Task and Behavior-Related Variables Are Encoded by the Postrhinal and Medial Entorhinal Cortex During Non-Spatial Associative Learning

Ingeborg Nymoen, Frederik Sebastian Rogge, Ann Margareth Aasen, Sverre Grødem, Anders Malthe‐Sørenssen, Mikkel Elle Lepperød, Torkel Hafting, Marianne Fyhn, Kristian Kinden Lensjø

TL;DR

This study investigates the role of the medial entorhinal cortex (MEC) and postrhinal cortex (POR) in non-spatial associative learning. Researchers found that while both regions adapt their neural activity during learning, MEC exhibits a more organized, anatomically clustered representation of behavior-related information, extending its known role beyond spatial processing.

What Problem Does This Paper Address?

The medial entorhinal cortex (MEC) is famously known for its role in spatial navigation, with its grid cells forming a neural map of an animal's environment. However, its involvement in other forms of learning, particularly non-spatial associative learning, has been less clear. This paper aims to bridge this gap by exploring whether MEC neurons can encode information about learned associations between visual cues and outcomes, and how these representations develop during the learning process. The study also investigates the postrhinal cortex (POR), a region known to integrate visual stimuli with outcomes and which is connected to MEC, to understand the comparative roles of these two brain areas.

Methods

The researchers employed a combination of techniques to record and analyze neural activity in mice during a visual associative learning task.

  • Behavioral Task: Mice were trained on a Go/NoGo visual association task. They learned to associate drifting gratings (visual cues) with different outcomes: a rewarding milkshake, an aversive bitter solution, or no outcome. The task involved learning these associations and then experiencing a reversal where the cue-outcome pairings were changed.
  • Two-Photon Calcium Imaging: This technique was used to record the activity of individual neurons in the MEC and POR. Genetically encoded calcium indicators (GCaMP) were expressed in these neurons, allowing their activity to be visualized as changes in fluorescence.
  • Generalized Linear Model (GLM): A GLM was used to predict neural activity based on various factors, including behavioral variables (licking, speed, brain motion), stimulus types (plus, minus, neutral cues), and outcomes (reward, aversive). This helped to identify which factors were driving neural responses.
  • Agglomerative Clustering: This unsupervised machine learning technique was applied to group neurons with similar response patterns (tuning profiles) based on their GLM weights. This allowed for the identification of functional cell clusters.
  • Support Vector Machine (SVM) Classifier: SVMs were used to decode visual information from the stimulus-related clusters, assessing how accurately the type of visual cue could be predicted from neural activity.
  • Mutual Information Analysis: This was used to quantify the amount of information about the animal's behavioral choice (licking or not) encoded in neuronal populations.
  • Linear Mixed Effects Model: This statistical model was used to analyze the correlation structure within and between functional clusters, both before and after training, accounting for individual animal and day-to-day variability.
  • Anatomical Clustering Analysis: Techniques were used to determine if the functionally identified clusters also showed spatial organization within the brain tissue.

Datasets

Not explicitly detailed as separate datasets. The study utilized neural recordings from mice trained on the described visual association task.

Key Insights and Ideas

  • MEC's Role Beyond Spatial Navigation: The study strongly suggests that MEC is not solely a spatial processing center. It can encode complex, non-spatial associative learning and behavior-related variables, challenging previous assumptions.
  • MEC's Organized Representation of Behavior: A key finding is that MEC neurons form distinct functional clusters that are anatomically organized and maintain strong correlations, particularly for reward-related behaviors. This suggests a more structured internal representation compared to POR.
  • Learning-Induced Neural Changes: Both POR and MEC neurons show significant changes in their activity patterns as learning progresses. This includes developing biases towards rewarded trials and encoding information about behavioral choices.
  • POR as a Higher-Level Visual Integrator: POR neurons show more immediate and stable tuning to visual cues from the start of training, suggesting it acts as a more direct visual association area compared to MEC.
  • Behavioral Choice Precedes Action: The study provides evidence that neural activity in certain MEC clusters precedes the actual licking behavior, indicating that these neurons might be involved in predicting rewards or making the decision to act.

