Emotion lecture 5

Question 1

Describe a study which attempted to study emotional responses in flies

Answer 1

Trapped flies in a little plexiglass arena and had a paddle attached to a motor wave back and forth over the glass to simulate an overhead sweeping stimulus (aerial predator). This study quantitatively studied the behaviour of the flies after multiple presentations of the stimulus. It is remarkably effective in trigger an innate escape or avoidance response however this could reflect a reflex or something like fear- there needs to be more building blocks to establish emotion.

Question 2

Describe how the building blocks were established in the fly experiment (5)

Answer 2

Scalability: When only a few/ single shadow passes, freezing behaviour is observed. When multiple shadows pass, hopping behaviour is observed.

Valence & Automaticity/ Priority: Shadows disperse flies from food away from the shadow

Persistence: Behavioral response persisted for seconds to minutes

Generalisation: Generalisation to different context (with or without food

Question 3

This study then modelled the behaviour of the fly. Describe this

Answer 3

Constructed a ‘leaky integrator’ model, as if the brain integrates the shadows that pass but it leaks over time. It must reach a certain level to trigger certain behaviours. There are multiple thresholds with different behaviours (i.e freezing > Running > Hopping). The model can also predict the probability of them persisting with their behaviour depending on how many shadows have passed and how much this ‘leaky integrator’ is filled. This is more complex than a reflex.

Question 4

What else does this study find possible evidence for regarding emotions?

Answer 4

An adaptive value of this emotion state:
Fly decision-making in real-life situations: Stay or flee is a life-and-death decision. Cost-benefit analysis through an “emotional state” leaky integrator model

Question 5

Describe the conditioned and unconditioned stimuli in a study which attempted pavlovian conditioning in flies

Answer 5

Either an electric shock (aversive stimuli) was paired with a neutral odor or sucrose (appetitive stimuli) was paired with a neutral odour.

Question 6

Name and describe a neural structure of a fly studied in this research with relevance to the study

Answer 6

A mushroom body cell:

• Kenyon MB cell is activated by odor synapses with MBON cells.
• The US (sucrose/ shock) activates a specific substrate of dopamine (DA) neurons
• A different subset of DA neurons for shock (γ4, γ5) vs. sucrose (γ2, γ3).
• DA facilitates synaptic transmission between kenyon MB cells and MBON (output neurons) cells in specific compartments => behaviour.
• MBONs hardwired to evoke approach (y2, y3) or avoidance (y4, y5). Odor alone too weak to evoke activity.

However if you pair it with a stimulus, then either approach or avoidance is triggered. If it is the shock, then these DA neurons trigger activity in y4 and y5 which are hardwired to trigger avoidance in the fly. The fly will then learn to avoid the odor in future. When the pairing is complete, then the synaptic transmission between the the Kenyon cell of the MB and the MBON is strengthened through this shock: Long term potentiation. At first the CS triggers activation in the specific compartment via dopamine which triggers the action, however after a while, it is facilitated so that the odor is doing it by itself.

Question 7

What does the association of the CS with positive or negative valence depend on? (3)

Answer 7

Same general process, same neurotransmitter (DA) just depends on:

a) Sparse representation of the odor by the Kenyon MB cells (small number of highly odor-specific neurons)
b) Compartment-specific synapses with MBONs that promote either approach or avoidance
c) Strengthening of these Kenyon-MBON synapses by compartment-specific DA release from specific DA neurons activated by shock or sucrose

Question 8

What implication does this pavlovian study in flies have for human neural research

Answer 8

It suggests that we should not associate dopamine with “reward” or a specific emotion or drive, as it has been shown to play a part in both avoidance and approaching behaviour. The circuit connectivity which is hardwired from birth plays a much bigger role in the fly.

Encoding of valence (reward/punishment) a given US (shock vs. sucrose) by DA neurons; Determined by the circuits in which specific DA neurons participate –whether these make synapses with neurons that mediate approach or avoidance. Valence is not encoded merely by the identity of a specific neurotransmitter.

Similar story regarding DA (reward AND punishment) has emerged in rodents and primates!

