Emotional memory

Question 1

Little Albert study

Answer 1

After 7 exposures to white rat (which previously elicited zero fear response) paired with hitting a metal bar (which caused checked breathing, arms raised, startle), presentation of rat alone caused Albert to cry and crawl away.
5 days later, they presented him with play blocks - no fear. Rats and rabbit - full fear response
Santa claus mask, dog, fur coat, cotton wool - milder fear response
He was then removed from hospital, so they couldn’t test attempts to remove the fear.

Question 2

Pavlovian conventional nomenclature

Answer 2

Neutral stimulus can cause orientation but not the unconditioned response (salivation).
After conditioning, the neutral stimulus becomes the conditioned stimulus.
Once the conditioned stimulus elicits the full UR, it is said to have replaced the US. This is ‘stimulus substitution’.
The CS-US stimulus substitution confers info about the sensory qualities of the US, ‘US value’.
The CS-UR association, on the other hand, is a simple stimulus-response and confers no info about the US identity.
CS-UR - CeN mediated, resistant to reinforcer (US) devaluation, cannot support second order conditioning or instrumental responding.
CS-US - BLA mediated, sensitive to reinforcer devaluation, supports second order conditioning and instrumental responding.

Question 3

Translationality

Answer 3

The ‘central fear state’ elicited in animals shares many of the same physiological symptoms to the human fear response - SCR, increased heart and ventilation rate, cessation of movement.

Question 4

Experimental methods

Answer 4

LeDoux’s cued fear conditioning (measures fear. Problems - the freezing response is not quite the UR, bc rats don’t freeze to footshock)
Davis’s potentiated startle (measures anxiety)

Question 5

The search for the neurobiological basis of fear

Answer 5

LeDoux 1985 - lesioning auditory cortex of rat had no effect on auditory fear conditioning. Lesioning thalamus and midbrain (both upstream) did. So where does auditory thalamus project to other than auditory cortex? LeDoux used tracing to identify subcortical sites, lesioned them in turn. The only one that affected conditioning was the amygdala.

Question 6

CeN amygdala projections and their functions

Answer 6

lateral hypothalamus—>sympathetic activation
Dorsal motor nucleus of vagus —> parasym activation (ulcers, urination)
VTA —> DA mediated behavioural arousal
PAG —> freezing
nucleus reticularis pontis caudalis—>augmented startle response

Kapp 1979 - lesioning CeN disrupted heart rate increases in auditory fear conditioning in rabbits.

Question 7

BLA functions

Answer 7

LeDoux lesions - impaired fear conditioning
Davis lesions - impaired potentiated startle response

So BLA responsible for sensory integration and computation.

Question 8

Redundant auditory fear pathways - what are they?

Answer 8

From medial geniculate nucleus of thalamus to

a) auditory cortex, then auditory association cortex, then lateral nucleus amygdala
b) straight to lateral nucleus amygdala

Same thing occurs for visual conditioning - superior colliculus, to pulvinar nucleus, to BLA.

Question 9

Redundant auditory fear pathways - how were they discovered?

Answer 9

Romanowski and LeDoux 1992 - lesioning just cortex spares conditioning to a single tone. Severing just connections from thalamus to amygdala spares conditioning to a single tone. So two separate pathways.
Jarrell et al 1987 - conditioned to two tones, one paired with footshock, the other not. Lesioning the cortex made animals freeze to both tones - so the discrimination was gone, but the tone-fear association was still there.

Question 10

Redundant auditory fear pathways - why are they useful?

Answer 10

• optimising transmission speeds
• safety - if one fails, we’ll still respond to threats
• thalamo-cortico-amygdala pathway required for discrimination, thalamo-amygdala pathway for quick response
• representational accuracy - thalamo-amygdala connections come from associated, unaligned nuclei in the thalamus such as PIN [posterior intralaminar nucleus]. Projections to the cortex come from these associated unaligned nuclei AND from main sensory nuclei i.e. ventral medial geniculate. So the thalamo-amygdala system has a limited processing capacity in comparison.

[the main sensory vs associated sensory unaligned dissociation was proposed by Winter and Morest 1983, and supported by Romanowski and LeDoux]

Question 11

Hippocampus and contextual conditioning

Answer 11

Lesions of hippocampus impair freezing to context, but not to CS. Amygdala lesions impair both.
COntroversy though:
- is the hippocampus just required for performance of the UR? It does induce hyperactivity…
-some lesions produce retrograde but not anterograde amnesia (i.e. pretraining lesions don’t impair conditioning)
-lesioning the entire hippocampus can impair freezing to context /and/ CS.
-McNish et al 1997 - lesioning hippocampus does not impair fear-potentiated startle
-McNish et al - lesioning hippocampus does not disrupt contextual blocking
Blanchard et al 1977 - Hippocampal lesions impaired unconditioned freezing (in response to a cat)

Question 12

Intercalated cell mass

Answer 12

GABA-ergic cells between lateral and central nucleus. Medial ICM cells tonically inhibit CeN. LA causes lateral ICM cells to release GABA, onto the medial portion of ICM. This inhibits their inhibition, disinhibiting the CeN.

