Greener on the other side

The idea that certain molecules could act as retrograde synaptic messengers has been en vogue for some time. Several molecules such as nitric oxide, carbon monoxide and arachidonic acid metabolites have been proposed to diffuse from the postsynaptic neuron to act on the presynaptic cell but the evidence hasn't always been incontrovertible. Three recent papers have put a new spin on the study of retrograde messengers by showing that postsynaptically released cannabinoids can inhibit transmitter release in hippocampal and cerebellar neurons.

When a hippocampal neuron is depolarized, inhibitory inputs to this cell become transiently suppressed, a phenomenon termed 'depolarization-induced suppression of inhibition' (DSI). Wilson and Nicoll focused on the mechanisms responsible for DSI in hippocampal slices and found that cannabinoids released from the postsynaptic cell transiently depress GABA-mediated transmission. The authors determined that, although cannabinoid release is Ca²⁺-dependent, it does not seem to involve vesicle fusion, as botulinum toxin fails to block DSI.

Similarly, Ohno-Shosaku et al. studied DSI in cultured hippocampal neurons and found that endogenous cannabinoids were involved in synaptic suppression. Moreover, these researchers also established that strong depolarization (a stimulus that is commonly used to induce DSI) is not necessary to elicit suppression, as action-potential firing by the postsynaptic neuron was enough to produce DSI.

This form of synaptic suppression had so far been observed only at inhibitory synapses. However, Kreitzer and Regehr found that an equivalent phenomenon termed DSE occurs at excitatory synapses in the cerebellum. Depolarization of Purkinje cells can suppress parallel- and climbing-fibre input, and this suppression also depends on the action of endogenous cannabinoids. Moreover, they showed that presynaptic Ca²⁺ influx is reduced during DSE, indicating that cannabinoids might exert their action through the inhibition of presynaptic Ca²⁺ channels. Whether a similar mechanism is also involved in DSI remains to be tested, but is a likely possibility.

DSI and DSE arguably represent the first known function of endogenous cannabinoids. In addition, these three papers provide compelling evidence for a role of cannabinoids as true retrograde messengers. But these results also give rise to many new questions. If vesicle fusion is not necessary for DSI, then how are postsynaptic depolarization and Ca²⁺ influx coupled to cannabinoid release? How do DSI, DSE and the release of endogenous cannabinoids affect hippocampal and cerebellar function? Do they interact with other forms of synaptic plasticity? Are DSE or DSI involved in the behavioural effect of marijuana? These three studies are merely the beginning of the trip.