Question 9

What is required for such an important role for these specific circuits?

Answer 9

An innate basis for such connectivity to ground valence (the wrong response of approach or avoid would not be evolutionarily viable)

Question 10

Describe a study which investigates aggression and courtship in male flies

Answer 10

The structure P1 specific to males mediates both male flies’ aggressive behaviors towards other males, and courtship/mating toward females. Optogenetic stimulation of P1 leads to a persistent state in solitary flies, promoting aggression once a conspecific male is encountered, even 10 minutes later. This gives evidence that aggression/ courtship are not simply reflexes but persistent emotional states

Question 11

In what other species has similar research been done on this link between aggression and courtship?

Answer 11

In male mice mating and aggression controlled by colocalised and anatomically intermingled neurons in the ventrolateral subdivision of the ventromedial hypothalamus (VMHvl) (although flies and mice are separated by 500 million years of evolution!)

Question 12

How well is the amygdala conserved across evolution

Answer 12

Pretty darn well:
Non-mammalian species (reptiles, birds, fish, but not flies) have an amygdala-like brain region with similar circuits and function as the mammalian amygdala

Question 13

What amygdala nuclei are involved in fear conditioning?

Answer 13

Only 3–5 of of 12 amygdala nuclei involved in fear conditioning: Basolateral complex (BLA, including the lateral [LA], cell group), and central nucleus (CeA, containing lateral [CeL] and medial [CeM] subdivisions)

Question 14

What were these two nuclei thought to be involved in traditionally?

Answer 14

Basolateral complex: Aquisition of conditioned fear

Central nucleus: Expression of conditioned fear

Question 15

To what extent is the amydala required/ sufficient for fear?

Answer 15

Amygdala is necessary for acquisition and expression of conditioned fear (but not sufficient for conscious experience of fear)

Question 16

Is the amygdala involved in defensive or appetitive behaviours?

Answer 16

Distinct amygdala sub-populations control defensive vs appetitive behaviors (not visible on fMRIs)

Question 17

To what extent is the amydala required/ sufficient for defensive behaviour?

Answer 17

Not all amygdala: Hypothalamus necessary and sufficient for innate defensive behaviors (freeze, escape, avoidance)

Amygdala is more involved in learned fear, not innate responses.

Question 18

How may the amydala elicit behaviour?

Answer 18

The amygdala is important for the emotion state in fear conditioning and has output to pre-motor areas such as the the Peri-aqueductal grey which then initiates freezing behaviour for example. It has similar connections with other anatomical targets for other behaviour and bodily responses. This is know as the:
Global organismal response: Coordinated behavioral, autonomic, and endocrine response.

Question 19

Describe the classical view of how acquisition and expression of fear conditioning occurs

Answer 19

• Acquisition: Pathways conveying CS (cue) and US (shock) converge in the BLA (or LA), where they strengthen synaptic responses to the CS
• Expression: This information is then relayed to CeA(CE) as the final common pathway for output from the amygdala

Question 20

However what has more recent work indicated regarding the central nucleus?

Answer 20

Indicates distinct functions of CeAsubdivisions and cell types:

• CeL also plays a role in acquisition
• CeM coordinates output via inhibitory projection neurons (by disinhibiting target cells in downstream structures such as PAG) –global coordination
• Distinct CeA neuronal sub-populations controlling defensive vs. appetitive behaviors

Question 21

What evidence led to the claim that the amygdala is involved in both defensive and appetitive behaviour?

Answer 21

Progress thanks to novel neurotechnology (e.g. optogenetics) has revealed seven distinct cell types within different subdivision of CeA: Some involved in appetitive behavior, some in defensive behavior. Complexity of amygdala circuits: Cell-type specific techniques are necessary to identify and manipulate brain function so as not to mix circuit functions

Question 22

What has more recent research found to be true about the inner workings of the Basolateral complex?

