Atlas One · Human · Tissue

Synapse

The chemical synapse — a precisely engineered ~20 nm intercellular junction that is the fundamental unit of information transfer in neural circuits, spanning millisecond ion channel gating to multi-hour structural memory encoding.

The adult human brain contains an estimated 100–500 trillion synapses. The molecular machinery of vesicle exocytosis (SNARE complex) was elucidated in work recognised by the 2013 Nobel Prize (Thomas Südhof). LTP and LTD are the synaptic correlates of learning and memory.

~20 nmSynaptic cleft width
100–500 trillionSynapses in brain
1–5 msAP to PSP peak time
Nobel 2013SNARE/vesicle fusion
GO:0045202 · FMA:45021 · Atlas One / Scale 05 / Tissue

Synapse

The chemical synapse is the fundamental unit of information transfer in neural circuits — a precisely engineered nanoscale intercellular junction allowing one neuron to influence the membrane potential of another. At ~20 nm wide, it hosts processes ranging from millisecond-timescale ion channel gating (fast synaptic transmission) to multi-hour structural remodelling (LTP/LTD) that encodes long-term memory. SNARE/synaptotagmin-driven vesicle fusion, postsynaptic PSD scaffolding, and bidirectional Hebbian plasticity constitute the core molecular logic.

Overview

The chemical synapse is far below the resolution of conventional light microscopy yet is the site of processes central to all of neuroscience: from the millisecond-timescale opening of ion channels that underlie perception and movement, to the multi-day structural changes in dendritic spines that constitute the physical trace of a memory. The adult human brain contains an estimated 100–500 trillion synapses — orders of magnitude more than the number of stars in the Milky Way. This synaptic number, combined with the diversity of synapse types and the capacity for bidirectional strength modification, creates the combinatorial complexity that makes human learning possible.

The molecular machinery of synaptic vesicle exocytosis — the SNARE complex (synaptobrevin/VAMP, syntaxin-1, SNAP-25) and its Ca²⁺ sensor synaptotagmin-1/2 — was elucidated in work recognized by the 2013 Nobel Prize in Physiology or Medicine (Thomas Südhof and colleagues). Structural plasticity of dendritic spines has been directly imaged in living mice, revealing ongoing dynamics of spine formation and elimination that correlate with learning.

Two broad synaptic classes exist: excitatory glutamatergic synapses (onto dendritic spines; AMPA + NMDA receptors in the PSD; ~80% of CNS synapses) and inhibitory GABAergic synapses (onto shaft/soma/AIS; GABA-A/B receptors; gephyrin scaffold). Together they define the E/I balance that determines network excitability, information routing, and plasticity.

Anatomy — Presynaptic Terminal

ComponentFunction
Synaptic vesicles (SVs, ~40 nm)Store concentrated neurotransmitter: ~5,000–10,000 glutamate molecules; ~3,000 GABA molecules per vesicle. SV pools: readily releasable pool (RRP, ~5–20 docked), recycling pool, reserve pool.
Active zone (AZ)Electron-dense cytomatrix scaffold: Bassoon, Piccolo (large scaffolds), RIM1/2 (Rab3 effector, tethers vesicles, opens Ca²⁺ channels), ELKS/CAST, Munc13 (vesicle priming). Positions vesicles within 50–100 nm of Ca²⁺ channels for rapid coupling.
P/Q-type and N-type Ca²⁺ channelsClustered within 50–100 nm of docked vesicles. Ca²⁺ entry at active zone microdomain rises to ~10–100 µM within 0.2 ms of AP arrival, triggering exocytosis.
SNARE complexSynaptobrevin-2/VAMP2 (v-SNARE, vesicle) + syntaxin-1A (t-SNARE, target plasma membrane) + SNAP-25 (t-SNARE, target); form a four-helix bundle driving bilayer merger. Regulated by Munc18 (syntaxin chaperone) and Munc13 (priming factor).
Synaptotagmin-1/2Ca²⁺ sensor on vesicle membrane; two C2 domains bind 3–5 Ca²⁺ ions (affinity ~100 µM); Ca²⁺ binding displaces inhibitory clamp on SNARE → full fusion within ~0.2 ms.
Endocytic machineryClathrin/dynamin/AP2 recycle membrane after full collapse fusion. Kiss-and-run (pore closure without full collapse) is faster but deposits fewer transmitter molecules.

Anatomy — Postsynaptic Density and Cleft

The synaptic cleft (~20 nm) is not empty: it contains a dense extracellular matrix of proteoglycans and transsynaptic adhesion complexes (neurexin on presynaptic / neuroligin on postsynaptic) that align active zone and PSD and are genetically implicated in autism spectrum disorder (NRXN1, NLGN3/4 mutations).

