Synaptic change during memory formation is a multiscale process that converts fleeting electrical activity into lasting alterations in connectivity. At its core is synaptic plasticity, the capacity of synapses to strengthen or weaken based on activity, which underpins learning across species.
Molecular and structural changes at the synapse
At the molecular level, the NMDA receptor functions as a coincidence detector. When presynaptic glutamate release coincides with postsynaptic depolarization, NMDA receptors admit calcium into the dendritic spine. This calcium signal activates kinases such as CaMKII and triggers rapid AMPA receptor insertion into the postsynaptic membrane, increasing synaptic efficacy. These mechanisms define early-phase long-term potentiation described by Tim Bliss at the MRC Laboratory of Molecular Biology and Terje Lømo at the University of Oslo, who first characterized LTP in the hippocampus.
For changes to persist beyond hours, new gene transcription and protein synthesis are required. Eric R. Kandel at Columbia University demonstrated in Aplysia that long-term memory formation depends on protein synthesis and structural synaptic growth. Transcription factors such as CREB and immediate early genes like Arc remodel the cytoskeleton and scaffold proteins, producing spine enlargement or the formation of new spines and boutons. Not every activated synapse is stabilized: only those integrated into coherent patterns of activity and neuromodulatory context proceed to durable structural change.
Network-level and behavioral consequences
At the circuit scale, selective strengthening and weakening of synapses reshapes neural representations and forms memory traces or engrams. Susumu Tonegawa at the Massachusetts Institute of Technology used optogenetic tagging to show that specific ensembles of hippocampal neurons can be reactivated to retrieve a memory, indicating that synaptic modifications map onto identifiable cell assemblies. Homeostatic mechanisms described by Gina G. Turrigiano at Brandeis University help maintain overall network stability by scaling synaptic strengths up or down, preventing runaway excitation as individual synapses potentiate.
The consequences of synaptic change extend to development, culture, and health. Environmental enrichment increases dendritic spine density and improves learning in animal models, a phenomenon traced back by Marian C. Diamond at the University of California Berkeley to experience-dependent cortical plasticity. Conversely, chronic stress and neurodegenerative processes preferentially degrade synapses; synaptic loss correlates strongly with cognitive decline in human dementia, making synaptic maintenance a therapeutic target.
Understanding synaptic change clarifies why timing, repetition, and emotional context influence learning. Repeated, well-spaced activation promotes consolidation through both synaptic strengthening and gene-driven remodeling, while neuromodulators like dopamine gate which synapses become lasting. Clinically, this explains why interventions that enhance synaptic resilience or restore plasticity have potential to ameliorate cognitive disorders and why lifestyle factors that stimulate synaptic growth can support brain health.
The study of how synapses change during memory formation therefore bridges molecules to behavior, with robust experimental support from decades of work in molecular neuroscience, electrophysiology, and systems studies. Mechanisms are context-dependent and continually regulated, which is why translating basic findings into treatments requires careful attention to timing, cell types, and the broader physiological environment.