Answer 22

There are distinct neuronal subtypes in the BLA:

1. BLA neurons activated by aversive stimuli (e.g. foot shock) project to the CeA, and are necessary and sufficient for fear conditioning
2. BLA neurons activated by rewarding stimuli (e.g. nicotine) project to the NAc and support positive reinforcement

These different types of neurons are interspersed in BLA without any anatomical demarcation (found through marking genetically defined neurons, engaging them in reward/ fear conditioning for hour then killing the mouse and slicing the brain (?)).

Question 23

Has this finding regarding the BLA been found in other species?

Answer 23

Similarly interspersed neurons found in monkeys

Question 24

This research has obvious implications for debates about the role of amygdala in emotion. How does it explain some previous research and what issue does it exemplify?

Answer 24

Some meta analyses found evidence for activation in the amygdala for both negative and positive emotions. This research has some implications for applications of neuroimaging because while it can show activation in this general region, there would be no way to discern from fMRI why there was activation shown in that area for both valences (cannot detect the influence of different cell types.)

This ambiguity in the stimuli in which the amygdala is sensitive to led to the theory that there is a general role of the amygdala in ‘relevance detection.’ But human neuroimaging does not have sufficient cell-type/circuit-level resolution to pick up the fear and reward specific circuits in the amygdala.

Question 25

What are the challenges in studying innate fear? (2)

Answer 25

• Responses to innately threatening stimuli are much more variable and heterogeneous (e.g. risk assessment and avoidance of a predator instead of freezing to a CS)
• Often studied in different modalities than learned fear (predator odors instead of tones) or multimodal (presence of intact predators)

Question 26

What are the key findings regarding innate fear? Where has this brain area been observed before

Answer 26

Requires the ventromedial hypothalamus:
•Ventrolateral portion (VMHvl): Aggression and mating (mice-fly similarities) –neurons express the type 1 estrogen receptor (Esr1)

•Dorsomedial/central portion (VMHdm/c): Innate predator defense –neurons express the transcription factor SF1

Question 27

Describe the dissociation between learned and innate fear in the brain

Answer 27

Fear conditioning involves BLA of the amygdala (plus hippocampus and in bed nucleus of the stria terminalis)

Innate fear involves ventromedial hypothalamus (and different neurons than learned fear in hippocampus and in bed nucleus of the stria terminalis)

Question 28

To what extent are these __ neurons in the VMHdm/c required/ sufficient for defensive behaviours? Name three pieces of evidence

Answer 28

SF1 neurons in VMHdm/c are necessary and sufficient for the expression of innate defensive behaviors:

• Pharmacogenetic silencing reduces defensive responses to a predator
• Optogenetic activation evokes defensive behaviors (through downstream premotor nodes such as PAG) and increases autonomic and endocrine function
• SF1 neurons are activated in the presence of a predator (assessed by c-foslabeling)

Question 29

Do SF1 neurons exhibit markers of an emotion state? Describe the evidence

Answer 29

• Persistent post-stimulation avoidance of activation chamber => Valence and persistence
• Stimulation intensity: Low avoidance, intermediate freezing, high escape => Scalability
• Multi-component effect (behavioral, autonomic, endocrine) => Global coordination
• Interrupts ongoing mating or feeding –> Automaticity/priority

Question 30

Describe patient SM with bilateral lesions of the amygdala (no amygdala basically). What does this demonstrate?

Answer 30

• Intact concept of fear (language, knows that one would feel afraid if chased by a bear etc.)
• No Pavlovian fear conditioning of SCRs (but intact explicit memory of US-CS associations, compare with emotional persistence without explicit memory in hippocampus damage)
• Difficulty recognizing fearful expressions: Because she does not spontaneously attend the eye regions
• Less conscious experience of fear: No reported fear when watching scary film clips, touching snakes or spiders, walking through a “haunted house”
• Experiences fear of suffocating and panic induced by inhaling carbon dioxide (even exaggerated compared to control)

Demonstrates possible role of amygdala in only learned fear

Question 31

What conclusions were drawn from the meta-analysis of 60 years of human intra-cranial electrophysiology?

