Periodic Reporting for period 5 - NEODYNE (Decision making: from neurochemical mechanisms to network dynamics to behaviour)
Berichtszeitraum: 2024-12-01 bis 2025-11-30
Animals, including humans, are constantly engaged in making decisions, possibly, often without knowing about it. For instance, in a rich visual scene (like when driving through a busy street), our mind will inevitable select certain parts of the scene to which we will attend (and possibly remember later) whilst ignoring a vast majority of the remaining visual input. Alternatively, consider approaching someone in foggy weather, and you have decide whether this is a stranger or someone you know and would want to greet. Such kinds of choices are often referred to as perceptual decision making. Other decisions that come to mind are selecting a restaurant for dinner, a holiday destination, a career path, or whom to marry. Such decisions are referred to as value- or reward-guided, because we decide between the options based on how valuable ("rewarding") they appear to us.
The key parameters for decision making are thought to be carried in neuronal oscillations, rhythmic fluctuations in the activitiy of populations of neurons. Neuronal oscillations are thought be a key mode of information processing in the brain. While there have now been a number of studies investigating the relationship between neuronal oscillations and decision making, it is still rather unknown how the activity of neurotransmitters (chemical messengers between neurons) at the level of synapses translates into oscillations in ensembles of neurons and ultimately into decision making. These represent two different levels of investigation: neurotransmitter effects are usually assessed with respect to their effects at the synaptic or cellular level, while network dynamics (neuronal oscillations) are related to decision behaviour. Notably, a number of psychiatric disorders, such as major depressive disorders are characterized both by anomalies in neuronal oscillations, aberrant neurochemistry, and impairments in decision making. The impairments in decision making may even be a factor that supports persistence of a disorder, when individuals engage in self-defeating behavioural patterns. Notably, the neurochemical systems implicated in these disorders, and the targets of the drugs used for treatment are also heavily implicated in decision making. It is therefore important to understand how neurochemical activity translates into larger-scale dynamics (rhythmic activity of populations of thousands of neurons) to guide decision making.
This project will use non-invasive brain imaging (magnetoencephalography, MEG) to study neuronal oscillations in healthy humans while they engage in different decisions. This will allow us to investigate how the relevant parameters underlying a decision are represented in large-scale neuronal dynamics. To relate this to cortical neurochemistry, the activity of neurochemical systems will either be measured using magnetic resonance spectroscopy or manipulated by using drugs that target the neurochemical systems of interest. The wiring of brain micocircuits and their physiological properties turn them into a dynamical system, the behaviour of which can be challenging to predict intuitively. Therefore, neurochemical effects on large-scale dynamics and on behaviour are often hard to understand without an appropriate theoretical framework. To bridge the levels of investigation, biophysically detailed models of cortical circuits will be used. Variants of these models have been successfully used to understand neural dynamics and choice behaviour. The models can be used to simulate the effects of neurochemical manipulations to predict the effects at the level of large-scale dynamics and on behaviour. Thus, the aim of the project is to obtain a mechanistic understanding of the processes that translate the action of a neurotransmitter at a particular receptor into large-scale dynamics that carries information about choice-relevant parameters, and ultimately into the choice patterns we can observe in our human volunteers.
The key parameters for decision making are thought to be carried in neuronal oscillations, rhythmic fluctuations in the activitiy of populations of neurons. Neuronal oscillations are thought be a key mode of information processing in the brain. While there have now been a number of studies investigating the relationship between neuronal oscillations and decision making, it is still rather unknown how the activity of neurotransmitters (chemical messengers between neurons) at the level of synapses translates into oscillations in ensembles of neurons and ultimately into decision making. These represent two different levels of investigation: neurotransmitter effects are usually assessed with respect to their effects at the synaptic or cellular level, while network dynamics (neuronal oscillations) are related to decision behaviour. Notably, a number of psychiatric disorders, such as major depressive disorders are characterized both by anomalies in neuronal oscillations, aberrant neurochemistry, and impairments in decision making. The impairments in decision making may even be a factor that supports persistence of a disorder, when individuals engage in self-defeating behavioural patterns. Notably, the neurochemical systems implicated in these disorders, and the targets of the drugs used for treatment are also heavily implicated in decision making. It is therefore important to understand how neurochemical activity translates into larger-scale dynamics (rhythmic activity of populations of thousands of neurons) to guide decision making.
