobel Prize for gene targeting to manipulate the mouse
genome. So in this case, you basically manipulate any gene at will in mouse embryonic stem cells. And
then you generate a mutant mouse, nowadays, where even more quicker methods, like the CRISPR/
Cas genome editing system, for example, which can work directly in fertilised embryos in zygotes. So
you can manipulate any gene at will in a very quick way.
And so the rationale behind such experiments is that you, basically, impair a molecular process. And
that impairs maybe LTP. And then you can ask the question, does it impair learning and memory?
Week 4 © King’s College London 2019 4.
Transcripts by 3Playmedia
Slide 11
So one experiment that was done by Susumu Tonegawa, the Nobel laureate, who got actually his
Nobel Prize for discovery of antibody diversity and how this is being generated, he’s now very
interested in learning and memory. So one of his early studies when he turned into the learning/
memory field was to knock out an essential NMDA receptor subunit only in the hippocampus.
And so basically, we sent the NMDA receptor– consists of the NR1, GluN1 subunit. That is an essential
subunit for all of NMDA receptors. So if you knock out the gene that encodes GluN1, you block NMDA
receptors.
And you can make a conditional knock out. That’s indicated here. So you can basically introduce
with LoxP, recombination sequences in the NMDA receptor gene, in the GluN1 gene. And these LoxP
sequences are recognised by Cre recombinase, that cuts out when the sequence in between the
LoxP sequences and leads to knock out.
And now in that case, you get only a knock out when Cre recombinase is active. And Tonegawa
succeeded to have a mutant mouse that has an active Cre recombinase only in hippocampal area
CA1.
Slide 12
So this is indicated here, in the next cartoon. You can see, basically, this is a mouse that provides Cre
in area CA1. And in area CA1, there was no expression of the GluN1 subunit anymore.
This is an in situ hybridisation, which shows in control animals– nice expression of GluN1 in CA1,
CA3 and then the gyrus of the hippocampus. But you can see here with the signal in CA1 of the
hippocampus is severely reduced. It’s actually later confirmed by electrophysiology as completely
abolished in the CA1 unit.
Slide 13
What’s the consequence of knocking out NMDA receptors in area CA1? Well, it impairs spatial
memory. In the probe trial, after water maze trainings, these mice have been assessed. And you
can see here four groups of mice– the wild-type mice, which basically show in the memory probe
trial, that they search most of the time in the training quadrant. And then you have two other mouse
lines, which are just controls mouse lines that have a GluN1 flanked by LoxP sites– so they have also
a spatial memory when T29-1, just mice that have Cre recombinase. And they have also a spatial
memory.
And then, finally, we see one specific knockout mice for GluN1. They show a random search. So there
is no spatial memory.
So this is consistent with the finding where the pharmacological block of the NMDA receptors blocks
spatial memory. And it suggests where LTP– impaired LTP induction impairs spatial memory.
However, the catch is, of course, that blocking NMDA receptors can also block long-term depression.
So it’s unclear whether these impairments have resulted from impaired long-term depression or
impaired long-term potentiation.
Slide 14
Now an enzyme that has been very much implicated in long-term potentiation is CaM Kinase II, as
illustrated in part two. So CaM Kinase II, is the calcium calmodulin-dependent kinase. It has, as almost
every kinase, a regulatory domain and a catalytic domain.
Week 4 © King’s College London 2019 5.
Transcripts by 3Playmedia
Normally, the regulatory domain inhibits the catalytic domain. So the enzyme is not active. However,
when calcium levels rise in the post-synapse, calcium binds to calmodulin, and the calcium calmodulin
complex binds to a regulatory domain, inducing a conformational change. So now the enzyme can
phosphorylate substrates.
This configuration of the enzyme is dependent on the presence of high levels of calcium calmodulin.
So when calcium calmodulin levels drop back to baseline, the enzyme is inactive. However, the
enzyme can undergo an autophosphorylation, it can phosphorylate itself at threonine-286.
Threonine-286 is within the regulatory domain. And when you get phosphorylation of threonine-286,
you get a further conformational change that entraps the bound calcium calmodulin. So we’re now
at baseline levels of calcium calmodulin. There is still bound calcium calmodulin, and the enzyme
remains active.
