Reward, punishment and the VIP neurons

Wednesday, 11 January, 2023
Tags: News

Researchers at KOKI, Pázmány Péter Catholic University, and the BrainVisionCenter (BVC), founded in 2021 by Botond Roska, Balázs Rózsa, and ITM, in collaboration with colleagues at Washington University in St. Louis, have described a previously unidentified, cortical-wide principle that contributes to the success of early learning.


Perhaps even fashion doesn't change as fast as the thousand-faced science. Competition for funding is also growing, and the standards are getting higher. Even if a discovery is ground-breaking, to be published in the best journals, it must be validated using the latest methods and tools. Collaboration is necessary, because even the best-equipped team, with its many techniques and methodologies, and even the most advanced development teams, may not have enough experts to answer a given question in depth and in a multifaceted way.
The study published in eLife in December 2022 is also the result of a very successful collaboration. The journal's editor recommends it, saying that the authors have used an exceptional imaging technique and that the message of the work is a landmark.
At milestones, it is sometimes useful to pause for a moment and look back. For it is only appropriate to talk about the past when there is a continuation. And the inhibitory interneurons expressing vasoactive intestinal polypeptide (VIP), which will be discussed below, were discovered in our institute

And now, Zoltán Szadai, Ph.D. student, one of the first authors of the paper and member of Balázs Rózsa's Neuronal Networks and Dendritic Activity Research Group, will talk about the work published in eLife. 

-What was the big question you were looking for an answer to?

- Let me start from the beginning! One of the functions of the cerebral cortex is to build models of the external world, which can be used to predict what events we can expect in certain situations and their associated environments. (On this topic, it is worth reading the recent article by Balázs Ujfalussy.) The models should include information that indicates whether something useful or harmful to us could happen in a given situation. While we prefer the former, we would rather avoid the latter. But how such information is incorporated into the models created by the cerebral cortex is not yet understood.

- Similar perfectly understandable questions can lead to experiments that are incomprehensible to the layman!

 - Our experimental setup was very simple. In the published study, we modeled the above situations by investigating the function of a specific inhibitory cell, the vasoactive intestinal polypeptide (VIP) interneuron, during the application of a small reward (water) or a nasty stimulus (air blasts from the animal's eyes). 

- Why these?

- These VIP cells make up only 1% of the cortical cells, so they are a relatively small population, but their activity inhibits the very neurons that keep the main cells of the brain, the pyramidal cells, under constant inhibition. This means that when the VIP cells are active, the pyramidal cells are also activated. This is essential for learning, or in other words, for changing models of the outside world.

- So far, that makes sense. Inhibition of the inhibitor allows activity! But what did the mice have to learn?

- In the experiment, thirsty, or thirsty mice had to distinguish between two sounds. They were rewarded with water if they did so successfully, and an unpleasant gasp as a 'punishment' if they didn't. Both the reward and the punishment triggered strong activation of VIP cells in the entire cortical area. The fact that this happened in the entire cortical area suggests that VIP cells are transmitting organism-level information to cortical micro-networks which is important for survival. These micro-networks then generate new associations that help to remember the complex environment in which the reward or punishment occurred. 
From the results of this theoretically simple experiment, which is a little more difficult to carry out, we conclude that VIP cells create opportunities for the co-activation of pyramidal cell populations in the brain during events that are important to the organism, which facilitates the transformation of internal models.

- This was so clear and understandable that I don't know why you or others didn't realize it sooner!

 - I haven't mentioned that similar signals have been measured in certain brain regions - punishment in the amygdala, reward in the hippocampus, and medial prefrontal cortex - but the response, the activation, in these areas fits well into the presumed functional profile of the brain areas. As many non-brain scientists know, the amygdala has a role in the processing of punishment, and in the case of hippocampal measurements, the fact that a reward received while performing a specific orientation task in the area did indeed trigger neuronal activity could be demonstrated by the fact that random "reward" activated only a small proportion of cells, with the medial prefrontal cortex having a putative function in reward-punishment evaluation. So in neither case was it surprising, and in fact, it was expected, that when performing brain area-specific tasks, some cells in that brain area would be activated. In other cortical areas, such as the somatosensory cortex or the visual cortex, none of the studies to date have shown the activation of VIP interneurons in response to reward, casting doubt on the principle of cortical-wide signaling. 

 - So you have assumed that this cortical-wide signaling must be true for VIP interneurons, but for some reason, others have failed to show it!

- That is exactly right. And not only have we found the signals mediated by VIP cells, but we have also been able to explain why others have not seen them before!

 - Why not?

- It's actually quite simple. Because if the animal has already learned to associate the reward signal - in this case, sound - with the reward, then the VIP interneurons will activate in response to the sound rather than the reward itself.

- It would not have occurred to me, nor to others before you. I suppose it wasn't that anecdotally simple, otherwise, you wouldn't have needed to work with the Adam Kepecs group!

 - That's right. The members of Adam Kepecs' lab at Washington University in St. Louis are very good at studying mouse behavior, while we have the optical tools to study VIP cells, a type of cell with a small cell body and a sparse distribution. Although VIP cells can be studied by electrophysiological methods, their small size and location pattern make them difficult to study. Our three-dimensional two-photon scanning microscopes, on the other hand, can capture activity in large volumes from all cells expressing calcium-binding fluorescent dye molecules introduced by gene manipulation methods. And from deeper regions that are difficult to access with a microscope, we recorded the population-level activity of VIP cells using fiber optics.

- Well-deserved recognition from the editor of eLife! So you have answered this question as you should. But in science, one answer is always the basis for another question!

- In the sequel, we will investigate the pathway by which information about reward and punishment reaches the cortex and VIP interneurons. We are also interested in how different VIP cell subtypes modulate pyramidal cell activity during the performance of different behavioral tasks. 

- What gives this question particular relevance?

- In certain conditions, such as the development of autism and schizophrenia, some theories suggest that the brain's internal models of the external world are disturbed. The balance between the brain's predictive functions, which try to predict the future, and its representational functions, which process stimuli from the outside world, is upset.
Our results may help to design experiments investigating the neurophysiological changes associated with these conditions.


Zoltán Szadai, Hyun-Jae Pi, Quentin Chevy, Katalin Ócsai, Florin D. Albeanu, Balázs Chiovini, Gergely Szalay, Ádám Kepecs, Balázs Rózsa (2022). Cortex-wide response mode of VIP-expressing inhibitory neurons by reward and punishment. eLife




Zoltán Szadai, Hyun-Jae Pi, Quentin Chevy, Katalin Ócsai, Florin D. Albeanu, Balázs Chiovini, Gergely Szalay, Ádám Kepecs, Balázs Rózsa (2022). Cortex-wide response mode of VIP-expressing inhibitory neurons by reward and punishmenteLife, 11. Doi: 10.7554/eLife.78815.