The brain is sculpting signal out of noise
Deep within the architecture of the human mind, a delicate balance between activation and restraint governs how experience becomes memory. Researchers at NYU Langone Health have now mapped the precise neural choreography between the entorhinal cortex and the hippocampus that keeps memories stable during learning — a discovery published in Science that illuminates why, when this balance fails, the mind can no longer distinguish a genuine threat from a phantom one. The work offers not only a clearer picture of how the brain holds onto what it knows while remaining open to the new, but a potential roadmap toward more precise treatments for the millions living with PTSD, schizophrenia, and other disorders of memory.
- The brain's ability to form stable memories depends on a three-part neural dance — excitation, inhibition, and disinhibition — that researchers have now decoded at the level of individual neurons.
- When this circuit breaks down, the consequences are severe: a soldier's brain may confuse a popping balloon with a bomb blast, unable to distinguish real danger from a false alarm.
- The NYU team, led by neuroscientist Jayeeta Basu, traced two types of long-range neural connections from the entorhinal cortex to the hippocampus, identifying how their interplay creates the stable 'place maps' essential for learning and recall.
- First author Vincent Robert described the finding as revealing how the brain selectively turns down inhibition in key microcircuits to amplify attention to what matters most.
- The research fills a long-standing gap in neuroscience by explaining how distant brain regions control local circuits — opening the door to treatments that target the specific excitation-inhibition imbalances underlying memory disorders.
A team of neuroscientists at NYU Langone Health has mapped the neural choreography that keeps memories stable during learning — a discovery published in Science with significant implications for treating disorders like PTSD and schizophrenia.
At the heart of the work is a conversation between two brain regions: the entorhinal cortex, which processes sensory information, and the CA3 region of the hippocampus, the brain's mapmaker for spatial memory. Two types of signals traveling between them work in tandem — one excitatory, one inhibitory — creating a balance that allows the brain to lock in learning while filtering out noise. A third layer, disinhibition, completes the circuit, producing the stable neural patterns that encode where we are and what we've learned.
Memory, the researchers emphasize, is not passive recording. It is pattern extraction — and those patterns must be stable enough to be useful yet flexible enough to adapt. When the system works, a person recalls where they parked their car. When it fails, a veteran's nervous system cannot distinguish a balloon pop from a blast, because the memory map has lost its precision.
Led by Jayeeta Basu, a recent Presidential Early Career Award recipient, and first author Vincent Robert, the team examined long-range neural connections that had been poorly understood — fibers releasing either glutamate or GABA that reshape local hippocampal circuits during learning. Their findings provide a mechanistic account of how the brain balances stability against change, and suggest that future treatments for memory disorders could be designed with far greater precision, targeting the specific imbalances in excitation and inhibition that underlie cognitive symptoms.
A team of neuroscientists at NYU Langone Health has mapped out the precise neural choreography that keeps memories stable as we learn—a discovery published in Science that could reshape how doctors treat memory disorders ranging from post-traumatic stress to schizophrenia.
The work centers on a conversation between two brain regions: the entorhinal cortex, which processes sensory information about the world around us, and the CA3 region of the hippocampus, which acts as the brain's mapmaker, encoding spatial memories and patterns. Researchers found that two types of neural signals traveling from the entorhinal cortex to the hippocampus work in tandem to stabilize these memory maps. One signal excites neurons, boosting their activity. The other inhibits them. Together, these opposing forces create a kind of neural balance that allows the brain to lock in what it's learning while filtering out noise.
The mechanism matters because memory isn't simply about recording experience like a video camera. It's about extracting patterns—learning that sugar water sits in the left corner of a maze, or that a particular street corner leads home. These spatial maps have to remain stable enough to be useful, yet flexible enough to adapt when the environment changes. When this system works properly, a mouse can navigate a maze or a person can recall where they parked their car. When it breaks down, the consequences are severe.