This autophosphorylation event has been thought to be an important event for maintaining
long- term potentiation. And therefore, there was interest to make a mutation that blocks for the
autophosphorylation. And actually, I myself did this work when I was in New York. As a post-doc, I
made a mutant mouse that has threonine-286 change to alanine.
So if you change the amino acid to alanine, the alanine cannot be phosphorylated. And it doesn’t have
a hydroxyl group. So you basically block the autophosphorylation at threonine-286. And you end up
with an enzyme that is normally active with a presence of calcium calmodulin.
Slide 15
So what happens to these animals? So they have severely impaired LTP in CA1 in the hippocampus.
So this is indicated here. So in panel A you see the open circles is LTP in normal wild-type mice. So
you have first STP and then followed by LTP.
In the mutants we have only some kind of PTP and maybe STP briefly, but mostly PTP, and then no LTP.
So LTP is completely abolished. Well, it was also shown in vivo, so in the alive animal where LTP was
completely abolished.
So the autophosphorylation of CaMKII at threonine-286 is fundamentally important for induction of
LTP. It’s not so clear whether it’s important for LTD. It seems to be rather important for LTP.
Slide 16
What happens to animals in the water maze? Well, you saw this graph also before, where basically
there is no improvement when the animals were trained in the water maze. And at the memory probe
trial, you can see on the right where mutants are shown in black, there is an equal surge in all four
quadrants, indicating a random surge. Whereas wild-type animals have more surge in TQ.
In this case, please note that the proximity is plotted, which is inversely correlated to the time spent
in the area– so if you have less proximity, when you spent more time basically, in the target quadrant.
And so in wild-type animals, they have selective surging in the target quadrant, while various mutants
have a random surge, indicating they have no spatial memory.
So these mutants were in lack of LTP induction and were lack spatial memory, further strengthening a
correlation between LTP and memory.
Slide 17
Finally, we can also manipulate the maintenance of LTP and ask the question whether this impairs an
Week 4 © King’s College London 2019 6.
Transcripts by 3Playmedia
established memory, which is work that has been done by Todd Sacktor, who is based in New York.
And so what he has found, he has identified a peptide that blocks protein kinase M zeta, where it’s
important for the maintenance of LTP. And this experiment shows where the peptide, ZIP, can block
the maintenance of LTP.
In panel D you see that LTP was induced and then recorded for many hours. So this is a late-phase
LTP. After a few hours the LTP consolidates into late-phase LTP when it’s being maintained.
And now ZIP is provided to a hippocampus slice or recorded in the hippocampus slice. And when the
ZIP treatment wipes out the LTP, the ZIP is therefore blocking the LTP maintenance mechanism. And it
was suggested to work via PKMzeta.
However, nowadays, we are not so certain. Because when PKMzeta is knocked out in mice, ZIP is
still able to block LTP maintenance. And therefore, it might act also on our molecules and not only on
PKMzeta.
In panel E is another behavioural task that requires the hippocampus– and to ask the question,
whether blocking of a ZIP erases hippocampus dependent memory. So the animals, in this case, were
placed with the rats. They’re placed on a rotating platform.
When the platform rotated into a particular area, the animal would receive a mild electrical foot
shock. And so what the animal learns from this task is to avoid to go into the area. It’s an active
avoidance task.
Basically, so then he has to move away from this area. It has to identify where the area is using these
cues in the box. So you can see that it requires a few training trials. And then the animal can avoid to
go into the danger zone.
So initially, it goes very quickly into a danger zone– within just maybe 10 seconds or so. And after
training, it needs more than 200 seconds to be in the danger zone.
So well, this memory can last a long time. So for example, it can be tested at 24 hours after the
training– so one day after the training. So when the animals are treated with saline, you can see that
they are still avoiding to go into the danger zone.
However, when our animals are injected with ZIP, well, there’s no memory, because now we the
animal is caught almost immediately back into the danger zone. So ZIP has erased the memory here.
And it has also erased LTP. So it has blocked the LTP maintenance. And so that’s a very strong
correlation between LTP maintenance and memory maintenance.
Slide 18
Taking these things together, there’s very strong evidence that LTP is a memory mechanism. So there
are theoretical grounds, loosely speaking. LTP follows Hebb’s postulate.
We have seen that LTP can be induced by behavioural training. Furthermore, blocking LTP induction
seems to block spatial memory formation. And further, blocking LTP maintenance seems to erase a
spatial memory.
