Neuromodulation of Cognition
Week 1: Dopamine (&introduction)
Aim of this course is to provide mechanistic insights into neuromodulation: more than just
describing, but explaining the neurobiology and cognition as two sides of the same coin.
Goals of the course:
Explain how neurotransmitters shape cognition
Hypothesize the cognitive effects of drugs
Critically evaluate neuromodulation studies
Understand cognitive deficits in neurological disorders
Exam questions (6) require you to explain (in +/- 150 words) for example:
A mechanism related to neuromodulation
Effect of a drug on cognition
Effect of a deficit in neuromodulation on cognition
Use the word BECAUSE!
With each answer, 3 important elements need to be present, each worth ½ a point.
Agonist: activates the receptor it binds to
Antagonist: inhibits the receptor it binds to
Cognitive control: the balance between stability and flexibility.
Dopamine (DA): A neuromodulator. It influences brain processes, but does not really carry
information on its own. It is (among others) implicated in:
Reward, goal proximity, motivation, learning , attention / distractibility, executive
functions/cognitive control
The influence of dopamine on cognition follows an inverted u-shape. Each task has its own
inverted u-curve, with different levels of dopamine being optimal. This makes the story more
complex, because an increase in dopamine can have different effects for different tasks.
Low levels of dopamine:
ADHD
Parkinson’s disease
Medium levels of dopamine = optimal
High levels of dopamine:
Schizophrenia
Drug overdosing
Tonic dopamine: background/baseline
Phasic dopamine: additional release in response to a stimulus/event
Low tonic boosted phasic
High tonic reduced phasic
Presynaptic autoreceptors inhibit dopamine release to regulate the amount of dopamine
present in the synaptic cleft.
Dopamine is produced in the midbrain (substantia nigra, ventral tegmental area) and
projected to the PFC (stability) and the striatum (flexibility). The precursor amino acids of
DA are tyrosin and L-DOPA.
1
,Dopamine receptor subtypes:
D1= D1 + D5
Low affinity/sensitivity for dopamine D1 is dominantly activated when DA levels
are moderate to high
D2 = D2 + D3 + D4
High affinity/sensitivity for dopamine D2 is activated when DA levels are low (or
extremely high)
DA in the PFC
Dual-state theory
State 1: Gate Open
o Related to D2 receptors (D2 agonist decrease GABA activity)
o Multiple network representations (unstable)
o Signal-to-noise ratio is lowered, i.e. there is more noise
o You are more flexible
o It is harder to focus, you are easily distractible (noise
can cause you to switch)
GABA interneurons activity decreases the impact of further excitation and bring on State 2.
State
2: Gate Closed
o Related to D1 receptors
o One dominant network representation (robust)
o High signal-to-noise-ratio, i.e. there is less noise
o You have clear focus on one task, you are very stable
and can maintain tasks despite distractors
o It is difficult to switch to a different task
Neural gain
Throughout the whole course, neural gain is a reoccurring
phenomenon which can be linked to all out topics. The general
idea of neural gain is that high gain makes the contrast
between active and inhibited units more distinct (low signal-to-
noise ratio), while low gain makes the difference between the
two less clear (high signal-to-noise ratio). Because irrelevant
information is suppressed in high neural gain state, this is
reflected in lower brain activity overall.
High neural gain = stability (D1 receptors/state 2)
Low neural gain = flexibility (D2 receptors/state 1)
Cognitive functioning
An increase in D1 activity (D1 agonists) leads to enhanced
focus, good maintenance of rules/choices/goals, but
impairment on switching tasks.
Administration of D1 antagonists lead to predominant D2 activity, which leads to more
random responding, increased distractibility, but no problems in switching or other flexible
tasks.
The effect of dopamine on cognitive functioning has large variability, because it follows
the inverted u-shape. Therefore, DA manipulations will have very different and conflicting
influences on cognitive functions, depending on the optimal values of the implicated brain
2
, areas and tasks. Also, DA manipulation effects interact with baseline dopamine levels, also
resulting in varying effects.
One of the ways in which baseline dopamine levels can vary is polymorphisms in the
COMT gene. This is an enzyme that breaks down dopamine that is present in the synaptic
cleft. People with the Met allele break down faster, and therefore have lower baseline
dopamine levels (only D2 receptors are activated, state 1, flexibility), while people with the
Val allele break down slower, and therefore have higher baseline dopamine levels (also D1
receptors are activated, state 2, stability).
