Classical Conditioning Flashcards
What is an important issue in systems neuroscience?
Understanding the changes in the brain that underlie memory
What changes in the brain are we looking to examine?
Likely to be changes in synapses
e.g. in LTP where synapses become more powerful
What is an issue with examining synapses?
There are trillions of synapses in the brain that underlie many different behaviours
- some behaviours are complex e.g. those mediated by the hippocampus where LTP has been demonstrated
What is a good strategy for studying the neural basis of memory?
Picking a simple task to start with
Why is it best to pick a simple task?
Hopefully only a small number of synapses are involved here
–> gives us a chance of identifying them, and finding out how they work
If the same forms of synaptic plasticity underlie both simple and complex learning, understanding simple tasks will help with complex behaviour
What simple form of learning is easiest to study?
Classical conditioning
How does classical conditioning work?
Before training
UCS produces UCR
During training
Pair CS + UCS
After training
CS produces CR
What are the different types of CC?
- A good (e.g. food) or bad (e.g. shock) UCS
-Produces either salivations or limb withdrawal - Autonomic vs skeletal
-Heart rate, dry mouth vs movement - Movements of body parts
-Limb withdrawal, eyeblink
What type of CC has been proven the easiest to study?
Eyeblink conditioning
How is eyeblink conditioning studied in humans?
UCS = puff of air to the eye
CS = tone
CR and UCR = movement of eyelid
What is the nictitating membrane?
Some animals e.g. cats or birds, have a third eyelid
This nictitating membrane can be clasically conditioned
Why is the nictitating membrane CR preferred over an eyelid CR?
Eyelid responses have interference from voluntary movements like winks and spontaneous activity which don’t affect the nictitating membrane
-NMR is an autonomic response
What is the preferred animal of choice for the NMR ?
Rabbits- easy to train, have a nictitating membrane, responses cause little-to-no distress
What happens in an NMR rabbit experiment?
Typically 100 trials per day for the rabbit
CS (tone) comes on about 0.5s before the UCS (shock/puff of air) but both terminate at the same time
Day 1: response on UCS only
Day 3: Evidence of the nictitating membrane moving before the UCS (somatosensory stimulus) comes on i.e. shows a CR
Day 5: CR clear to see - effect= eyelid closes in anticipation of the US
What are two important features of eyeblink conditioning?
- The UCS overlaps with the CS - this is called ‘delay conditioning’
-if there’s a gap between the end of the CS and the start of the UCS its called ‘trace conditioning’ (invovled more structures so is less simple-not the case for NMR) - The UCS is usually a mild brief shock delivered to skin around the eye (‘periorbital shock’)
-means that closing the eye have no effect on the UCS, this is as it should be because CC is defined in this way- the CR does not affect the UCS otherwise its avoidance learning, a form of instrumental learning
How do we know which parts of the brain are involved?
Anterograde amnesia studies
Accidental finding from HM who had a surgery resulting in the bilateral removal of the temporal lobe – this was successful in stopping his epilepsy but unexpected side effect = his new memories now only lasted for 1 minute
What did patient H.M. demonstrate about STM and LTM?
H.M’s STM was intact but LTMs could not be formed
What type of memory is eyeblink conditioning?
LTM
Does damage to the hippocampus and surrounding tissue prevent eyeblink conditioning?
No
What is the evidence from Weiskrantz and Warrington (1979) that damage to the hippocampus and surrounding tissue does NOT prevent eyeblink conditioning?
Examined patients with anterograde amnesia
Evidence of normal learning, even though the patients could not remember the aparatus that was delivering the UCS
What more recent evidence from Gabrieli et al. (1995) proves that the hippocampus and surrounding tissue does NOT prevent learning?
More recent report
More subjects and control groups
It was verified radiologically that there was damage to the medial temporal lobe
Again, the rates of conditioning were the same in both groups
So we can be sure this type of learning is not due to the hippocampus
Is there more than one type of LTM?
Yes, LTM is divided
Squire et al. (1993)
“major distinction is between conscious memory for facts and events and… nonconscious memory, including skill and habit learning, simple classical conditioning, the phenomenon of priming.. expressed through performance rather than recollection”
What potential brain regions could eyeblink conditioning depend on?
Forebrain (cerebral cortex mainly)
Brainstem, pons and medulla
Cerebellum
What is the evidence against the invovlement of the forebrain?
Delay NMR conditioning still possible in rabbits lacking either hippocampus or cerebral cortex
Also possible in rabbits with forebrain separated from brainstem and cerebellum (‘decerebrate’ preparation)
Mauk & Thompson (1987)
Who researched the brainstem & cerebellum’s involvement?
Thompson’s group
Ref: Thompson (1983)
McCormick & Thompson (1984)
What two techniques were used to examine the involvement of the brainstem and cerebellum?
Electrophysiological recording
Lesions
Explain the electrophysiological recording research methodology. (McCormick & Thompson 1984)
Carried out electrophysiological mapping during conditioning
- they systematically record from multiple units in cerebellar cortex and the deep cerebellar nuclei
Control group- unpaired CS and UCS presented as frequently as in the training condition but they were unrelated
Experimental group- paired CS and UCS
Explain the findings from electrophyiological reocrdings. (McCormick & Thompson (1984)
Recording MUA from the cerebellum
Control group- no response before the UCS but NMR to the shock in the CR
Experimental group- even on day 1 there is a little NMR which is very strong by day 3
Where did McCormick & Thompson (1984) localise the brain acitivty to?
Recorded from the deep cerebellar nuclei close to cerebellum
Found that almost all cerebellar output goes through these deep cerebellar nuclei
What is a limitation of electrophysiological recording?
Recording does not establish a central role- this correlation does not equal causation
What did Thompson (1983) find with lesions?
Large lesions of the entire cerebellum plus the output of the deep cerebellar nuclei
These lesions abolish any previously learned CR and they can never be relearned -> in-activating this part of the brain makes conditioning impossble
So, this area seems to be central to classical conditoning
What did Yeo, Hardiman & Glickstein (1985) find from lesion studies?
Even small lesions confined to the deep cerebellar nuclei produced a similar effect of abolishing the CR
Lesions confined to the anterior portion of the interpositus nucleus are effective - so we can localise the crucial output for CC to this area
What are the important controls from these lesions as discussed in Yeo (1987) that are assessed in behaviour?
After lesion = selective loss of CR but the UCR is not affected
This is important in showing that the lesion does not produce a simple motor deficit- it shows we are not just paralysing the nictitating membrane because it still moves perfectly well
- this is a type of implicit control we get from lesion studies
Lesions do not make the animals deaf (can still hear the tone)- we know they are not deaf because unilateral lesions only affect conditioning to the eyes on the same side of the head , we can still get the CR on the ipsilateral side of the head
What is the conclusion therefore from lesion studies?
That loss of the conditioned response appears to be genuine loss of the memory trace (the ‘engram’)
Effects we have observed are because we have reached the synapses in the brain where the memory trace is stored
How much do we know about where in the brain are the plastic synapses that mediate this conditioning?
Humans- medial temporal lobe damage (hippocampus, amygdala & parahippocampal regions) prevents formation of new long-term memories, but has little effect on eyeblink conditioning
Rabbits- forebrain not needed for delay NMR conditioning
But both electrophysiological and lesion studies implicate the cerebellum
