What is everybody doing? Social prediction, categorization, and monitoring in the Prefrontal Cortex of the Macaque adopting a new human-monkey (H-M) interactive paradigm.

Project Information

HUMO

Grant agreement ID: 648734

DOI

10.3030/648734

Project closed

EC signature date 7 December 2015

Start date 1 March 2016

End date 31 August 2021

Funded under

• EXCELLENT SCIENCE - European Research Council (ERC)

Total cost

€ 1 028 750,00

EU contribution

€ 1 028 750,00

Coordinated by

UNIVERSITA DEGLI STUDI DI ROMA LA SAPIENZA
Italy

Periodic Reporting for period 4 - HUMO (What is everybody doing? Social prediction, categorization, and monitoring in the Prefrontal Cortex of the Macaque adopting a new human-monkey (H-M) interactive paradigm.)

Reporting period: 2020-09-01 to 2021-08-31

Summary of the context and overall objectives of the project

Primates live in social groups in which they face cognitive problems that they must learn to deal with if they want to survive and procreate. They need to understand the meaning of other individuals’ behavioral signs and they need to learn and maintain in memory the knowledge of their previous interactions with their group mates. Why is it important to study the Frontal Pole cortex (FPC), area 10, in monkeys?
The FPC in humans has a homologous area in other primates, but not in other mammals that means that the frontal pole of non-human primates and humans derives from a common ancestor. In humans, the FPC continues to develop during childhood and adolescence and it has been suggested that its abnormal development might result in a variety of disorders such as autism spectrum disorders, attention deficit and hyperactivity disorder (ADHD), and schizophrenia.
In the first part of our project, we record the activity of the neurons of the FPC in the context of a social interactive task, specifically in a task requiring both the monitoring and the evaluation of the choices of other agents. Elucidating the function of FPC can be crucial to understanding the neural basis of social cognition and also to understand its importance in giving an evolutionary advantage to primates.
We also test the hypothesis that the monitoring role of this area can have a role in learning. Its signal could implement a mechanism for single-trial learning, both for individual learning, and likely also for observation learning. Single-trial learning can be considered a unique primate ability. We also hypothesized that the representation of goals around the time of the feedback of an action could contribute, in the FPC, to create the necessary linkage between learned events and outcome (correct and errors) for generating fast learning in one trial.
This hypothesis is in line with the result of a past neuropsychological study that showed that monkeys with PFC lesions were impaired in the one trial learning typical of object-in-place scenes(OIP) task. In this task, monkeys see a series of images in the background, called scenes. Two letters are presented always in the same place for each scene and many scenes are sequentially presented with different letters. The monkey’s task is to learn which letter of the two presented is correct in each scene. To understand the role of FPC in learning we recorded recording in FPC of two monkeys during an object-in-place scenes learning task to investigate the neural correlate of learning based on single events. We were also interested in the observational learning aspect that is the neural substrate of learning from other individuals as a human agent or a computer.

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

We recorded the neural activity using two main paradigms with chronic arrays, the non-match-to-goal (NMTG) task and the object-in-place (OIP) scenes learning task, recording from the Frontal Pole.
We trained the monkeys to perform a social variant of the non-match-to-goal task (NMTG), designed to study the ability of monkeys to monitor human agent choices and computer choices generated by a cursor. Under experimental control, the monkey and a human agent alternated each other’s roles as actor and observer. As actors, the monkeys had to choose between two peripheral targets (PT) displayed on the two right and left sides of the touchscreen. They had to follow a non-match-to-goal rule, discarding the target chosen in the previous trial for the alternative one. As an observer, they had to observe the human agent or the computer agent in a controlled condition who followed the same rule, and to monitor his/its choice to perform its own future choice correctly. The computer controlled cursor acted similarly to the human agent.
With this paradigm, we found that both during the monkeys' trials and the observation of the human agent FPC neurons represented the spatial goal (right/left) around the end of the movement that is when the choice is made explicit and the goal is accomplished.
By comparing human and computer trials we found that immediately after the movement separated populations of FPC cells were dedicated to the representation of the action goals in the human (social) and the computer (nonsocial) indicating an important difference between the interaction with living and non-living agents. Importantly, the fact that this prefrontal area encoded differently human and computer spatial goals shed light on the role of the area in the representation of other’s goals once these are accomplished. These results were presented in a symposium that organized by Genovesio at the 2020 FENS and by Genovesio at International Symposium on Biology of decision making (SBDM). Nougaret presented the results in a symposium at the SINS meeting (http://www.sins.it/medias/943-final-program.pdf)
Second, we analyzed the sessions from the object in place (OIP) paradigm in which the monkeys learned the task in two modalities: alone or through observation. We found that during the delay period the neuronal activity underwent the largest changes between the first and the second presentation of a given problem. This result can support the hypothesis that FPC is involved in one trial learning. This result shows for the first time a neural substrate of the one-trial learning, a process in which area 10 appears to play a pivotal role, as reported by lesion studies. We found also a learning effect at the level of the direction signal, the direction tuning increased with learning moving between the first presentation of the scene(first run) to the last (last run). These results were presented by Nougaret at the Neurofrance conference and at SFECA conference.
Besides the recording and the electrophysiological analysis, we published five articles. In one article in Behavioral Brain Research (Ferrucci et a., 2019) we compared one-trial learning in the OIP task within different reward contexts. Our results showed that the amount of upcoming reward could speed the learning.
In another second article, we provided evidence that monkeys can show one-trial learning in the OIP paradigm both by observing a human agent and a ghost agent represented by a cursor controlled by a computer.
These two papers paved the way to address the question from a neurophysiological point of view in the Frontal Pole.

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

In the past years, we have presented the evidence that monkeys can learn from human models, provided that the human model received a reward through a vicarious reward mechanism and that monkeys are able to take turns with humans and monitoring the human behavior, we extend now that to the observation of a computer agent. In one previous neurophysiological study, we had shown that the Frontal Pole cortex encoded a monkey’s goals, not when they were established, but much later when the feedback about whether the goal chosen was the correct one arrived. This finding suggested that the FPC plays an important role in monitoring or evaluating the goals of our actions. The primary question to answer in this project was whether the monitoring activity observed in the FPC is specific to the monkey's own goals or applies to others as well. Our task, with an interactive component, was designed to study this process and to address this issue. It allowed us to demonstrate that FPC monitors both self-generated decisions and other agents’ decisions assigning to FPC a social function as well.

humo-summary.jpg