Synapses form through a multi-step process where a growing nerve fiber reaches a target cell, makes contact using specialized adhesion proteins, and then both sides of the connection recruit the molecular machinery needed to send and receive signals. This process, called synaptogenesis, begins before birth and continues well into adolescence, with different brain regions following different timelines. It also happens on a smaller scale throughout adult life as the brain learns and adapts.
How Two Neurons Make First Contact
Synapse formation begins when a growing axon (the long output fiber of a neuron) extends toward its target. The tip of the axon, called a growth cone, navigates through brain tissue using chemical signals until it reaches the right neighborhood. Once there, it needs to recognize the correct target cell and lock on.
That recognition depends heavily on a pair of proteins that act like a lock and key across the tiny gap between neurons. One protein sits on the surface of the sending neuron, and its partner sits on the receiving neuron. When these two proteins bind together, they trigger changes on both sides of the contact point. Experiments have shown just how powerful this interaction is: when researchers placed the receiving-side protein on a non-neuronal cell, nearby neurons formed functional sending terminals onto that cell as if it were another neuron. The reverse also worked. Placing the sending-side protein on a non-neuronal cell caused nearby neurons to build receiving structures against it. These proteins essentially trick the cellular machinery into building a synapse wherever they appear.
Other adhesion molecules, including members of the cadherin family, also contribute to this initial handshake, though many of these proteins play roles in earlier stages of brain development too, making it difficult to tease apart their specific contribution to synapse building.
Building the Sending Side
Once initial contact is made, the sending neuron (the presynaptic side) needs to assemble an “active zone,” a dense cluster of proteins right at the contact point that will handle the release of chemical messengers. This assembly involves three core steps: generating the necessary proteins and transporting them down the axon, combining those proteins into functional complexes, and anchoring those complexes at the membrane directly opposite the receiving neuron’s receptors.
These steps don’t always happen in the same order. Some proteins arrive pre-packaged in small transport packets that travel down the axon, ready to fuse into a working release site almost immediately. Others are assembled on location. The end result is a tightly organized patch of membrane surrounded by tiny bubbles called synaptic vesicles, each loaded with neurotransmitter molecules waiting to be released when an electrical signal arrives.
Building the Receiving Side
On the other side of the synapse, the receiving neuron constructs what’s known as the postsynaptic density: a thick mat of proteins just beneath its membrane. The star player here is a scaffolding protein called PSD-95, which acts like a molecular pegboard. It has multiple binding sites that grab onto receptors, ion channels, signaling enzymes, and structural components, clustering them all together at exactly the right spot.
PSD-95 anchors the receptors that detect incoming neurotransmitter signals, including receptors for glutamate, the brain’s primary excitatory chemical messenger. By holding all these components in close proximity, PSD-95 ensures that when neurotransmitter is released from the sending side, the receiving side can respond quickly and reliably. How important is this scaffolding protein? When the gene encoding PSD-95 is disrupted in humans, 56% of affected individuals develop autism spectrum disorder, along with intellectual disability and other neuropsychiatric conditions. That gives a sense of how precisely synapse assembly needs to go for normal brain function.
Astrocytes Help Drive the Process
Neurons don’t build synapses alone. Astrocytes, the most abundant non-neuronal cells in the brain, secrete proteins that actively promote synapse formation. Among the first such factors identified were thrombospondins, a family of proteins that astrocytes release during peak periods of synapse building. These proteins act on a calcium channel component on the neuron’s surface, essentially giving neurons the green light to form new connections.
Interestingly, this signaling has a biological sex difference. In laboratory cultures, neurons from male animals respond strongly to one of these astrocyte-secreted proteins (thrombospondin-2), ramping up synapse formation, while neurons from female animals do not show the same response. This suggests that the molecular recipe for synapse building may differ between sexes, even at the cellular level. Astrocytes also contribute later in the process by helping to eliminate excess synapses, using a separate set of receptors to engulf connections that are no longer needed.
Neural Activity Decides Which Synapses Survive
Forming a synapse is not a guarantee it will last. Newly formed synapses are essentially on probation, and their survival depends on whether they carry useful signals. When a synapse is active and successfully transmits messages, it gets strengthened. When it’s quiet or redundant, it’s marked for removal.
This activity dependence has been demonstrated clearly in experiments where nerve signaling was blocked with anesthetics. When researchers prevented electrical activity in a nerve during early development, the normal increase in signaling strength at connected synapses failed to occur. The frequency of spontaneous signaling events dropped significantly compared to controls. In other words, synapses that don’t get used don’t mature properly.
The flip side is also true. Increased activity can boost synaptic strength. This is the cellular basis of learning: repeated activation of a circuit makes its synapses more efficient, while neglected pathways weaken and may eventually disappear.
Pruning Trims the Excess
The developing brain massively overproduces synapses, then cuts back. This trimming process, called synaptic pruning, relies on the brain’s resident immune cells, called microglia. These cells patrol the brain and physically engulf synapses that have been tagged for removal.
The tagging system borrows from the immune system’s complement pathway, the same molecular cascade the body uses to mark bacteria for destruction. A protein called C1q attaches to weak or inactive synapses, initiating a chain reaction that deposits another complement protein on the synapse surface. Microglia recognize this tag through a specific receptor and swallow the marked synapse whole. Mice engineered to lack C1q, or the downstream complement proteins, or the microglial receptor itself, all end up with too many synapses and abnormal neural circuits.
There’s also a protective signal. Synapses can display a “don’t eat me” protein called CD47 on their surface. When microglia sense CD47, they leave that synapse alone. Removing the microglial receptor for CD47 leads to excessive, indiscriminate pruning. The balance between “eat me” complement tags and “don’t eat me” protective signals determines which synapses stay and which are eliminated. Astrocytes also participate in pruning through a separate mechanism involving their own set of engulfment receptors.
Timing Varies Across Brain Regions
Synapse formation doesn’t happen everywhere at once. In the primary visual cortex, a burst of synapse building between 3 and 4 months of age pushes synaptic density to 140 to 150 percent of adult levels by 4 to 12 months. After that peak, the count declines as pruning takes over.
The prefrontal cortex, responsible for planning, decision-making, and impulse control, follows a slower schedule. Synapse formation there begins around the same time as in the visual cortex, but doesn’t reach its peak until about 8 months and continues building through the second year of life. Pruning in the prefrontal cortex then extends into the mid-20s, which is one reason this region is among the last to fully mature.
Synapse Formation in the Adult Brain
Synaptogenesis doesn’t stop after childhood. Adults continue to form new synapses, particularly in the hippocampus, a region critical for memory. New neurons generated in the adult hippocampus go through maturation stages that closely mirror those seen during fetal development, assembling synapses using the same basic molecular toolkit.
There is one notable difference. During early brain development, neurons extend axons into target regions that are still growing and changing shape. As these regions expand, the axon branches get pulled apart, resulting in widely spread connection patterns. Neurons born in the adult brain don’t have this issue. They’re wiring into a stable, fully grown structure, so their connections tend to be more compact and targeted. This means adult-born synapses are structurally tighter, even though they follow the same fundamental assembly steps.