Abbreviations: LTP= Long-Term Potentiation

Neuro: 1:00-2:00Scribe: Sally Hamissou/ Sunita Jagani

Monday, March 16, 2009Proof: Dilly Willy / Dylan Vaught

Dr. LesterSynaptic PlasticityPage1 of 7

Abbreviations: LTP= long-term potentiation, CS= Conditioned Stimulus, US= Unconditioned Stimulus, UCR= unconditioned response, CR= conditioned response, AP= action potential

Synaptic Plasticity

I.Synaptic Plasticity [S1]

We will follow along with Dr. Wyss’s lecture about learning and memory so we will go back into the hippocampus. He used Pavlov’s dog to show procedural learning in the cerebellum (e.g. the conditioned eye blink reflex).
Humans do most of their learning in an associative manner. We can take Pavlov’s dog and extend it to declarative memory, things that are transiently stored in parts of the brain, like the hippocampus, and then stored in the cortex.
We know this because we all have pneumonics that we use to remember things.
The reason associative memory is useful is because it is like storing information by imagining it like a puzzle, the information is different parts of the puzzle but we only need to see little parts of the puzzle to recall the whole memory or the whole puzzle to identify what the puzzle is.
We can store things by association things get associated with one another and we don’t need to remember the whole thing just a part of it to get everything to come back out.
We will use Pavlov’s dog again, and if you keep that in mind, what he will try to show us at the cellular level of current understanding with how associative memories may be stored through the hippocampus and then on into the cortex.
Associative means we are associating two things together to form a memory. We look for mechanisms that may be two pathways, come together, and change something. At the molecular level, the change is at the receptor the NMDA receptor
It’s an important receptor for remembering and forgetting.
The NMDA is a molecular plasticity switch in the brain.
Highest density of NMDA receptors is in the hippocampus
Wednesday’s lecture about addiction, think about addiction as a form of associative learning. Those with addiction problems have associated cues that trigger cravings.

II.Learning objectives [S2]

Understand the properties of long-term potentiation (LTP) that define it as a model of experience-dependent synaptic plasticity
Might have heard of LTP. This is an activity dependent (experience dependent) change in synaptic strength. The synapse may keep stimulating synapses to get a response of x, and if we do something to the synapse that may involve learning or associating pathways, we change that response and it gets bigger and stays bigger for a long time.
Discuss the characteristics that make NMDA receptors coincident detectors cable of initiating associative information storage (Pavlov’s dog)
NMDA is a glutamate receptor; remember glutamate is primary excitatory transmitter in the brain.
Describe the relationship between NMDA receptors, LTP and behavioral memory
How this is involved in real learning

III.Learning & memory in taxi drivers [S3]

Study done in humans
Took cab drivers from London and gave them a memory recall task.
They had to imagine picking up passengers at some location and dropping them off somewhere in London. They had to recall it based on spatial information; they did not have a map to trace the route, so they had to use spatial recall. When they did this, they found that the left temporal lobe lit up, where you find the hippocampus, which is involved in recall.
Short and LTP- the hippocampus is important for changing short term information into long term information.
The case about HM - had good short term memory (as long as he didn’t get distracted) but couldn’t put it into long term memory
Hippocampus provides a special role in limbic system that is associated with the Papez’s circuit for encoding spatial information, so there are place maps in that area.
During recall you expect to see parts of the brain associated with specific tasks light up.
Memory that is visual, you will see visual cortex responding.

IV.Papez Circuit/Loop [S4]

Know this Papez circuit within the limbic system and this is the only bit of anatomy we need to know and it illustrates a lot of parts in the limbic system
Hippocampus is in the temporal lobe, we get information leaving by the fimbria and fornix around to the mamillary bodies. We have the mammillothalamic tract to the anterior thalamus. Up to anterior cingular cortex throughout cortex and back out to hippocampus- know this memory circuit
Other information that feeds into circuit: amygdala, temporal lobe, other parts of the brain, striatum, olfactory bulb

V.Place cells / maps [S5]

