Memory (lecture based) Flashcards
What is Hebbian learning?
“when an axon of cell A excites cell B and repeatedly or persistently takes part in firing it…. A’s efficiency, as one of the cells firing B, is increased”
Neurons that fire together, wire together
Does LTP have anything to do with memory?
It is suggested that synaptic strength might be enhanced by concurrent activation in pre- and post-synaptic neurons (exactly what happens in LTP)
-> Need synaptic input and depolarisation of receiving cells so, in a sense, what Hebbian proposed does actually map on to what is happening in the reality of LTP
Describe the evidence in favour of link between hippocampal LTP and memory
Widely used ‘Moris Water Maze’ which examines the role of the hippocampus in spatial learning
= relatipnship between hippocampal LTP and memory
The experiment
Tank filled with water; submerged platform
Water made opaque by milk
Rat placed in water and swims around; eventually finds platform
The rat can find out where it is in relation to things that surround it in the room
Eventually the escape latency gets shorter (finding the platform and escaping is quicker)
Rat then was put back in different locations
4 different start locations (N, S, E and W) used randomly
Eventually learns to swim directly to platform regardless and this is known as spatial learning
Second study measures spatial transfer test (after a number of trials the platform is taken away and we wait to see how long the rat spends swimming around in the quadrant of the maze where the platform was- still a measure of spatial learning)
Both studies were disrupted by hippocampal lesions- so we know hippocampus invovled in memory
Spatial learning in this task invovles the hippocampus, this task is a perfect paradigm to look at the relationship between LTP and spatial memory
Studies showing a link between LTP and spatial learning: Morris (1989)
3 days of training; 8 x 2 min trials per day
Animal removed if failed to find platform in 2 min
Received bilateral injection of NMDA (LTP involves the induction of NMDA) antagonist into hippocampus; two doses – low and high
High dose significantly lengthened Escape Latency on Day 3 reflected a struggle to learn
Low dose – no effect
Examined the effect of same manipulations with the same doses on hippocampal dentate gyrus LTP
High dose almost completely blocked LTP
Low dose – no effect
So the ability to produce dentate gyrus LTP reflected the problems that animals had at learning in the morris water maze. Initial evidence that there is a link betweeen ability to learn spatial position and synaptic potentiaion in the hippocampus
One of the earliest pieces of evidence that linked these two things together
Studies showing a link between LTP and spatial learning: Wilson & Tonegawa (1997)
Used ‘knockout’ mice
Used cell-restricted Cre–loxP recombination
Leads to the deletion of a particular gene in a cell population which express a particular gene product (in this case a variant of CaMKII)
W + T produced mice lacking a critical subunit of the NMDA receptor (NMDAR1) in CA1 cells of hippocampus
Also looked at the ability of cells in the CA1 region to exhibit LTP by tetanic stimulatio of the fibers coming in to the CA1 cells
Findings
Tetanic stimulation of Schaffer collateral → CA1 synapses failed to induce LTP
LTP eliminatated
They also examined how it affected behaviour: Tested animals in Morris Water Maze
Received usual training sessions; then given spatial transfer test
Measure – do they swim in the correct quadrant?
Normals focused search on trained location
NMDAR1 knockouts did not
What do Wilson and Tonegawa (1977) demonstrate?
