How the brain learns and applies rules: sequential neuronal dynamics in the prefrontal cortex

Understanding how the brain learns and applies rules is the key to unraveling the neural basis of flexible behavior. A new study from the University of Toyama, Japan, reveals that our ability to follow procedural rules is encoded in the evolving dynamics of neuronal activity in the medial prefrontal cortex (mPFC).

The research team, led by Assistant Professor Shuntaro Ohno at the Faculty of Medicine, University of Toyama, Japan, recorded neuronal activity in mice learning a Y-maze task. As learning progressed, distinct sequences of neural activation emerged in the mPFC that could predict whether the mouse succeeded or failed at obtaining a reward. Their findings were published in Volume 18, Article 56 of Molecular Brain on July 1, 2025.

The researchers placed the mice individually in a Y-shaped maze and allowed them to explore the maze without any restriction in the beginning. The branched arms were called the “Zone,” where each mouse had to wait and then respond to a light cue to navigate its way to the water container called the “Port,” and finally claim a water reward by licking within the predetermined time. As their training progressed, the mice became faster and more successful in obtaining rewards, although the physical paths they took remained the same. Meanwhile, the scientists recorded hundreds of mPFC neurons through calcium imaging, capturing how neural populations changed during the learning process.

To comprehend this complex neural data, the team developed iSeq—a novel computational tool that applies convolutional non-negative matrix factorization to automatically detect neuronal sequences from imaging data without any prespecified behavioral labels. These sequences represent ordered patterns of neural activation spanning several seconds. The analyses revealed that in the initial phases of training, the sequences were less predictive. However, by day 6, the dynamics of sequences differed significantly between successful and unsuccessful reward acquisition moments in mice that had mastered the task, even before the action occurred.

“The development of iSeq allowed us to observe the brain’s internal organization of behavior in unprecedented detail,” explained Dr. Ohno. “We found that as the animals learned, their prefrontal cortex dynamically restructured neural activity patterns to emphasize actions that reliably led to rewards.”

Furthermore, the researchers observed that the composition of neurons participating in each sequence changed across days of training. In other words, the set of cells that formed the sequence on day 1 was not the same as the one on day 6, indicating that the mPFC continually reorganized its neural circuits as the behavior became refined. This flexible reconfiguration, rather than the reuse of fixed neural assemblies, reflects the brain’s capacity to adapt its internal representations.

“These results suggest that the brain does not store a rule as a static template,” noted Dr. Ohno. “Instead, it continuously updates sequential activity patterns to link meaningful sensory cues, actions, and outcomes—essentially learning how to learn.”

These results bridge the gap between neural activity and behavioral rule execution. They suggest that a procedural rule—analogous to a cascade such as stimulus → action → reward—is represented in the brain as a chain of neural events. This chain is not fixed but evolves as the animal becomes competent; the brain reorganizes neural sequences to align with successful behavior. Understanding the underlying mechanism offers new insight into how cognitive control, learning, and adaptation are instantiated in neural circuits.

The implications of this study extend beyond basic science. Insights into how rules are encoded and updated might inform rehabilitation strategies after a brain injury, or how artificial intelligence might mimic this flexibility. Moreover, the computational method iSeq could become a tool for investigating sequence-based neural dynamics in other areas of the brain and could be extended to other species.

While the study was conducted in mice, it provides a foundational framework for understanding how the human brain learns to execute rules and adapt its behavior. The findings highlight the importance of temporal patterns in brain activity, and how they might form the basis of flexible, learned behavior rather than static connectivity alone.

***

Reference
DOI: 10.1186/s13041-025-01230-w

About University of Toyama, Japan
University of Toyama is a leading national university located in Toyama Prefecture, Japan, with campuses in Toyama City and Takaoka City. Formed in 2005 through the integration of three former national institutions, the university brings together a broad spectrum of disciplines across its 9 undergraduate schools, 8 graduate schools, and a range of specialized institutes. With more than 9,000 students, including a growing international cohort, the university is dedicated to high-quality education, cutting-edge research, and meaningful social contribution. Guided by the mission to cultivate individuals with creativity, ethical awareness, and a strong sense of purpose, the University of Toyama fosters learning that integrates the humanities, social sciences, natural sciences, and life sciences. The university emphasizes a global standard of education while remaining deeply engaged with the local community.

Website: https://www.u-toyama.ac.jp/en/

About Assistant Professor Shuntaro Ohno from the University of Toyama
Dr. Shuntaro Ohno is a neuroscientist at the University of Toyama, Japan, whose expertise lies in developing advanced analytical and computational methods to decode brain activity. His research focuses on integrating and analyzing in vivo calcium imaging and electrophysiological data to uncover how neural populations represent and update procedural knowledge. In the reported study, he applied these analytical frameworks to reveal how neuronal sequence dynamics in the medial prefrontal cortex evolve during rule learning. His work contributes to the understanding of cognitive control and neural adaptation, and he continues to explore how temporal patterns of neural activity support flexible behavior.