Don’t’ remember details but just know at a single cell level there are cells in the hippocampus that are place cells
If you put an animal in a novel environment, the cells start to encode for direction or position within the environment
If the animal is facing a certain way, some cells will prefer that certain orientation, as in visually direction sensitive cells in the retina. So depending on what direction the animal is facing, the cells fire AP or become silent.
If an animal wanders around this environment, some cells respond strongly when it is in a particular location and then turn off, and another set will turn on at that location, so it seems that it encodes space.
[SQ]: visual feedback and dependence [A]: It requires visual information to form place cell. Once it is encoded in the same environment, and it doesn’t have to rely on visual cues totally. The exact details of the place maps are being worked out. We do know that they extend from the hippocampus to the cingular cortex, so the entire area is involved.
We have to get from the cells (synapses) in hippocampus to understand how learning happens

VI.Hebbian learning….in theory [S6]

Hebb was a psychologist.
Idea is that if you have repeated conjoint activity between pre and postsynaptic cells, then you can have a change in synapses that connects the cells.
The pre and postsynaptic cell is active, the synaptic strength changes and that is Hebb’s postulate, and several years later that was demonstrated physiologically.

VII.Synapses are plastic [S7]

How do we make synapses remember things? We encode things by pattern of AP. For a location on the body, it will ultimately result in an AP being initiated and going up to the brain.
The frequency of AP may tell you something about intensity.
The frequency or pattern of activity may be important to tell if you should modify the synaptic strength, maybe you should make it stronger or weaker.
At NMJ: stimulate one AP at the motor neuron and recording synaptic response to NMJ.
Stimulate one AP and in the next one the response stays stable, you get a similar response. High frequency train of AP, the responses get bigger and stay bigger for a while then fade back down.
One idea about changing the synaptic path, you alter the pattern of information coming back
Synapses “remember” previous activity
Only remembers things for a second or so then goes back to normal, which is good for muscles cell to contract and then go back to resting position and to do that reliably.
Any memory control occurs centrally

VIII. Can synaptic plasticity explain learning? [S8]

Does it result in a different outcome if done in CNS? Remember Dr. Wyss told us about learning and memory, so maybe the rain deals with this differently.
Pavlov’s dog slide we have seen before- we don’t have to go into too much other than remember what is the CS and these US
CS is something that produces a neutral weak outcome but can be altered when paired with the US because the US produces a strong response
We will map those inputs onto the hippocampus

IX.Learning Objective #1 [S9]

Understand the properties of long-term potentiation (LTP) that define it as a model of experience-dependent synaptic plasticity

X.Figure of hippocampus [S10]

Hippocampus is down in the temporal lobe.
Circuit where information goes through, cycles around, and passes back out. You might have recurrent looping of information, with information coming in and looping back out several times and we decide if it goes out to permanent storage.
Orientation information comes in from entorhinal cortex, the adjacent cortex, and it starts to feed into the proper neo cortex.
The Neo cortex is a 6 layered structure, and the hippocampus cortex has 3 layers and does connect with neo cortex via the entorhinal cortex.
Input comes in, synapses on one set of cells in the dentate gyrus, which is sometimes not referred to as hippocampus proper.
Those cells synapse with another set of cells, CA3 pyramidal cells, and those make connections with CA1 pyramidal cells.
So there are 3 synapses all serially organized and all of them can undergo LTP
Information comes out, these cells go to the subiculum and we can see how it loops into the entorhinal cortex and out. The axons bifurcate and the information can pass out via the fornix and fimbria can go from subiculum cells to the CA3 pyramidal cells.
The CA3 takes information to the other hippocampus but this is our main output to the fornix and back around to the mammary bodies.
Overall: know 3 synapses all have glutatmate and all can undergo plasticity.
As Dr. Wyss said, if we lesion CA1 part, ischemic damage. So if you lesion one part of the serial pathway, the information doesn’t loop around anymore.

