Striatal interneuron microcircuits gate reinforcement to stabilize adaptive choice

Every day, we make countless small choices, from deciding which route to take to work to picking a meal from a menu. Some of these choices are learned through trial and error, a process where our brains weigh the success or failure of an action against the outcome it produced. This learning happens in a specific region of the brain called the striatum, a deep structure that acts as a hub for processing information about rewards and punishments. Within this hub, a type of nerve cell known as a spiny projection neuron acts as the main messenger, sending signals that drive our behavior forward. Scientists have long understood that these messenger cells are controlled by excitatory inputs, but the role of the local inhibitory cells—those that act as brakes or filters within the striatum itself—has remained a mystery. Understanding how these local brakes work is crucial because they might be the key to how we adapt our habits when the rules of the world change, shifting from a successful strategy to a new one when our old ways stop working.

A team of researchers recently turned their attention to the dorsomedial striatum to uncover how these local inhibitory circuits influence learning. They focused on a specific type of learning task where an animal had to push or pull a lever to get a reward, but the rules of the game would occasionally flip. Sometimes pushing was the right move, and sometimes pulling was, and the animal had to figure out the change based on whether it received a reward or not. By watching the activity of different inhibitory nerve cells during this process, the scientists discovered a precise internal circuit that links the immediate result of a single trial to the broader strategy the animal uses to make choices. They found that two specific types of inhibitory cells, distinguished by the proteins they contain, react in opposite ways to the same events. One type, which contains a protein called somatostatin, becomes active when a trial goes wrong or when a reward comes unexpectedly. The other type, containing a protein called tyrosine hydroxylase, does the opposite: it quiets down when a trial fails and springs into action when a reward is received.

The researchers then mapped out how these two cell types talk to each other and to the main messenger cells. They found that when the tyrosine hydroxylase cells are active, they suppress the somatostatin cells. This suppression removes a brake, allowing the main messenger cells to fire more freely. This chain of events creates a pathway where a good outcome triggers a specific sequence that boosts the activity of the cells responsible for driving action. To test if this circuit was actually necessary for learning, the scientists temporarily silenced these cells in living animals. When they stopped the somatostatin cells from working, the animals began to make the wrong choice repeatedly, sticking to a suboptimal strategy even when it no longer paid off. Conversely, when they silenced the tyrosine hydroxylase cells, the animals struggled to learn from positive rewards, failing to reinforce the correct behavior. These experiments showed that this specific microcircuit acts as a gate, deciding whether the brain should update its behavior based on a single trial's outcome.

The study also looked at what happens inside the cells after these behavioral changes occur. When the somatostatin cells were inhibited, the researchers observed that the connections between the excitatory inputs and the main messenger cells became stronger over time. This strengthening suggests a physical change in the brain's wiring that could explain why the animals kept making the same wrong choice long after the experiment ended. The findings indicate that this disinhibitory circuit does more than just react to a single event; it transforms the immediate result of a trial into a lasting change in policy. By gating the flow of reinforcement through this specific local loop, the brain can stabilize adaptive choices, ensuring that we learn from our successes and failures in a way that guides our future actions.