Answer 31

•Aspects of emotion can be evoked many cortical and subcortical sites
•Some location specificity for emotion (amygdala, insula, ACC, basal ganglia)
•Some specificity for distinct categories, for example
=>Intense sadness in basal ganglia (subthalamic nucleus)
=>Mirth and laughter in supplementary motor area

Question 32

What limitations were there to this meta-analysis? (4)

Answer 32

Very small samples,
extent and location of stimulation poorly controlled,
effects variable
Subject to sampling bias- put there for a reason

Question 33

How can human neuroimaging of emotion focus on very different aspects of emotion ? When are these different aspects useful?

Answer 33

Emotion perception –how people perceive emotional stimuli (without experiencing them)
Stimuli: Facial expressions, body movements, voices, Theory of Mind

Emotion concepts –how people think about emotions (semantic knowledge)
Stimuli: For example, reading stories, categorizing stories. Relevant for the debates between core affect, basic emotion, appraisal theory

Subjective emotion experience –as a proxy for actual emotion states.
Stimuli: Various, e.g. chills from music, shocks, heat, fear conditioning, live tarantula Relevant for interoceptive theories and subcortical basic emotion theories

Question 34

Describe research which examined the emotion concepts aspect of emotion research

Answer 34

Participants were given 10 stories each from 20 different emotion categories (e.g apprehensive, fearful, joyful) and had to categorise these different stories according to questionnaires. One questionnaire was related to core effect where you just rated the story as negative vs positive and mild vs intense, another model was ranking then on the basic emotions, another was one based on appraisal where you ranked the stories based on these 38 higher level concepts (e.g safe, expected etc). Brain recordings were also taken during the study.

They then constructed ‘representational dissimilarity matrix’ on the basis of the 20 different emotion stories and the dimensions recorded by the three conditions. This was then compared to the RDM compiled by the brain activity recorded. It was found that the Appraisal theory dimensions predicted the data better and the Dorso Medial prefrontal cortex was most informative in this.

Question 35

Give a criticism of this emotion concept study

Answer 35

It makes sense thinking about emotions suits appraisal theory best- its essentially circular reasoning and DMPFC would be expected when thinking about emotions.

Question 36

Describe a subjective emotion study which attempted to find evidence for Bud Craig’s theory

Answer 36

Examined the experience of getting strong feelings of chills from music. Subjects were exposed to classical music pieces and clicked a button when they felt ‘chills’. While it is difficult to name what exact emotion is experienced during chills, rCBF-PET activity increases in a number of regions related to reward and body representations, including

• Ventral striatum (VStr)
• Orbitofrontal cortex (Of)
• Midbrain (PAG?)
• Insula (In)
• Anterior cingulate cortex (AC)

Bud Craig used the activation of the insula as evidence for his theories.

Question 37

Timo criticised the fact that many human studies do not track strong emotional states as is possible in animals. Describe a study he described which does track these strong emotions

Answer 37

Subjects were placed in an fMRI scanner and a tarantula was placed in the end in compartments, getting progressively closer. Results showed the following regions track physical proximity of the tarantula (correlated with subjective fear ratings):

• Midbrain (PAG?)
• Dorsal ACC
• Amygdala

Question 38

Describe a study which attempts to dissociate objective stimulation and subjective feeling

Answer 38

Participants had heat applied to their forearms and it was assumed that this noxious input would activate a pathway of nociceptive pain and that this activation would be linked to the intensity of the heat. Participants had to give pain ratings of different temperatures.

ratings were obviously higher for higher temperatures however when they controlled for stimulus intensity and NPS response, Residual variability in subjective ratings explained by activity in additional regions: PFC, insula etc. This demonstrates that there are ways of dissassociating stimulus-driven nociception vs. subjective pain.

Question 39

Summarise the difference in results obtained from emotion concept studies and emotion states studies

Answer 39

• Many studies focus on emotion concepts rather than emotion states: Pattern analyses support appraisal theories (vs. basic emotion or dimensional taxonomies) and invoke dorsomedial prefrontal cortex
• Few studies on intense emotion states in the scanner reveal widespread subcortical and cortical activity, involving midbrain (PAG?), amygdala, ventral striatum, insula, anterior cingulate cortex (remember our tour through the emotional brain)