PSD ComponentRole (excitatory/glutamatergic synapse)
PSD-95 (MAGUK scaffold)Master organiser; clusters NMDA receptors (PDZ1/2 binds GluN2 C-terminus), anchors CaMKII and nNOS; ≥300 copies per PSD.
AMPA receptors (GluA1/2/3/4)Mobile — their number in the PSD is the primary determinant of synaptic strength. LTP → more surface AMPARs; LTD → fewer. GluA2-containing receptors are Ca²⁺-impermeable (normal); GluA2-lacking → Ca²⁺-permeable (AMPA receptor trafficking in plasticity).
NMDA receptors (GluN1/GluN2A/2B)Anchored by PSD-95; voltage-dependent Mg²⁺ block relieved only at depolarised potentials (coincidence detector). Essential for LTP induction; more stable in PSD than AMPARs.
CaMKII (α/β)~30% of PSD protein mass; 12-subunit holoenzyme; activated by Ca²⁺/CaM during LTP → autophosphorylation (T286) → constitutive activity → GluA1 Ser831 phosphorylation → AMPAR conductance ⇧ and trafficking to PSD.
HomerClusters mGluR1/5; links to IP₃ receptors on ER within spine (“spine apparatus”) → local Ca²⁺ store for plasticity and protein synthesis.
ShankConnects Homer and PSD-95 into unified scaffold; mutated in autism (SHANK1/2/3) and schizophrenia.

Inhibitory (GABAergic) postsynaptic specialisation. Located on soma, dendrite shaft, or axon initial segment. Gephyrin anchors GABA-A receptors (Cl⁻ channel, fast inhibition) and GABA-B receptors (GIRK/K⁺ channel + ↓cAMP, slow inhibition). No dendritic spines; no PSD-95.

Function — SNARE Vesicle Fusion Cycle

The complete cycle of vesicle exocytosis and recycling at the active zone:

VESICLE CYCLE AT THE ACTIVE ZONE 1. DOCKING (Rab3-RIM interaction, syntaxin/Munc18 complex) SV docks at active zone within 50-100 nm of P/Q-type Ca2+ channels 2. PRIMING (Munc13 opens syntaxin; SNARE zippering begins) Synaptobrevin-2 (v-SNARE on SV) + Syntaxin-1A + SNAP-25 (t-SNAREs on plasma membrane) --> Partial SNARE bundle (N-terminal; ~4 layers zipped) --> Vesicle in readily releasable pool (RRP) 3. Ca2+ TRIGGERING AP --> P/Q-type Ca2+ channels open --> [Ca2+] spikes to ~10-100 uM at AZ microdomains --> Synaptotagmin-1 C2A + C2B domains bind 3-5 Ca2+ ions (affinity ~100 uM) --> Ca2+-bound syt1 DISPLACES complexin inhibitory clamp --> Accelerates SNARE C-terminal zippering to completion (~0.2 ms) --> Bilayer merger and pore opening 4. EXOCYTOSIS ~5,000-10,000 NT molecules flood 20 nm cleft within 0.1 ms NT binds postsynaptic ionotropic receptors within 0.2 ms 5. CLEARANCE Glutamate: EAAT1/2 (GLT-1, GLAST) on astrocyte perisynaptic processes -- >90% uptake GABA: GAT-1 on presynaptic terminal; GAT-3 on astrocyte 6. ENDOCYTOSIS + RELOADING Full-collapse fusion: clathrin/dynamin/AP2 --> endosome --> SV budding (adaptin, synaptojanin) Kiss-and-run: pore closure, SV retains identity SV refilled: VGluT1/2 (glutamate) or VGAT (GABA) via H+-electrochemical gradient (V-ATPase)

Total time from AP arrival to postsynaptic PSP peak: 1–5 ms for fast synapses. The RRP is depleted within ~10 APs at 100 Hz; recycling pool sustains transmission during sustained activity.

Plasticity — LTP and LTD

Long-Term Potentiation (LTP). The canonical form at CA1 Schaffer collateral synapses (first demonstrated by Bliss and Lømo, 1973):

