Our next presentation is entitled Single Neuronal Basis for Interactive Social Behavior in a Primate Model. Amy Wang presenter and our discussant is Joseph Nemat. Hi everyone. Thank you so much for this opportunity to share some of my work with you. My name is Amy Wang and I'm a rising fourth-year med student working with Dr. Zeve Williams at Mass General Hospital and we're interested in the neuronal basis for interactive social behavior. Clinically, we're interested in this topic because there are several neurocognitive disorders that impact social functioning. For example, autism spectrum disorder, depression, and schizophrenia. And these disorders, as you know, can be quite treatment refractory. So to move towards more targeted therapies, it's important to identify the neuronal mechanisms behind the dysfunction. However, the neuronal basis for social behavior in general is not very well understood. While we've identified many brain areas implicated in the social brain, as you can see, our understanding remains quite general, especially compared to our thorough characterization of, for example, the basal ganglia circuitry. This knowledge has really allowed for the implementation of neuromodulatory procedures such as deep brain stimulation and helped a lot of patients with movement disorders. One of the major limitations in social neurobiology is the inability to study these neuronal processes in humans in real time at a fine resolution. For this reason, we run our social tasks and perform neuronal recordings in rhesus macaques, who in the wild live in large colonies and form long-lasting friendships. This image really captures the specific aspects of social behavior that we're interested in. First of all, reciprocity. These two animals are grooming the third, perhaps with the hopes that the third will return the favor. We're also interested in group behavior, which brings additional complexity to social behavior. In groups, animals need to cooperate and avoid conflict to reach mutually favorable goals, and this requires the ability to predict another agent's intentions and actions. So we believe that studying group behavior will allow us to dissociate and study the core components of social behavior. To bring this dynamic into the laboratory, we devised a structured social task. We built an apparatus, which basically looks like a small table with a rotating top, to allow the animals to deliver rewards to each other over multiple trials. You're looking at an overhead view of the setup, and in each trial, one animal is designated actor. Once the food reward is loaded, the handle is positioned in front of the actor and locked in place. The trial is initiated by the press of a pedal, which unlocks the lock mechanism, allowing the actor to turn the handle freely. And then the actor makes a decision, and the recipient animal, in this case monkey B, reaches for reward, thus concluding the trial. Now, in a future iteration, perhaps the handle goes to monkey B. So the question here is, does monkey B care that monkey A at one point rewarded him, or is he going to give to his friend monkey C? Now, this basic interaction is central to economic game theory and social decision‑making theory, and these types of interactions are important for understanding normal and atypical social development. So first we looked at behavior. What did the animals do? The primates showed preferences for specific individuals. This is an overview of reward allocation in a single session. And as you can see, monkey A and B demonstrated strong preference for each other, less so for C. On a trial by trial basis, each animal's preferences fluctuated between the other two with locally sustained patches of preference. Zooming out, we see that each animal's preferences varied from a session to session basis and by differing amounts. So the size of the bubble shows the degree of preference. Next, we looked at strategic behavior, such as tit for tat. Tit for tat is a strategy basically mirroring the opponent's actions. So if you give to me, then I'll give back to you, but if you don't give to me, then I'm not going to give back to you. And in the same example session, you can see that monkey B gave back to A 76% of the time and monkey A, likewise, gave back to B almost 60% of the time. In the gray is a chance distribution of probability of enacting the tit for tat strategy. And in this session, monkey B used the strategy significantly higher than would be predicted by chance. Now, the behavior is very robust and honestly quite fun to watch. But a major question in our lab is how do we get from this level of understanding incrementally to this level of understanding? We recorded from the dorsal anterior cingulate during task performance. We chose this area because it's been implicated in social behavioral disorders such as autism and prior studies using functional imaging as well as single unit recordings have demonstrated its role in encoding social features. So during each session, we recorded from a single animal. We found three main neuronal responses in the cingulate. This plot is a parastimulus histogram showing firing rate and below it, this raster plot is stratified by which animal was reaching for reward. So our first response is increased activity when the recorded animal himself reached for reward. So self