DA in the striatum/ basal ganglia
The striatum has approximately 11x more D2 receptors than the PFC.
The basal ganglia works with a gated mechanism. By default, the gate between sensory
input and the working memory is closed, which allows for maintenance. Phasic DA or
reinforcement learning triggers the gate to open, which allows for updating.
The direct and indirect pathways of how DA in the striatum influences the PFC
Direct pathway
o D1 activity in striatum ↑ Globus pallidus/substantia nigra activity ↓ thalamus
activity ↑ Gate open, PFC activity ↑
o Parkinson’s disease: this pathway deteriorates Rigidity
Indirect pathway (= baseline)
o D2 activity in striatum ↑ External globus pallidus activity ↓ Globus pallidus /
substantia nigra activity ↑ Thalamus activity ↓ Gate closed, PFC activity ↓
o Huntington’s disease: this pathway deteriorates Flexibility/chorea
With higher levels of DA, there is a dominance of D1 and there is greater flexibility in the
PFC.
This mechanism is the opposite of the PFC mechanism, in which flexibility is the default
state, and focus is actively triggered. This contrast shows the overall balance between
stability and flexibility in the brain, sub served by the striatum and the PFC, respectively.
There is also top-down influence from the PFC on the striatum. Glutamate activity from the
PFC can stimulate tonic DA release, dampening the phasic activity (but not in
schizophrenia). Even when the striatum causes great cortical excitability, the PFC can still
prefer stability and the striatal influence will be attenuated.
Reward
When learning about a reward (learning phase), there is a clear spike in DA activity 200ms
after receiving a reward. This spike disappears when the reward is established/expected.
DA codes subjective reward value (not objective), because the value is related to
environment/context/preferences/state/emotion/personality/etc.
The DA response codes prediction error signals. This works with a model-free and model-
based mechanism. There is a general tendency to expect reward after it has happened
(higher DA response). Also, knowledge (such as cues) can influence the prediction that you
make about your reward expectations and therefore also your prediction error. DA response
is very high after a reward following a contextual cue.
Strangely, aversive cues also increase DA response. There are speculations trying to explain
this counterintuitive finding:
It is a homeostasis rebound after a DA decrease
There is a modality overlap between the rewarding and aversive cues (you should try to
make a research design that does not have this overlap to test this)
3
Week 1: Dopamine (&introduction)
Aim of this course is to provide mechanistic insights into neuromodulation: more than just
describing, but explaining the neurobiology and cognition as two sides of the same coin.
Goals of the course:
Explain how neurotransmitters shape cognition
Hypothesize the cognitive effects of drugs
Critically evaluate neuromodulation studies
Understand cognitive deficits in neurological disorders
Exam questions (6) require you to explain (in +/- 150 words) for example:
A mechanism related to neuromodulation
Effect of a drug on cognition
Effect of a deficit in neuromodulation on cognition
Use the word BECAUSE!
With each answer, 3 important elements need to be present, each worth ½ a point.
Agonist: activates the receptor it binds to
Antagonist: inhibits the receptor it binds to
Cognitive control: the balance between stability and flexibility.
Dopamine (DA): A neuromodulator. It influences brain processes, but does not really carry
information on its own. It is (among others) implicated in:
Reward, goal proximity, motivation, learning , attention / distractibility, executive
functions/cognitive control
The influence of dopamine on cognition follows an inverted u-shape. Each task has its own
inverted u-curve, with different levels of dopamine being optimal. This makes the story more
complex, because an increase in dopamine can have different effects for different tasks.
Low levels of dopamine:
ADHD
Parkinson’s disease
Medium levels of dopamine = optimal
High levels of dopamine:
Schizophrenia
Drug overdosing
Tonic dopamine: background/baseline
Phasic dopamine: additional release in response to a stimulus/event
Low tonic boosted phasic
High tonic reduced phasic
Presynaptic autoreceptors inhibit dopamine release to regulate the amount of dopamine
present in the synaptic cleft.
Dopamine is produced in the midbrain (substantia nigra, ventral tegmental area) and
projected to the PFC (stability) and the striatum (flexibility). The precursor amino acids of
DA are tyrosin and L-DOPA.