In conditions like schizophrenia and post-traumatic stress disorder, the stability and precision of these memory circuits deteriorate. A soldier's brain might misfire when a balloon pops at a party, triggering a freezing fear response because the neural system has confused the sound with a bomb blast. The memory map has become unstable, unable to distinguish between a genuine threat and a false alarm. Understanding how the brain normally maintains this stability opens a path toward interventions that could restore it.
The research team, led by Jayeeta Basu, an assistant professor in the departments of Psychiatry and Neuroscience, focused on neurons with long extensions that reach across distant brain regions—the kind of long-distance wiring that had been poorly understood. They identified two specific types of these long-range connections: one that releases glutamate, an excitatory neurotransmitter, and another that releases GABA, an inhibitory one. The excitatory signal drives activity in the hippocampus, but it also triggers local inhibition that fine-tunes the firing pattern. The inhibitory signal then suppresses that local inhibition, effectively disinhibiting the circuit and encouraging activity. This three-part dance—excitation, inhibition, and disinhibition—creates the stable neural patterns that encode place maps.
Vincent Robert, the study's first author and a postdoctoral scholar in Basu's lab, described the finding as a dissection of how the brain amplifies attention to important sensory information by selectively turning down the volume on inhibition in key microcircuits. The team examined these interactions at the level of individual neurons, mapping how long-range inputs reshape local circuits to support learning. The result is a mechanistic understanding of how the brain balances stability against change—how it holds onto what it knows while remaining open to new information.
Basu, who recently won the Presidential Early Career Award for Scientists and Engineers, emphasized that the work fills a substantial gap in neuroscience. For years, researchers knew that the entorhinal-hippocampal circuit was critical for both forming memories and retrieving them by completing patterns from partial cues. What was missing was a clear picture of how long-range neural inputs control the local circuits that make this possible. The new study provides that picture, suggesting that future treatments for memory disorders could be designed with far greater precision, targeting the specific imbalances in excitation and inhibition that underlie cognitive symptoms.
Notable Quotes
Our study fills in a substantial gap in understanding how long-range inputs control neuronal circuits essential for memory recall.— Jayeeta Basu, senior study author
The team detailed a circuit mechanism that fine-tunes the dialogue among excitation, inhibition, and disinhibition in service of context-dependent memory formation.— Vincent Robert, first study author
The Hearth Conversation Another angle on the story
Why does it matter that these two signals arrive at the same time? Couldn't the brain achieve stability with just one?
The timing is everything. Excitation alone would create noise—too much firing, too little specificity. Inhibition alone would suppress learning. Together, they create a filter. The excitatory signal says "pay attention to this," and the inhibitory signal says "but only to this precise pattern." The brain is sculpting signal out of noise.
You mentioned that memory maps have to be stable yet flexible. How does the brain know when to hold firm and when to adapt?
That's the elegant part. During learning, the balance tips toward excitation—the brain is more receptive, encoding new information. At rest, the balance maintains itself, keeping the map intact. It's not a switch; it's a dial. The same circuits that stabilize also allow for gradual updating as the environment changes.
In PTSD, you said a balloon pop triggers a bomb response. Is the circuit broken, or is it just calibrated wrong?
It's calibrated wrong. The circuit itself is intact, but the memory map has become unstable—it's conflating different experiences, different contexts. The brain can't distinguish between the party and the battlefield. Restoring stability might mean helping the circuit re-encode those memories with proper context, so the fear response fires only when it should.
Does this research suggest a drug target?
Not yet, but it points in a direction. If you understand that disinhibition is key to stabilizing place maps, you might develop compounds that modulate GABAergic signaling in specific ways. The challenge is precision—you'd want to restore stability without disrupting the flexibility the brain needs for normal learning and adaptation.
How did they actually observe these signals firing at the cellular level?
They used techniques that allow them to record from individual neurons while the animal is learning—watching the electrical activity as the brain encodes spatial information. They could see which neurons fired together, in what sequence, and how long-range inputs shaped that activity. It's like listening to an orchestra and finally understanding how the brass section coordinates with the strings.