Evidence that disruption of NMDA-receptor mediated transmission in hippocampus affects spatial learning
Since LTP also affected, provides evidence for involvement of hippo-campal LTP in spatial learning and in memory
However their data was correlational
Studies showing a link between LTP and spatial learning: Castro et al. (1989)
Induction of LTP prior to learning should disrupt learning - taking away the flexibility of the system to learn
effect of inducing LTP in dentate gyrus on performance in Morris Water Maze
3 groups; control plus 2 tetanised groups
Animals received low frequency stimulation of perforant pathway each day to measure excitatory post synaptic potential in dentate gyrus
After 5 days of baseline, tetanised groups received: Repeated tetanic stimulation of perforant pathway –14 days
Findings- change in EPSP got gradually bigger across the days = the intensity of the LTP got stronger across those days
Tested behaviour on Day 19 [at peak of LTP induction]
Morris Water Maze– escape latency
Given 12 (60 s) trials in same day
Little evidence of learning in tetanised group- escape latency not changing but for control group the escape latency decreases as the animal is learning
At peak of LTP in paradigm, the animal has no ability to learn
BUT
Tested behaviour of all 3 groups on Day 34 [after decay of LTP]
Given 12 (60 s) trials
->Evidence of comparable learning in all groups!! Animals able to learn one LTP dropped back to near baseline
Evidence against a link between hippocampal LTP and memory: Abeliovich et al. (1993)
More knockout mice; this time lacking ‘protein kinase C’
[Intracellular] enzyme which alters proteins
Looked at effect on LTP and learning
-> Mice lacking protein kinase C failed to show LTP in thr CA1 subfield of the hippocampus
Unimpaired in Morris Water Maze
9 days of training; 4 x 1 min trials per day
Animal removed if failed to find platform in 1 min
No significant difference between ‘wild type’ and knockout animals in escape latency - blocking LTP didn’t affect learning
What reviews also show negative findings towards LTP and the hippocampus?
Eichenbaum (1996)
‘the take-home message…. is that the major approaches used to date have failed to provide compelling evidence for a connection between LTP and memory’
Shors and Matzel (1997)
‘there is little empirical evidence that directly links LTP to the storage of memories’
Bannerman et al. (2006)
‘neither the original hippocampal LTP/spatial learning hypothesis nor…. can adequately explain all of the available data’
Why are there discrepancies in the findings?
One idea is that all of the data asessed here has been entirely correlational- not causation
Maybe hippocampal LTP [and LTP in general?] has nothing to do with memory
Major factor = uncertainty about exact role of hippocampus in [spatial] memory- we dont understand enough about the hippocampus or its role in spatial learning to understand what the manipulations actually do to the hippocampus or to functioning
What other structure is involved in learning ?
Amygdalar complex which is closely related to the hippocampus and is involved in classical conditioning
What is classical conditioning?
US → UR [food → salivation]
CS + US → UR [food + bell → salivation] [conditioning phase]
CS → CR [bell → salivation]
Learning the association between US [food] and CS [tone] = establishing a simple ‘memory’
According to Pavlov (1927):
Conditioning requires changes in the pathway that mediates the CS, due to convergence of inputs from US
What did Rogan et al. (1997) find about the amygdala and learning?
‘Conditioned fear response’
US → UR [foot shock → fear response – ‘freezing’]
CS + US → UR [foot shock + tone → freezing] [conditioning phase]
CS → CR [tone → freezing]
Conditioned fear involves the amygdalar complex
This is a great study as classical conditioning is a straight forawrd behavioural paradigm and the amygdala is a straight forward neural structure
How does conditioned fear link to LTP in Rogan et al. (1997)?
(a) A tone [CS] evokes a field potential in amygdala; early part (the ‘dip’) reflects activity in a pathway bringing auditory information from thalamus
(b) During training, CS and US are paired [US just after CS]
(c) After training, CS produces larger amplitude ‘dip’ in field potential; CS also produces freezing
Important: repeated presentation of CS [tone] alone leads to ‘extinction’ of freezing response, and reduction in size of dip
Get an assocaition with the conduction of the conditioned fear, but also with the extinction of the conditioned fear as well
What did Schroeder and Shinnick-Gallagher (2005) find about the amygdala and LTP?
Conditioned fear
CS: white noise tone
US: foot shock
Paired 10 times a day for 2 days
Stimulated pathway from cortex to (lateral) amygdala
fEPSPs were measured in naïve control and unpaired and fear-conditioned animals; 1 day and 10 days post-testing
animals continued to express conditioned fear for at least 10 days after induced
So the behavioural side is long lasting
Synaptic side
->For the stimulation group, every point in the stimulation curve the field potential is bigger, both at 10 days and at 1 day,
so across time, driving the inputs to the amygdala to produce a bigger field potential response and long lasting potentiation
Lasts a similar amount of time as the behavioural responses in the animals
