Bloodgood and Sabatini were “playing around” with a sensitive and photoactivatable fluorophore when they noticed that some dendritic spines seemed to be decoupled from the rest of the dendrite. Fluorophore activated inside these spines was slow to leak out, and fluorophore activated outside was slow to diffuse in. Large changes in this diffusion barrier occurred spontaneously in organotypic slice cultures, with more blockage being induced by drugs that favor excitatory transmission.
Sure enough, diffusion restriction was induced in individual dendrites by pairing two excitatory signals from pre-synaptic and post-synaptic sources.The clampdown at the dendrite neck may involve cross-linking of an actin mesh, or blockage by mitochondria or smooth endoplasmic reticulum. Either way, says Sabatini, “if a spine can hold onto [molecules such as] active kinases, this will have a big impact on signal integration.” An isolated spine may be able to retain activated messengers until the next electrical spike arrives to boost the signal further. And truly isolated spines may also be able to act as independent electrical units that store or boost sub-spike electrical inputs.