Your brain is carrying more spare wiring than anyone suspected. In a study that upended a long-standing assumption in neuroscience, researchers at the Massachusetts Institute of Technology found that roughly 30 percent of synapses in the adult mouse cortex are sitting dormant. They contain the molecular hardware needed to transmit signals but remain electrically silent, apparently waiting for the moment the brain needs them to learn something new.
The finding, published in Nature, challenges the textbook view that these so-called silent synapses disappear after early childhood development. If the result extends to humans, it means the adult brain maintains a large, flexible reserve of pre-built connections that can be switched on without tearing apart the circuits already in use.
What the MIT team actually found
The researchers used a technique called eMAP, which physically expands preserved brain tissue so that individual protein clusters become visible under super-resolution microscopy. They applied it to more than two thousand synapses from layer 5 pyramidal neurons in the primary visual cortex of adult mice.
About 25 percent of those connections were missing AMPA receptors, the molecular gates that allow ions to flow when a synapse fires under normal conditions. Without them, a synapse stays electrically quiet even when the neuron on the other side releases a signal. But those same synapses still had NMDA receptors, a second type of receptor that opens only under specific voltage conditions. That combination, NMDA present but AMPA absent, is the classic signature of a silent synapse.
The team also noticed a structural pattern. Many of the dormant connections sat on filopodia, thin finger-like protrusions that extend from dendrites. Filopodia are common in developing brains, but neuroscientists had not expected to find them persisting in large numbers in mature cortex. In the Nature paper, the researchers reported that these structures serve as a physical scaffold for the silent connections, giving them a recognizable shape distinct from the mushroom-headed dendritic spines that characterize active synapses.
MIT’s institutional summary placed the overall figure at about 30 percent, folding in electrophysiology experiments alongside the protein-imaging data. That number represents a strikingly large fraction of cortical wiring sitting in standby mode.
Why this matters beyond the lab
The concept of silent synapses is not new. Experimental evidence for their existence was first published in 1995, when two independent groups, Isaac et al. and Liao et al., showed that some connections in young brain tissue failed to produce signals at resting potential yet responded when the postsynaptic cell was artificially depolarized. Later work demonstrated that activity-dependent insertion of AMPA receptors can convert a silent synapse into a functional one, a process closely tied to long-term potentiation, the cellular mechanism behind memory formation.
What the MIT study changed is the scale and the timing. Before this work, the prevailing view held that silent synapses were a feature of early brain development, present in abundance during the critical periods of infancy and childhood but largely pruned away by adulthood. Finding that nearly a third of cortical synapses remain dormant in grown mice suggests the adult brain is far more structurally prepared for change than neuroscientists had assumed.
That has potential implications well beyond basic science. If the adult human brain also maintains a large pool of recruitable connections, it could help explain how people continue to acquire complex skills, recover function after stroke, or adapt to sensory loss later in life. It could also open new avenues for understanding what goes wrong in conditions where plasticity is impaired, from neurodegenerative diseases to age-related cognitive decline.
What remains uncertain
The 30 percent figure comes from one brain region in one species. The imaging focused on layer 5 pyramidal neurons in the mouse primary visual cortex, and as of June 2026, no published data from the same group confirm whether the proportion holds in other cortical areas, subcortical structures, or in awake, behaving animals. Extending the finding to humans requires caution. No human tissue measurements of adult silent synapse prevalence appear in the study or its supporting materials, and earlier reviews that discuss AMPA-silent synapses have generally focused on developmental stages and pathological conditions rather than healthy adult cortex, though the precise scope of any individual review should be verified independently.
The mechanism by which silent synapses get activated during actual learning also lacks direct demonstration in this dataset. Earlier studies showed that rapid AMPA receptor recruitment can convert silent synapses into active ones in brain-slice preparations, but the MIT paper did not include behavioral experiments tracking whether specific silent synapses switch on during a learning task. That causal link, from dormant synapse to new memory, remains an inference supported by decades of plasticity research rather than a result proven in this particular study.
A related open question is spatial organization. Do silent synapses cluster near recently active connections, which would let the brain recruit new capacity in a targeted, local fashion during skill acquisition? Or are they scattered randomly across a neuron’s dendritic tree, which would require a different organizational logic? The published data do not resolve this, and future work will need to combine structural imaging with activity measurements to map how filopodia relate to functional circuits.
There is also the question of stability. The current work provides a snapshot of adult cortex, not a time-lapse. It is not yet clear whether individual filopodia-based synapses persist for months, appear only transiently during specific windows of experience, or cycle between silent and active states as AMPA receptors traffic in and out. Each scenario would imply different rules for how the brain manages its reserve of plastic connections.
Finally, the role of silent synapses in disease remains speculative. Because they are abundant during development and re-emerge in certain pathological conditions, some researchers have proposed that abnormal regulation of these dormant contacts could contribute to neurodevelopmental disorders or maladaptive plasticity after injury. The MIT findings add weight to the idea that silent synapses are a normal feature of adult circuitry, but they do not yet show whether shifting the balance between silent and active connections helps or harms brain function in specific clinical settings.
Separating what is solid from what is still a hypothesis
The strongest evidence here is the peer-reviewed Nature paper, which provides direct protein-level imaging of thousands of individual synapses. That dataset is the primary source for the 25 percent figure and the filopodia finding. The slightly higher 30 percent estimate in MIT’s press materials incorporates electrophysiology results and should be understood as a rounded institutional summary, not a contradiction.
Background claims about what silent synapses are and how they become active rest on a well-established literature stretching back nearly three decades. A commentary in Nature Reviews Neuroscience placed the MIT findings in the context of this earlier work, treating the adult prevalence result as a significant extension of known principles rather than a wholesale rewrite of the field.
Readers should keep two distinct claims separate. The first is anatomical and molecular: a substantial fraction of synapses in adult mouse visual cortex have NMDA receptors but no detectable AMPA receptors and often reside on filopodia-like structures. This claim rests directly on the imaging and appears robust within the sampled tissue. The second is a functional interpretation: that these silent synapses constitute a built-in reserve the brain can rapidly convert into active circuitry during learning.
The functional interpretation is plausible and consistent with prior plasticity studies, but it is necessarily more tentative. The current data do not track individual synapses through a learning episode or show AMPA receptors being inserted into specific filopodia in a living animal. Until such experiments are performed, the idea that these structures are the primary substrate for new memory traces should be viewed as a leading hypothesis, not a settled fact.
For now, the central takeaway is that adult cortex appears far more structurally prepared for change than previously thought. Rather than having to dismantle and rebuild existing synapses to encode new information, the brain may be able to recruit a pool of pre-formed, NMDA-only connections, turning them on by adding AMPA receptors when experience demands it. That vision of a flexible yet stable circuit architecture is already shaping the next wave of experiments aimed at linking silent synapses to moment-to-moment learning and long-term behavioral change.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