Key Results

  • Task Learning: Mice successfully learned the visual association task, demonstrated by predictive licking behavior that emerged with training.
  • Neural Tuning Changes: Both POR and MEC neurons showed dynamic changes in their response profiles during learning. POR neurons maintained stable cue tuning, while MEC neurons developed robust tuning towards rewarded trials.
  • Functional Clusters: Six task-relevant functional clusters were identified. Clusters C1 and C2 were strongly linked to licking behavior, while C3-C5 were related to stimulus processing.
  • Anatomical Organization in MEC: Clusters C1 and C2, particularly those related to behavior, showed significant anatomical clustering and separation in MEC, but not in POR.
  • Visual Information Encoding: Decoding accuracy for visual stimuli was significantly higher in POR than in MEC, suggesting POR encodes more detailed visual information. However, MEC showed a learning-related increase in stimulus encoding for the "plus" cue.
  • Behavioral Choice Information: Mutual information between neuronal responses and behavioral choice (lick vs. no-lick) increased with learning in both regions, especially within reward-related clusters.
  • Correlation Structure: Within-cluster correlations were stronger in MEC than in POR, and this difference was more pronounced for behavior-related clusters. These correlations were also maintained during periods of darkness (no task).
  • Task Reversal: Reversing cue-outcome associations led to significant changes in network responses, including a decrease in overall task-responsive neurons but a slight increase in stimulus representation.

Limitations and Future Work

  • Visual Information in MEC: While MEC showed some stimulus encoding, it was limited compared to POR. The authors suggest that spatial context might gate visual information integration in MEC, or that MEC's role in non-spatial tasks might be more focused on behavior and outcome rather than detailed visual features.
  • Origin of Reward Signals: The exact origin of reward modulation signals to MEC remains elusive, though anatomical evidence points to amygdala input.
  • Future Directions: The authors suggest future studies should explore the precise projections from POR and the amygdala to MEC, and investigate how cognitive processes can be disentangled from motor actions in these neural circuits. They also propose that future work could investigate how the transition from non-spatial to spatial contexts might alter MEC's processing mechanisms.

Summary updated

Abstract

Abstract The medial entorhinal cortex (MEC) is pivotal in spatial computations and episodic memory. In particular, an animal’s position can be decoded from the activity of entorhinal grid cells. However, it remains elusive whether MEC could play a more general role in different types of associative learning and how the representations develop during the learning process. It has been shown that the postrhinal cortex (POR), which is directly connected to MEC, integrates visual stimuli with salient outcomes. Here, we use a non-spatial visual association task to investigate whether MEC neurons represent low-level visual cues during learning. Using a Go/NoGo visual association task, we recorded neural activity in MEC and POR throughout the learning phase as mice associated drifting gratings with rewarded, aversive, or neutral outcomes. Our findings reveal that the neural tuning curves in both the POR and MEC change with the learning of the task. From the start of training, the POR neurons exhibited response tuning to the visual cues, and the tuning was stable to cue orientations during learning. In contrast, MEC neurons did not initially respond very strongly to visual cues but developed a robust tuning toward the rewarded trials. While the MEC’s representation of visual information was limited, it encoded other task elements. A large fraction of the neurons formed distinct functional clusters that were either activated or suppressed by reward-related behavior. Remarkably, these clusters segregated anatomically in MEC and maintained strong within-cluster correlations before and after training. Notably, although the same functional clusters were apparent in the POR, they did not show any anatomical structure as in the MEC. Task reversal induced significant changes in network responses across both regions, with a decrease in overall task-responsive neurons but a slight increase in stimulus representation. Strikingly, information about the choice to lick emerged with learning in both brain areas, and most significantly within the functional cell clusters representing reward consumption and plus-cue stimulus. Our results demonstrate that although neurons in MEC and POR develop behavior-modulated tuning during learning of a non-spatial visual association task, the MEC exhibits stronger within-cluster correlations and anatomical organization. Conversely, the POR population exhibits less structural organization and more specific stimulus-tuning, which is reflective of being a higher visual association area. Our findings reveal that the MEC can encode task– and behavior-related variables beyond spatial information.
BibTeX
@misc{nymoen2025task,
  author = {Ingeborg Nymoen and Frederik Sebastian Rogge and Ann Margareth Aasen and Sverre Grødem and Anders Malthe‐Sørenssen and Mikkel Elle Lepperød and Torkel Hafting and Marianne Fyhn and Kristian Kinden Lensjø},
  title = {{Task and Behavior-Related Variables Are Encoded by the Postrhinal and Medial Entorhinal Cortex During Non-Spatial Associative Learning}},
  year = {2025},
  doi = {10.7554/elife.106095},
}
Figure