XI.Long-term potentiation (LTP) [S11]

Don’t worry about the wave forms just worry about the size of the synaptic response and whether they are bigger or smaller
This is the first study that was done in vivo in rabbits.
Used 2 stimulating electrodes that stimulated two separate inputs into the dentate gyrus
One of the pathways was controlled and was stimulated once every 30 seconds and the response stayed stable
Referring to before wave: so before and after, it didn’t change, it stayed the same size. Referring to after wave: this synaptic response got bigger and we’re recording the synaptic response in the dentate gyrus. It gets bigger here and why does it get bigger? Because they gave high frequency trains, a burst of AP to the stimulating electrode and now the AP comes down to synapse much faster so the response gets bigger and it stays bigger and it ultimately saturates. That makes sense because you only have a certain number of synapses they can only get to a certain size depending on the mechanism.
The responses got bigger and stayed bigger for days and weeks in vivo so the animal has a completely changed hippocampus. This was really cool because no one had demonstrated a change in synapses that had lasted that long.
To this day, this is the best model that we have.
This experiment hasn’t been done in humans but this shows that hippocampus is actively involved in learning memory tasks and recalling.

XII.Properties of LTP [S12] ***for this part of the lecture I tried to follow along with which graph he was referring to. He bounced around a lot but I tried my best to group everything with their respective graphs***

We have to learn a little about the properties of this graph.
Think of Pavlov’s dog and you will remember this.
Pavlov’s dog, we have pathways coming into hippocampus, associating information, and we have NMDA receptor (the molecular switch).
Pavlov dog
CS, the weak input, produces a small response. The teacher, the food, elicits a strong UCR.
If we give a high frequency trained to a weak response, it doesn’t do anything because it is weak. There must be a threshold of activity before you can get plasticity.
Referring to cooperativity graph: weak before/after no change. Now if you stimulate the weak at the same time you stimulate the strong, the strong potentiates and gets bigger and stays bigger, this is long lasting plasticity. Referring to associativity graph: But since you are doing this at the same time as the other, it makes the weak one stronger. You are associating information in those two pathways and this really happens. The strong has conditioned the weak, long term plasticity of the weak.
The idea of these first two here (A & B) that there is some sort of threshold of activity, if not provided you don’t potentiate. But it may be provided by this one, telling this one to change, so there is some amount of activity and that relates to NMDA receptor.
Referring to specificity graph: important because it satisfies Hebb’s postulate you need pre-synaptic and post-synaptic activity. You can stimulate the strong one all you want but if you don’t pair it with the US, then the US never learns. So we need to pair them and that is the same as Pavlov’s dog.
If you don’t ring the bell, then they don’t learn to associate the bell with the food.
This tells you that you need presynaptic activity and (now referring to graph B) you have post-synaptic activity that spreads down and influences the synapse.
[SQ]: Explain difference between A and C on the figure. [A]: Referring to graph A- In any form of learning you need both stimuli. Referring to graph C- if this is the bell, you will never learn if you are not presenting it at the same time. If you keep ringing the bell and never give the meat, you never learn the association of salivation with the bell. So you allow these to go together to allow the graph C- pre and post is tricky. Graph A- the postsynaptic activity is due to the strong input that spreads down here and presynaptic activity is releasing glutatmate and you need to post-synaptic activity glutamate, graph B the reason you need that is…

XIII. Learning Objective #2 [S13]

Discuss the characteristics that make NMDA receptors coincident detectors cable of initiating associative information storage (Pavlov’s dog)
These are the required to activate NMDA receptors because that is the switch (continuing answer to question).

XIV.Cellular mechanisms underlying LTP [S14]

This is LTP. Here’s our synaptic response- we are given a trained highfrequency input. We potentiate the response, it stays bigger. We worry about the induction- and how does the change happen. And then worry about why it stays big afterwards. During the high frequency train, something happens. But, then we go back to our normal stimulation, then it stays big. It remembers the train.

XV.Excitatory synaptic transmission [S15]

Illustrate - what goes on at these glutamatesynapses in the brain.
2 types of post-synaptic receptors
AMPA receptor – doesn’t really matter, not the NDMA receptor.
NDMA receptor
Remember-under normal condition-everyday resting membrane potential, NDMA does not participate in synaptic transmission, hardly at all if any. It’s the other receptor.
Here is the synaptic response, if you block the NDMA, it doesn’t change the synaptic response. Those receptors aren’t playing much of a role in the response.

