Examining form and function of dendritic spines

Abstract

The majority of fast excitatory synaptic transmission in the central nervous system takes place at protrusions along dendrites called spines. Dendritic spines are highly heterogeneous, both morphologically and functionally. Not surprisingly, there has been much speculation and debate on the relationship between spine structure and function. The advent of multi-photon laser-scanning microscopy has greatly improved our ability to investigate the dynamic interplay between spine form and function. Regulated structural changes occur at spines undergoing plasticity, offering a mechanism to account for the well-described correlation between spine size and synapse strength. In turn, spine structure can influence the degree of biochemical and perhaps electrical compartmentalization at individual synapses. Here, we review the relationship between dendritic spine morphology, features of spine compartmentalization and synaptic plasticity. We highlight emerging molecular mechanisms that link structural and functional changes in spines during plasticity, and also consider circumstances that underscore some divergence from a tight structure-function coupling. Because of the intricate influence of spine structure on biochemical and electrical signalling, activity-dependent changes in spine morphology alone may thus contribute to the metaplastic potential of synapses. This possibility asserts a role for structural dynamics in neuronal information storage and aligns well with current computational models.

Figure 1

Biochemical compartmentalization in dendritic spines. The spine neck may offer enhanced compartmentalization of biochemical signalling at synapses. (a) The lateral mobility of surface glutamate receptors is attenuated across longer spine necks and at the postsynaptic density. (b) The spine neck imposes diffusional constraints on cytosolic signalling proteins. These mobility restraints imposed by the spine neck create spine-specific compartmentalization of cytosolic signalling and surface receptor dynamics.

Figure 2

Dissociation of spine size and synaptic strength. (a) The release of glutamate was reduced for 48 hours specifically on the spine marked (1). This leads to a homeostatic enhancement of synaptic strength, as assessed by 2P-uncaging of MNI-Glutamate. The size of the synaptic currents induced by 2P-uncaging is shown in the bottom panel. (b) The significant enhancement of the amplitude of synaptic currents onto “silenced” spines was not accompanied by any measurable changes in spine volume. Adapted from [15].