Lukacs, I. | bioRxiv Science

bioRxiv2024

Synaptic targets and cellular sources of CB1 cannabinoid receptor and vesicular glutamate transporter-3 expressing nerve terminals in relation to GABAergic neurons in the human cerebral cortex

Cannabinoid receptor 1 (CB1) regulates synaptic transmission through presynaptic receptors in nerve terminals, and its physiological roles are of clinical relevance. The cellular sources and synaptic targets of CB1-expressing terminals in the human cerebral cortex are undefined. We demonstrate a variable laminar pattern of CB1-immunorective axons and electron microscopically show that CB1-positive GABAergic terminals make type-2 synapses innervating dendritic shafts (69%), dendritic spines (20%) and somata (11%) in neocortical layers 2-3. Of the CB1-immunopositive GABAergic terminals, 25% were vesicular-glutamate-transporter-3 (VGLUT3)-immunoreactive, suggesting GABAergic/glutamatergic co-transmission on dendritic shafts. In vitro recorded and labelled VGLUT3 or CB1-positive GABAergic interneurons expressed cholecystokinin, vasoactive-intestinal-polypeptide and calretinin, had diverse firing, axons and dendrites, and included rosehip, neurogliaform and basket cells, but not double bouquet or axo-axonic cells. CB1-positive interneurons innervated pyramidal cells and GABAergic interneurons. Most glutamatergic synaptic terminals formed type-1 synapses and some were positive for CB1 receptor concentrated in the presynaptic active zone, unlike in GABAergic terminals. From the sampled VGLUT3-positive terminals, 60% formed type-1 synapses with dendritic spines (80%) or shafts (20%) and 52% were also positive for VGLUT1, suggesting intracortical origin. Some VGLUT3-positive terminals were immunopositive for vesicular-monoamine-transporter-2, suggesting 5-HT/glutamate co-transmission. Overall, the results show that CB1 regulates GABA release mainly to dendritic shafts of both pyramidal cells and interneurons, and predict CB1-regulated co-release of GABA and glutamate from single cortical interneurons. We also demonstrate the co-existence of multiple vesicular glutamate transporters in a select population of terminals probably originating from cortical neurons and innervating dendritic spines in the human cerebral cortex.

neuroscience↗

bioRxiv2019

An Emergent Neural Coactivity Code for Dynamic Memory

Coincidental spike discharge amongst distributed groups of neurons is thought to provide an efficient mechanism for encoding percepts, actions and cognitive processes1-3. Short timescale coactivity can indeed bind neurons with similar tuning, giving rise to robust representations congruent with those of the participating neurons4-6. Alternatively, coactivity may also play a role in information processing through encoding variables not represented by individual neurons. While this type of emergent coactivity-based coding has been described for physically well-defined variables, including percepts and actions7-10, its role in encoding abstract cognitive variables remains unknown. Coactivity-based representation could provide a flexible code in dynamic environments, where animals must regularly learn short-lived behavioural contingencies. Here, we tested this possibility by training mice to discriminate two new behavioural contingencies every day, while monitoring and manipulating neural ensembles in the hippocampal CA1. We found that, while the spiking of neurons within their place fields is organised into congruent coactivity patterns representing discrete locations during unsupervised exploration of the learning enclosure, additional neurons synchronised their activity into spatially-untuned patterns that discriminated opposing learning contingencies. This contingency discrimination was an emergent property of millisecond timescale coactivity rather than the tuning of individual neurons, and predicted trial-by-trial memory performance. Moreover, optogenetic suppression of plastic inputs from the upstream left CA3 region during learning selectively impaired the computation of contingency-discriminating, but not space-representing CA1 coactivity patterns. This manipulation, but not silencing the more stable right CA3 inputs, impaired memory of the contingency discrimination. Thus, the computation of an emergent, coactivity-based discrimination code necessitates plastic synapses and supports dynamic, two-contingency memory.

neuroscience↗