'Synaptic autocontrol'.

Abstract

Dale's principle, as stated by the English neuroscientist Henry Dale, posited that a presynaptic neuron will have the same chemical action on all its connections regardless of the class of target cell. Dale's statement was often interpreted as a single synapse releasing a single class of chemical messenger. There is now a vast body of experimental evidence that challenges Dale's principle. New observations reveal that we do not completely understand the complexity and extent of the mechanisms in place to control the activity of postsynaptic and presynaptic elements. Existing modelling strategies still do not take into account several of these new variables. To begin with, there are many synapses that cannot be reduced to either simple excitatory (glutamate) or inhibitory (GABA) inputs. For instance, it is now clear that, under some conditions, these two transmitters may be co-released by the same synapse (Gutiérrez, 2005). In addition, synapses in many brain regions co-release peptides; new work in this issue of The Journal of Physiology by Hawes et al. (2017) provides a unique perspective on peptide actions in the striatum. Most investigators implicitly assume that synaptic plasticity encoded as long-term potentiation (LTP) or long-term depression (LTD) is an ongoing feature in neuronal circuits and is a first step in the formation of functionally connected neuronal ensembles. By either reinforcing or weakening synapses, synaptic connections can be preferentially tuned to the flow of activity within these ensembles to produce specific activity patterns and reverberations. These circuits are shaped by synaptic changes that occur during learning and conform to procedural and declarative memories underlying goal-directed actions, routines and habits. Addiction, in some ways, falls into this category. Yet, we know that neuropeptides play key roles in many of the circuits implicated in addiction. What are the specific contributions of these peptides? Do they interact with activity-dependent plasticity mechanisms such as LTP/LTD? Hawes et al. (2017) provide a novel answer to this long-standing question. They show that corticostriatal LTP is modulated by dynorphin. Dynorphin is one of the peptides co-released by direct pathway striatal projection neurons (dSPNs), which synthesize and release the fast channel-gating transmitter GABA. The other co-released peptide by dSPNs is substance P (SP). In contrast, indirect pathway neurons (iSPNs) release enkephalins. Using voltammetry Hawes et al. show that dynorphin decreases dopamine release via κ-opioid receptors (KORs), and using optogenetics they also show that a small fraction of channel rhodopsin transfected D1-Cre dSPNs are capable of releasing enough endogenous dynorphin to decrease dopamine release. This reduction in dopamine release results in a decrease in corticostriatal LTP in neighbouring SPNs, suggesting that opioids modulate learning by controlling the number of synapses that develop LTP. This means that local release of peptides may tune circuits by impacting ongoing plasticity between specific synapses. One wonders if similar mechanisms can be found for LTD. Further in their Discussion, Hawes et al. reveal a potentially more complex scenario, that dSPNs also release SP which has opposite actions to dynorphin: increase in glutamate, acetylcholine and dopamine release. What is the possible outcome of releasing peptides with opposed actions? The authors’ speculation is interesting and experimentally testable: their spatio-temporal profiles may be different, SP acts briefly and before dynorphin, implying that LTP or LTD may be facilitated during specific temporal windows due to peptides released, perhaps interfering with spike-timing-dependent plasticity. This would determine the size of the neuronal ensembles that are formed, thus helping the subject to make hierarchical learning: augmenting or filtering different inputs at a given moment depending on context. In the case of dynorphin, this property may serve to control ensemble activation during addictive behaviour, limiting a runaway enhancement of striatal activity in response to drugs of abuse. Clearly, this mechanism can be explored as a therapeutic target for treating addiction. More generally, the field should not only investigate whether peptides released by synapses facilitate or restrain synaptic transmission, but also direct efforts towards whether additional processes such as synaptic plasticity are being modified. None declared. CONACyT Frontera-57, DGAPA-UNAM IN205610.