encoding. Our second response is increased activity when either of the other two animals reached for reward. So general other encoding. Another way to conceptualize this is non‑self. So anyone but self. And the third is very interesting to us. It's increased activity when a specific other animal reaches for reward. So in this plot, it's monkey A shown in the blue. Why is this neuron important? In order to maintain ‑‑ in order to demonstrate social preference or act on your social strategies, it's necessary to maintain a mental representation of specific other individuals. So I can't possibly reciprocate if I don't know who is helping me out in the first place. This neuron represents that basic social variable. Finally, we looked at the distribution of these three responses at the population level. We recorded from a total of 292 neurons and of these, 120 or 46% were significantly more active than the other neurons. And they were significantly modulated by recipient identity. You can see the distinct sub‑populations here. So our specific other encoding neurons are shown in the purple. General other in green. And there's very little overlap between cells that encode self versus encode others. And we're looking at how these distinct signals come together to generate more complex social behavior. So in summary, in our study, we found that the primate model, there's still a lot of behavioral controls we need to look at. However, we have strong evidence that the animals act specifically towards unique individuals. And we have demonstrated a population of neurons in the dorsal anterior cingulate which encode information about these specific individuals. We're currently analyzing the role of this fundamental signal in social computations such as, of course, preference and reciprocity. To further investigate the role of the cingulate within the social circuitry, in the coming weeks, we're applying deep brain stimulation to the area to assess our ability to modulate group interactive behavior. Thank you all for your time. And not so fun to study social neuroscience by yourself, I suppose. So thank you to the lab, especially Dr. Williams and my post-doc mentor, Dr. Baez. Thank you very much. It's a real pleasure to have been asked to speak on this paper. What the authors have done here, I think, is really tremendous and very innovative work. And I was excited to be able to read it and think about it. This is unique, to do this kind of an experimentation not just on a single primate, but on three primates acting socially around a table with spinning food, I think is a heroic task that really required quite a bit of preparation, training, and a very thoughtful design. It brings up a very important topic. The social brain hypothesis has demonstrated increasingly that the growth of the primate neocortex is not related necessarily to increasing environmental challenges that primates were facing, but rather to increasing complexity of social interactions. If we look across primate groups, what we see is that the size of the neocortex and its ratio to the size of the medulla is directly correlated with the complexity of the social interactions that each species engages in. With humans here, you can see the square at the top. Despite this, in doing primate research, most of the studies that we've undertaken have excerpted very simple perceptual or motor tasks and done these in isolation out of the social context in which they are typically practiced. The reason for this is clear. To show you a very simple example from my lab, this is the task that we typically use, the Simon task, and there are two potential actions in a setting of three or four potential contexts. That's a tremendously simple and straightforward task that allows it to be easily studied. By comparison, this example from their work, there are three separate primates being recorded from. There are, by my calculation, 24 potential actions in different contexts. That's a tremendously complicated and thoughtful way to approach social interaction. The innovation of their design is being able to pare this down into individual actions and important elements to study which are independently discernible and meaningful. I think that's really a tremendous task they've undertaken. Thus far, I won't repeat their findings, but it's demonstrated that individual neurons in cingulate cortex are able to identify specific other individuals. That's an important correlation. The work that they've suggested doing DBS stimulation at the same time potentially can show that these are necessary correlations between cingulate activity and identification and response in social interaction. I think that promises both to be an important research finding, which I'll be very interested to follow, and also potentially an important therapy for things like depression and autism, which are important diseases. Finally, since imitation is the greatest form of flattery, I want to report to you guys that I immediately after reading this work, went and repeated the experiment on the only primate colony that I have access to, which is this. Thus far, results have been mixed, but we're continuing to do this work. So thank you very much for really fantastic work. Thank you.

neurocognitive disorders

rhesus macaques

structured social task

neuronal recordings

deep brain stimulation