Groundbreaking study sheds light on how our brain learns and adapts

Neurobiologists employing advanced visualization methods have uncovered how alterations within our synapses and neurons occur. The results illustrate the way information is handled within the brain’s circuitry, providing valuable insights for neurological conditions and brain-inspired AI systems.

How do we acquire new knowledge? How do responsibilities at a new position, words to the newest popular song, or navigation to a friend’s residence get stored in our minds? The general response is that our brains make adjustments to process new information. To adopt a new behavior or remember newly learned information, the brain’s networks adapt.

These alterations are coordinated through trillions of synapses — the links between single nerve cells, known as neurons — where communication in the brain occurs. In a meticulously organized process, fresh information strengthens some synapses with new data while causing others to weaken. Neuroscientists who have thoroughly examined these changes, called “synaptic plasticity,” have discovered various molecular mechanisms that drive this plasticity.

However, the understanding of the “rules” determining which synapses experience this process remained unclear, a mystery that ultimately governs how information learned is stored in the brain. Neurobiologists from the University of California, San Diego, William “Jake” Wright, Nathan Hedrick, and Takaki Komiyama have revealed important information regarding this process.

The principal financial backing for this multi-year research was supplied by various grants from the National Institutes of Health and a training grant. As reported on April 17 in the journal Science, the scientists employed an advanced brain visualization technique, utilizing two-photon imaging, to closely study the brain activity of mice and monitor the actions of synapses and neurons during learning tasks.

The new images, allowing observation of individual synapses like never before, showed that neurons do not adhere to a single set of rules during learning episodes, contrary to traditional beliefs.

Neuroscientists have meticulously examined how synapses possess access solely to their own “local” information, yet together they contribute to the development of extensive new learned behaviors, a dilemma known as the “credit assignment problem.” The problem is similar to single ants performing designated tasks without awareness of the broader objectives of the whole colony.

The latest data provides encouraging perspectives for the future of artificial intelligence and the brain-inspired neural networks on which they function. Usually, a whole neural network operates under a unified set of plasticity rules; however, this study suggests potential novel approaches to creating advanced AI systems by employing various rules within individual units.

Regarding health and behavior, the results may provide a novel approach to addressing issues like addiction, post-traumatic stress disorder, and Alzheimer’s, as well as neurodevelopmental disorders such as autism.