XVI.LTP depends specifically on NMDA receptor activation [S16]

If block NDMA, at the time I give high frequency train, it completely blocks the plasticity, you don’t get LTP.
You get a transient change, but it goes goes back to normal.
Wash the blocker of, you get LTP. That tells you that, the NDMA is at the synapse, doesn’t do much until you give high frequency train is given.
That’s where the idea that this is molecularplasticityswitch. It turned on by high frequency stimulation, associative high frequency stimulation, it causes the synapse to change, and it turns back off again.

XVII.What is special about NMDA receptors? [S17]

Why does that happen? And how does it fit in with associative?
Remember different types of channel- NDMA is a ligand-gated channel and also it needs voltage as well.
It also needs voltage because Mg trapped in the channel under normal conditions. That’s why it doesn’t have a role in normal synaptic transmission. Mg has to be kicked out of the channel, so that post-synaptic depolarization kicks it out.
NDMA receptor is a closed channel, waiting for glutamate, glutamate opens the channel, as soon as it is open- Mg plugs it up, so it doesn’t play much of role.
But, if we give a lot of depolarization, this is our strong input now depolarizing that cell really strongly, it kicks the Mg out. Now, ions come in through that channel.

XVIII.NMDA receptor: a molecular switch [S18]

This is trying to put all of those together. This needs to make sense to you. Remember, Pavlov’s dog to begin with CS and US. This is represented by our weak input which is CS, and strong input as US.
Hebb’s postulate, you need both conjoined pre-synaptic and post-synaptic activity.
NDMA needs joined pre-synaptic and post-synaptic activity. It can explain everything- it can put it together.
We call the NDMA co-incidence detector.
If presence here at this synapse, well- its present at both that’s why they both can change.
But, if it present here (not sure where he’s pointing at), then it can detected pre-synaptic activity coming from here because that’s glutamate being released. It needs glutamate to open it. And it can detect the spread of depolarization from the strong input down the dendrite. So it spreads, it depolarizes and it kicks Mg out and it can be activated.
It’s a co-incidence detector. It will not work if the two conditions are satisfied.
So if I don’t stimulate the input, it doesn’t matter the how much depolarization spreads the dendrite, the channel will never open because there is no transmitter there. You need channel to be open, to get the ions to flow through.
Again, if I stimulate this pathway, and provide glutatmate, it’s a weak input, doesn’t depolarize enough, so it never kicks Mg, so it doesn’t get activated. So both conditions satisfy this.
At the molecular level, and the circuit level, so we understand quite a bit how these process could accounts for associate information storage.
[sq]- is there Ca input?
[A]- One of the peculiarities NDMA receptor is that they do flux in Ca. When you kick out Mg, the current is carried by Ca. And Ca does turn on various things and causes the changes to happen.
[sq]- Something about the slide and pre-synaptic and post-synaptic on the diagram.
[A]Happens on the same dendrite. They are both pre-synaptic. They are both the synapse. We are only talking about what’s happening at the weak synapse. Think about p. dog, it’s the strong unconditioned input that trains the weak or conditioned input.
Diagram on the board- Thisis our big synapse. A little synapse. This one releases glutamate, and this one we get lot of depolarization that spreads upwards, it can passively spread so this one is depolarized. The same thing happens here, this one satisfies itself because it is a strong input, so it has glutamate and depolarization and that’s why it potentiated anyway in an unassociated manner.
But one of the ideas it has been shown, that it can happen with single input. So one idea is that now that the reason you see this potentiate as an unphysiological, we have never this massive input stimulated together to potentiate itself. It’s representative of an unconditioned input, which we don’t really know what it looks like.
Depolarization and transmitter, and the key is NDMA receptor.
If I ask questions- if you understand this diagram- this is the majority of what you need to know to understand NDMA and associative plasticity. You should be able to work yourself through this.
[SQ] – Something about weak response and depolarization
[A]:Weak input has glutamate, so the glutamate binds to the receptor, but it’s so weak t hat it doesn’t depolarize that part of the dendrite at all.
[SQ]: If you block the NDMA receptor, what happens?
[A]: If you don’t block the NDMA, you don’t get any plasticity happening at all. You stopped GLU from activating it. You still would depolarization, but preventing glutamate can’t bind to the NDMA receptor, which is the antagonist does.

XIX.How does the NMDA receptor cause a change in synaptic strength? [S19]