1
,Dopamine receptor subtypes:
D1= D1 + D5
Low affinity/sensitivity for dopamine D1 is dominantly activated when DA levels
are moderate to high
D2 = D2 + D3 + D4
High affinity/sensitivity for dopamine D2 is activated when DA levels are low (or
extremely high)
DA in the PFC
Dual-state theory
State 1: Gate Open
o Related to D2 receptors (D2 agonist decrease GABA activity)
o Multiple network representations (unstable)
o Signal-to-noise ratio is lowered, i.e. there is more noise
o You are more flexible
o It is harder to focus, you are easily distractible (noise
can cause you to switch)
GABA interneurons activity decreases the impact of further excitation and bring on State 2.
State
2: Gate Closed
o Related to D1 receptors
o One dominant network representation (robust)
o High signal-to-noise-ratio, i.e. there is less noise
o You have clear focus on one task, you are very stable
and can maintain tasks despite distractors
o It is difficult to switch to a different task
Neural gain
Throughout the whole course, neural gain is a reoccurring
phenomenon which can be linked to all out topics. The general
idea of neural gain is that high gain makes the contrast
between active and inhibited units more distinct (low signal-to-
noise ratio), while low gain makes the difference between the
two less clear (high signal-to-noise ratio). Because irrelevant
information is suppressed in high neural gain state, this is
reflected in lower brain activity overall.
High neural gain = stability (D1 receptors/state 2)
Low neural gain = flexibility (D2 receptors/state 1)
Cognitive functioning
An increase in D1 activity (D1 agonists) leads to enhanced
focus, good maintenance of rules/choices/goals, but
impairment on switching tasks.
Administration of D1 antagonists lead to predominant D2 activity, which leads to more
random responding, increased distractibility, but no problems in switching or other flexible
tasks.
The effect of dopamine on cognitive functioning has large variability, because it follows
the inverted u-shape. Therefore, DA manipulations will have very different and conflicting
influences on cognitive functions, depending on the optimal values of the implicated brain
2
, areas and tasks. Also, DA manipulation effects interact with baseline dopamine levels, also
resulting in varying effects.
One of the ways in which baseline dopamine levels can vary is polymorphisms in the
COMT gene. This is an enzyme that breaks down dopamine that is present in the synaptic
cleft. People with the Met allele break down faster, and therefore have lower baseline
dopamine levels (only D2 receptors are activated, state 1, flexibility), while people with the
Val allele break down slower, and therefore have higher baseline dopamine levels (also D1
receptors are activated, state 2, stability).
DA in the striatum/ basal ganglia
The striatum has approximately 11x more D2 receptors than the PFC.
The basal ganglia works with a gated mechanism. By default, the gate between sensory
input and the working memory is closed, which allows for maintenance. Phasic DA or
reinforcement learning triggers the gate to open, which allows for updating.
The direct and indirect pathways of how DA in the striatum influences the PFC
Direct pathway
o D1 activity in striatum ↑ Globus pallidus/substantia nigra activity ↓ thalamus
activity ↑ Gate open, PFC activity ↑
o Parkinson’s disease: this pathway deteriorates Rigidity
Indirect pathway (= baseline)
o D2 activity in striatum ↑ External globus pallidus activity ↓ Globus pallidus /
substantia nigra activity ↑ Thalamus activity ↓ Gate closed, PFC activity ↓
o Huntington’s disease: this pathway deteriorates Flexibility/chorea
With higher levels of DA, there is a dominance of D1 and there is greater flexibility in the
PFC.
This mechanism is the opposite of the PFC mechanism, in which flexibility is the default
state, and focus is actively triggered. This contrast shows the overall balance between
stability and flexibility in the brain, sub served by the striatum and the PFC, respectively.
There is also top-down influence from the PFC on the striatum. Glutamate activity from the
PFC can stimulate tonic DA release, dampening the phasic activity (but not in
schizophrenia). Even when the striatum causes great cortical excitability, the PFC can still
prefer stability and the striatal influence will be attenuated.
Reward
When learning about a reward (learning phase), there is a clear spike in DA activity 200ms
after receiving a reward. This spike disappears when the reward is established/expected.
DA codes subjective reward value (not objective), because the value is related to
environment/context/preferences/state/emotion/personality/etc.
The DA response codes prediction error signals. This works with a model-free and model-
based mechanism. There is a general tendency to expect reward after it has happened
(higher DA response). Also, knowledge (such as cues) can influence the prediction that you
make about your reward expectations and therefore also your prediction error. DA response
is very high after a reward following a contextual cue.
Strangely, aversive cues also increase DA response. There are speculations trying to explain
this counterintuitive finding:
It is a homeostasis rebound after a DA decrease
There is a modality overlap between the rewarding and aversive cues (you should try to
make a research design that does not have this overlap to test this)
3