LTP INDUCTION (E-LTP, within minutes): High-frequency stimulation (HFS, 100 Hz, 1 s) or theta burst stimulation --> AMPAR-mediated depolarisation of postsynaptic membrane --> NMDAR Mg2+ block relieved at depolarised potentials --> Ca2+ influx through NMDAR (~10-50 uM in spine) --> Ca2+/CaM --> CaMKII activation (T286 autophosphorylation) --> GluA1 Ser831 phosphorylation --> AMPAR conductance increase --> Rab11/NSF-dependent AMPAR exocytosis to PSD surface --> Structural spine enlargement (F-actin remodelling via Rac1/Cofilin) LTP CONSOLIDATION (L-LTP, >1 h, transcription-dependent): CaMKII --> MAPK/ERK --> CREB Ser133 phosphorylation --> Arc (AMPAR endocytosis regulation; limits LTP magnitude) --> BDNF (TrkB --> MAPK/PI3K --> local mRNA translation in dendrites) --> CPEB --> polyadenylation of dendritic mRNAs (CaMKII-alpha, Arc, GluA1) --> New protein synthesis --> stable structural LTP (new spine heads; new synapses) LTD INDUCTION (at CA1): Low-frequency stimulation (LFS, 1 Hz, 900 pulses) --> Moderate Ca2+ via NMDAR (lower peak than LTP) --> Preferential activation of PP2B (calcineurin) and PP1 --> GluA2 Ser880 dephosphorylation --> GRIP release --> AMPAR endocytosis --> Spine shrinkage --> silent synapse or synapse elimination mGluR-LTD (Group I mGluR; Gq --> IP3/DAG; prominent in hippocampus): --> Local dendritic mRNA translation (CPEB, FMRP-regulated) --> AMPAR endocytosis; dysregulated in Fragile X (excess mGluR-LTD)

LTP and LTD represent bidirectional Hebbian plasticity — the molecular substrate of learning-related synaptic modification. Together they implement a sliding threshold (BCM theory) that maintains neurons in a dynamic range.

Pathology

Alzheimer's Disease — Early Synaptic Loss

Amyloid-β oligomers (soluble, not plaques) bind neurexin and PSD-95 → activate complement C1q tagging of synapses → microglial CR3 phagocytosis of dendritic spines. Soluble Aβ also activates NMDARs → Ca²⁺ overload → synaptic depression (chemical LTD). Synaptic density loss precedes neuron loss by years and correlates more strongly with cognitive decline than plaque or tangle burden.

Epilepsy — Runaway Excitation

E/I imbalance: ⇧glutamatergic release or ↓GABAergic inhibition → AMPAR/NMDAR hyperactivation → synchronised burst firing propagating as seizure. Ion channel mutations (SCN1A Nav1.1, KCNQ2/3, GABA-A subunits) and synaptic protein mutations (STXBP1/Munc18, PRRT2) cause genetic epilepsies via disruption of the synaptic release machinery or E/I balance.

Schizophrenia — NMDA Hypofunction

NMDA receptor hypofunction (particularly on PV⁺ interneurons → disinhibition of pyramidal cells) disrupts gamma synchrony and working memory encoding. Complement-mediated over-pruning of synapses during adolescent synaptic pruning (C4A overexpression, GWAS finding) is implicated in excess synapse elimination. Cognitive symptoms correlate with prefrontal synaptic density.

Autism Spectrum Disorder

Mutations in transsynaptic adhesion molecules (NRXN1, NLGN3/4) and PSD scaffold proteins (SHANK1/2/3, SYNGAP1, DLGAP1) disrupt synapse formation, maturation, and E/I balance. Many ASD genes converge on mTOR pathway → excess local dendritic protein synthesis → excessive mGluR-LTD → over-weakening of synapses (similar to Fragile X).

Fragile X Syndrome

Loss of FMRP (FMR1 gene; CGG repeat expansion) → disinhibition of mGluR-dependent local mRNA translation → excess AMPAR endocytosis → exaggerated mGluR-LTD → immature, thin dendritic spines → intellectual disability, autism features. mGluR antagonists (MPEP, fenobam) and lithium partially rescue the synaptic phenotype in animal models.

Tetanus Toxin (TeNT) — Inhibitory Synapse Block

TeNT cleaves VAMP-2/synaptobrevin on GABAergic and glycinergic interneuron terminals → blocks inhibitory neurotransmitter release → loss of inhibitory control of motor neurons → spastic paralysis (lock-jaw, opisthotonus). Botulinum toxins cleave VAMP, SNAP-25, or syntaxin at neuromuscular junctions → flaccid paralysis (opposite phenotype).

References

  • Südhof TC. Neurotransmitter release: the last millisecond in the life of a synaptic vesicle. Neuron. 2013;80(3):675–690. doi:10.1016/j.neuron.2013.10.022 · PubMed 24183019
  • Holtmaat A, Svoboda K. Experience-dependent structural synaptic plasticity in the mammalian brain. Nat Rev Neurosci. 2009;10(9):647–658. doi:10.1038/nrn2699 · PubMed 19693029
  • DeFelipe J, López-Cruz PL, Benavides-Piccione R, et al. New insights into the classification and nomenclature of cortical GABAergic interneurons. Nat Rev Neurosci. 2013;14(3):202–216. doi:10.1038/nrn3444 · PubMed 23406968
  • Kandel ER, Koester JD, Mack SH, Siegelbaum SA. Principles of Neural Science. 6th ed. McGraw-Hill; 2021. mhprofessional.com