Skip to main content
European Commission logo print header

Prefrontal and cingulate interactions in cognitive control: reversible inactivation and electrocorticograms

Final Report Summary - REVERSIBLE COGNITION (Prefrontal and cingulate interactions in cognitive control: reversible inactivation and electrocorticograms.)

Introduction

The standout feature of primate cognition is its remarkable flexibility, permitting rapid adaptation to a wide range of situations. Cognitive control is a set of processes that allow behaviour to vary adaptively and efficiently depending on current goals and the environment in which we seek them. Understanding it is fundamental to knowing how our decisions can be adapted and flexible.
Cognitive control is applied and adapted through the process of learning to learn. Organisms begin life able to learn about their environment and hence act more efficiently within it. Only advanced species with prefrontal cortex (PFC) are able to learn to learn more efficiently, gaining efficient, flexible and prospective decision making that goes beyond simple maximisation of immediate rewards.
The project ‘Reversible Cognition’ was designed to reveal and detail the processes of learning to learn and the cognitive control that results, studying related behaviour as well as how it emerges from interactions between a widespread network of structures in the cortex of the brain, including PFC as well as parietal and cingulate cortical areas. Gaining such detailed knowledge about the brain, and in particular intervening in the brain in order to understand how different structures are critical, requires use of an animal model. Reversible cognition used macaque monkeys, animals with highly developed PFC and cognitive control skills. Both the researcher and the host laboratory had extensive experience with this species, and placed an emphasis on refinement of laboratory monkey use throughout the project.

Objectives

The initial scientific objective of the project was to develop and test a complicated task for macaque monkeys that tests the use of cognitive control, and to show how a training strategy focussing on learning to learn can lead to efficient performance of this task. Next the project involved development of a method of longitudinal chronic neurophysiology (ECoG) to be able to record physiological signals from the brain whilst monkeys completed this task. A major objective was therefore to use ECoG to show how different regions of the frontal lobes collaborate to support performance. Finally, these techniques are to be combined with reversible lesions in the monkeys, something not possible in human subjects. The lesions are necessary to reveal the critical contributions of important regions, and to show which elements of the interaction between regions are crucial for task performance. This rare combination of behaviour, neurophysiology and lesions in the same animals is necessary to show how large-scale cortical dynamics reflect and are critical for the flexible and efficient cognitive control that results from learning to learn.
The project also had a number of training and personal development objectives, and in particular was designed to provide the researcher with new expertise from the Host Laboratory in electrophysiological recording techniques. This element of the project has been a significant success, providing the researcher with excellent training, a rare and powerful combination of methods, and the platform to create a significant career in the field. The combination of the researcher’s previous experience with the expertise of the host laboratory has been fruitful for both the researcher and the host laboratory, and the project showcases the benefits of the Marie Curie Fellowship funding model.

Results

We designed a protocol, the Task Set Manipulation task (TSM), to show how monkeys can efficiently switch without being cued between different but related tasks by using task sets. Task sets are sets of rules and protocols specific to each task, and the TSM was inspired by work in human subjects manipulating task sets.
TSM is complicated for monkey and human subjects alike, and an early aim was to understand what processes contributed to the ability to successfully carry out TSM. Therefore we first taught the monkeys a different protocol to the TSM. In the training protocol, the task switches were clearly cued so that the monkey knew when to change task. We then observed the learning of the monkeys, looking for signs of learning to learn.
Right from the start monkeys are able to learn the tasks to a behavioural criterion, so a naïve monkey is very much able to learn in this sort of protocol. But after working over a period of months, monkeys vastly improved their learning rates for each new task. This phenomenon is referred to as learning set, and it represents a strategy that allows efficient switching between different problems, enhancing behavioural flexibility. In an important result, we showed that monkeys could use this learning set to transfer without cost from the training protocol to the full TSM. TSM is a very different protocol, and ostensibly much harder. Successful transfer to the TSM demonstrates the flexibility of cognitive control provided by the learning set, and represents a dramatic advance on previous research on the transfer of learning sets. Monkeys having never encountered a protocol requiring them to learn and relearn multiple task sets with being told when to switch showed extremely proficient performance thanks to their previously acquired learning set. This powerful result strongly suggests that learning-set-like phenomena are at the heart of the acquisition of flexible cognitive control abilities necessary to manipulate task sets.
We used these results to study the processes in the brain necessary to apply flexible and efficient cognitive control. Monkeys were implanted with a grid of electrocorticography (ECoG) electrodes to provide recordings from the surface of the brain that allow us to monitor activity from many different regions of the brain at the same times. ECoG recordings provide information about the activity of large groups of neurons. They provide in particular two types of information. First they tell us about oscillations in the local field around each electrode. These field oscillations are the focus of intense research at the moment, and are increasingly being proposed as a communication method within the brain. Our recordings revealed oscillations at the beta frequency (20-30Hz) during the task in the PFC and parietal regions of the brain. In other work we have already linked these oscillations to a range of control mechanisms, and we are currently working on the precise role of these oscillations in cognitive control and learning to learn. A second signal derived from ECoG recordings is the event related potential. This is the summation of local population activity, and can provide information about the overall role of the recorded part of cortex. We recorded such potentials in response to the feedback that told the monkey whether each trial was correct or incorrect. In the medial part of frontal cortex, these feedback potentials were able to differentiate between correct and incorrect feedback. Both of these results from ECoG recordings show how the brain is reflecting information necessary for cognitive control. How these signals interact between brain areas to produce the specific elements of the TSM remains the subject of the on-going project, which is funded beyond the period of Marie Curie financing. In the crucial final phase, we will combine ECoG and behaviour with reversible lesions to targeted portions of our network of interest, to understand how individual regions contribute critically to the action of the network.

Impact and Societal Implications

In humans, alterations in cognitive control are thought to make a critical contribution to the development of a number of conditions that are currently only partly understood. Further, the process of learning to learn must be crucial to improving education and training, and to helping patients recover from serious brain injury or disease. Deficits in cognitive control and learning to learn are associated with a number of psychological and neurological conditions, including schizophrenia, Alzheimer’s disease, and Parkinson’s disease. There are therefore numerous health and societal benefits to understanding this process, and ‘Reversible Cognition’ has and will contribute to this understanding in numerous ways. In particular, certain important questions, such as the critical contribution of individual brain areas, can only be answered using an invasive approach that demands use of an animal model.
The project has already inspired much follow-up research. In particular it is hoped that the data from this project can be applied in humans in a comparative approach. Relating the results of highly controlled monkey lesion studies to non-experimentally controlled brain damage in human subjects is an important part of the use (and justification) of the monkey approach, specifically because human brain damage is rarely controlled, and difficult to interpret without direct comparators in a controlled model. The final phase of this project, using controlled reversible lesions, will particularly help in this effort, and our laboratory is currently developing collaborations with world-renowned neurosurgeons in order to investigate the clinical value of such an approach.
When such direct comparisons are made, it is possible to test a number of strategies, treatments, or interventions in the monkeys that might permit them to resist the effect of the lesions, but would be ethically problematic to try with human patients. Interventions that prove successful might then be more justified in the human subjects. This is a perspective of ‘Reversible Cognition’, and an element that will be put in place with continuing funding of the project.