This project will use non-invasive brain imaging (magnetoencephalography, MEG) to study neuronal oscillations in healthy humans while they engage in different decisions. This will allow us to investigate how the relevant parameters underlying a decision are represented in large-scale neuronal dynamics. To relate this to cortical neurochemistry, the activity of neurochemical systems will either be measured using magnetic resonance spectroscopy or manipulated by using drugs that target the neurochemical systems of interest. The wiring of brain micocircuits and their physiological properties turn them into a dynamical system, the behaviour of which can be challenging to predict intuitively. Therefore, neurochemical effects on large-scale dynamics and on behaviour are often hard to understand without an appropriate theoretical framework. To bridge the levels of investigation, biophysically detailed models of cortical circuits will be used. Variants of these models have been successfully used to understand neural dynamics and choice behaviour. The models can be used to simulate the effects of neurochemical manipulations to predict the effects at the level of large-scale dynamics and on behaviour. Thus, the aim of the project is to obtain a mechanistic understanding of the processes that translate the action of a neurotransmitter at a particular receptor into large-scale dynamics that carries information about choice-relevant parameters, and ultimately into the choice patterns we can observe in our human volunteers.
In a set of studies, we have recorded cortical activity in healthy humans using MEG (Magnetoencephalography). Participants were recorded with MEG under the influence of different drugs that targeted particular neurochemical systems, in particular acetylcholine, GABA, glutamate, and noradrenaline. We compared the effects of these drugs to placebo. Cortical activity under the influence of drugs was recorded both in the resting state, to investigate basic physiological effects on cortical dynamics and while participants performed different tasks probing decision making and learning. Blocking cholinergic transmission at muscarinic receptors with biperiden affected neural oscillations recorded at rest. Notably, biperiden did not affect reward-guided decision making when all choice attributes were explicitly available to participants, but it impaired the appropriate calibration of learning rates for updating estimates of the true underlying reward probability in an uncertain environment. This was accompanied by an elimination of the crucial update signal in high-beta power in lateral prefrontal cortex.
An early version of these results is available as a preprint:
https://www.biorxiv.org/content/10.1101/2024.09.20.614105v1(öffnet in neuem Fenster)
We are currently finalizing the manuscript for publication in a peer-reviewed journal.
Enhancing GABAergic activity prolonged neuronal timescales at rest, with a tendency toward stronger effects in the frontal cortex. Dynamic network analyses revealed that GABAergic effects were most pronounced in the frontal default mode and dorsal attention networks. In both networks, increased GABAergic signaling markedly prolonged timescales. Modulation of NMDA glutamate receptor activity produced no significant changes on cortical activity at rest. Increasing NMDA transmission affected both perceptual and reward-guided decision making, with choices being more strongly influenced by the available sensory and reward information, respectively. No effects on learning were observed. Increasing GABAergic transmission decreased the effect of reward information on decision making and reduced the use of explicitly available information during reward-based learning. We are still analyzing the MEG data for this larger set of studies and hope to submit a manuscript for publication by the end of the year.
An early version of these results is available as a preprint:
https://www.biorxiv.org/content/10.1101/2024.09.20.614105v1(öffnet in neuem Fenster)
We are currently finalizing the manuscript for publication in a peer-reviewed journal.
Enhancing GABAergic activity prolonged neuronal timescales at rest, with a tendency toward stronger effects in the frontal cortex. Dynamic network analyses revealed that GABAergic effects were most pronounced in the frontal default mode and dorsal attention networks. In both networks, increased GABAergic signaling markedly prolonged timescales. Modulation of NMDA glutamate receptor activity produced no significant changes on cortical activity at rest. Increasing NMDA transmission affected both perceptual and reward-guided decision making, with choices being more strongly influenced by the available sensory and reward information, respectively. No effects on learning were observed. Increasing GABAergic transmission decreased the effect of reward information on decision making and reduced the use of explicitly available information during reward-based learning. We are still analyzing the MEG data for this larger set of studies and hope to submit a manuscript for publication by the end of the year.
Combining non-invasive recordings of cortical dynamics in humans with neurochemical approaches and algorithmic and biophysical modelling. Together with our modelling collaborators, existing cortical network models are being revised and extended to incorporate novel theoretical and experimental findings. This will allow us to even more precisely take into account the underlying biological complexities at the microcircuit level. We are currently in the process of testing these models against human behaviour. Once data acquisition for the pharmaco-MEG studies (see above) is completed, these cortical network models will also be applied to the neurochemical effects on behaviour and neural activity. If successful, this work will result in a detailed understanding of how cortical circuits transform task-relevant input into a decision, and how this is governed by neurotransmitters at the synaptic level. In the end, the hope is that this will also contribute to understanding why and how these processes (impaired decision making and abnormal neural oscillations) go awry in disease.