Endocannabinoids are unconventional neurotransmitters, travelling from a postsynaptic to a presynaptic neuron — the opposite direction to typical chemical messengers. This signalling is known to be important for the survival, differentiation and function of the presynaptic neuron, but there is speculation about the synaptic site of production of endocannabinoids and how they elicit their responses. Katona and colleagues go some way to clarifying the molecular architecture of the endocannabinoid system at glutamatergic synapses, and also implicate this pathway in the plasticity of these synapses.
Thought to be produced postsynaptically in response to specific physiological stimuli, endocannabinoids have several potential cellular targets, two of which are the cannabinoid receptors CB1 and CB2. Of these receptors, so far only CB1 has been found at the presynapse. If endocannabinoids act on CB1 presynaptically, as has been proposed, these molecules are prime candidates as retrograde signals in synaptic plasticity. In the hippocampus, the predominant endocannabinoid thought to be involved in synaptic plasticity is the fatty acid derivative 2-arachidonoyl-glycerol (2-AG), the synthesis of which is thought to be mediated by the enzyme diacylglycerol lipase-
(DGL). The authors found that the highest expression levels of DGL were indeed in the hippocampus, and in particular at the heads of dendritic spines — sites of synaptic plasticity for glutamatergic synapses — of hippocampal principal cells.
...endocannabinoids...are prime candidates as retrograde signals in synaptic plasticity.
Dendritic spines are made up of discrete functional subunits that individually contribute to separate aspects of synaptic plasticity. Could the site of DGL expression be narrowed even further? High-resolution immunoelectron microscopy revealed that DGL production at glutamatergic synapses is highest in the perisynaptic annulus that surrounds the postsynaptic density, with the active site of DGL being located on the intracellular side of the membrane. Interestingly, the authors noted that the levels of DGL expression decreased gradually along the spine neck. This finding lends support to the theory that the narrow spine acts as a barrier to ensure synapse-specific plasticity mechanisms.
The localization of DGL on dendritic spine heads suggests that the receptors for 2-AG — that is, CB1 receptors — should be located in close proximity on the opposite side of the synapse. Indeed, immunocytochemical analyses revealed the localization of CB1 receptors to these specific axon terminals in wild-type mice but not in CB1-knockout mice. This precise positioning implicates 2-AG in retrograde signalling at glutamatergic synapses.
On the basis of these findings, the authors propose that CB1 receptors could control glutamate release presynaptically, and so be involved in homosynaptic plasticity of excitatory synapses and heterosynaptic plasticity between excitatory and inhibitory contacts. It will be interesting to see whether other biosynthetic routes to 2-AG exist at other synapses and, if they do, whether their molecular architecture matches that of glutamatergic synapses.