Question 13

Extinction - neurological basis

Answer 13

VMPFC lesions impair extinction of cue-conditioning, but not contextual conditioning. [In fact, they impair recall of extinction - rats extinguish as normal, but the CS-no-US association is forgotten]
In humans, VMPFC thickness and activity is correlated with fear extinction
In rats, infralimbic bursting predicts recall of extinction two days later
Klavir et al 2012 - LFS inactivated dACC in rhesus monkeys, found better recall of extinction.
PFC projects onto extinction neurons in amygdala, which inhibit CeN from driving into fear state.
hippocampus inhibits these extinction neurons, so disinhibits CeN neurons.
PFC promotes extinction, hippocampus prevents it.

Question 14

Amygdala and fear in humans

Answer 14

Bechara et al 1995 - patients with amygdala damage learn a CS-US pairing but do not fear condition to it. Note that these patients were quite heterogenous though.
Delgado et al 2006 - fMRI studies have found amygdala activity in instructed fear conditioning (i.e. no actual shocks delivered, just told that a stimulus would predict a shock)

Question 15

Serial model of amygdala organisation - what is it?

Answer 15

Inputs to BLA, BLA processes and performs relevant computations (CS-US), BLA modulates CeN activity to generate output, CeN generates central fear state

Question 16

Serial model of amygdala organisation - which 3 lines of evidence are against it?

Answer 16

1. functional dissociations between amygdala subnuclei (killcross et al, BLA needed for conditioned punishment, CeN needed for conditioned suppression)
2. amygdala involvement in fear-unrelated functions (CeN lesions disrupt autoshaping)
3. involvement of structures beyond the original circuit in ‘emotional learning’. (NAc and aCC needed for autoshaping)

Question 17

Shaping and autoshaping

Answer 17

Shaping is where you reward behaviours increasingly close to the desired one until it’s reached
Bussey et al 1997 - rodent autoshaping - they start to approach the predictive light (the CS), you increase the delay between CS and US (food), until you omit US altogether. Rodents still licked the CS. aCC lesioned rats didn’t get the discriminated response - licked the CS- more than controls. mPFC lesioned rats had longer approach latencies.
Parkinson et al 2000 - Autoshaping is dependent on the CeN, not the BLA. This and the resistance to omission of US show it’s a pavlovian contingency. Also dependent on the core of the NAc.

Question 18

Second order conditioning

Answer 18

Eventually, a CS2 can substitute for a CS1 and produce the UR.
Hatfield et al 1996 - CeN lesions have no effect, but BLA lesioned animals cannot do this, i.e. cannot access the value of the US when given the CS2. This indicates that the CS1 never truly substituted for the US, and instead only a pavlovian CS1-UR association was built.

Question 19

If BLA only mediates CS-US, why is it needed for basic Pavlovian fear conditioning?

Answer 19

Because in fear conditioning, UR is not the same as CR. That is, footshock causes rat to run around and climb the walls, CS presentation causes freezing. This requires a CS-US association, because the freezing itself has value.

The CeN, then, is only required because it outputs to the PAG to actually mediate the freezing itself.

This is the serial model again. Useful for auditory conditioning, where CR is freezing. Less useful for other fear conditioning paradigms, where a parallel organisation may be more useful.

Question 20

Conditioned motivation

Answer 20

Umbrella term that includes conditioned suppression and pavlovian-instrumental transfer (basically just the opposite valence). BLA lesions impair specific conditioned motivation, CeN lesions impair general conditioned motivation.
General conditioned motivation is thought to be due to generalised excitation and facilitatory effect on behaviour produced by an appetitive CS.
Specific conditioned motivation is thought to be because presentation of the CS reactivates the CS-US memory, which reactivates the response-reinforcer association.
BLA lesioned animals cannot respond to conditioned reinforcers, when measured using second-order schedules or Acquisition of New Response, or using conditioned punishment.

Question 21

CeN - role in salience of stimuli and apportioning attention

Answer 21

1-lesions reduce orienting to a pavlovian CS paired with food
2-lesions increase exploration in an elevated plus maze, suggesting reduced attention to anxiogenic effect of being in the open arms.
3- impaired performance on 3-choice serial reaction time task, suggesting impaired attentional processing.

Put rats in a two-compartment cage, get them to shuttle between the two to avoid footshocks. This conditioned avoidance behaviour competes with the conditioned fear response (freezing). Rats will naturally fall on a spectrum between the two behaviours. If you take ‘poor avoiders’ that just freeze and lesion the CeN, you turn them into ‘shuttlers’. So the CeN was either involved in the general motivational value for the footshock or in the habitual behaviour brought about by the US, allowing sensory-specific, instrumental, BLA derived response (shuttling) to dominate.
BUT remember the CeN is needed for freezing behaviour…