Synaptogenesis

Synaptogenesis is the construction of a working contact between two neurons. An axonal growth cone and a dendritic filopodium make contact, transsynaptic adhesion molecules such as the neurexins and neuroligins bridge the cleft, and the two sides then differentiate in parallel: an active zone with docked vesicles and calcium channels on one side, a postsynaptic density packed with scaffolding proteins and glutamate receptors on the other. In cortex it peaks during development and is followed by extensive activity-dependent pruning, so adult synaptic number reflects the balance of ongoing formation and elimination rather than a fixed endowment.

Measurement has improved substantially. Post-mortem work counts synaptophysin or electron-microscopic profiles; rodent studies use two-photon imaging to watch individual dendritic spines appear and disappear over days. In living humans, PET ligands targeting synaptic vesicle glycoprotein 2A now allow regional synaptic density to be estimated in vivo, which has confirmed a long-standing observation from autopsy series that synaptic loss tracks cognitive decline in Alzheimer's disease more closely than plaque burden does.

The interpretive limit is that density is not function. A synapse counted is not a synapse that is correctly targeted, appropriately weighted or usefully integrated into a circuit, and PET gives a regional average that conceals which connections changed. More synapses can be worse, as pruning failure in neurodevelopmental conditions illustrates.

Where claims break down is in the leap from a culture dish to cognition. Counting puncta on cultured neurons after exposure to a compound is a screening assay; it is run at concentrations that need not be achievable in brain, on immature neurons without the surrounding circuit, and it says nothing about whether any behaviour changed.