Activity-dependent regulation of GABA release at immature mossy fibers-CA3 synapses: role of the Prion protein

In adulthood, mossy fibers (MFs), the axons of granule cells of the dentate gyrus (DG), release glutamate onto CA3 principal cells and interneurons. In contrast, during the first week of postnatal life MFs release -aminobutyric acid (GABA), which, at this early developmental stage exerts a depolarizing and excitatory action on targeted cells. The depolarizing action of GABA opens voltage-dependent calcium channels and NMDA receptors leading to calcium entry and activation of intracellular signaling pathways involved in several developmental processes, thus contributing to the refinement of neuronal connections and to the establishment of adult neuronal circuits. The release of GABA has been shown to be down regulated by several neurotransmitter receptors which would limit the enhanced excitability caused by the excitatory action of GABA. It is worth noting that the immature hippocampus exhibits spontaneous correlated activity, the so called giant depolarizing potentials or GDPs that act as coincident detector signals for enhancing synaptic activity, thus contributing to several developmental processes including synaptogenesis. GDPs render the immature hippocampus more prone to seizures. Here, I explored the molecular mechanisms underlying synaptic transmission and activity-dependent synaptic plasticity processes at immature GABAergic MF-CA3 synapses in wild-type rodents and in mice lacking the prion protein (Prnp0/0 mice). In the first paper, I studied the functional role of kainate receptors (KARs) in regulating GABA release from MF terminals. Presynaptic KARs regulate synaptic transmission in several brain areas and play a central role in modulating glutamate release at adult MF-CA3 synapses. I found that functional presynaptic GluK1 receptors are present on MF terminals where they down regulate GABA release. Thus, application of DNQX or UBP 302, a selective antagonist for GluK1 receptors, strongly increased the amplitude of MF-GABAA-mediated postsynaptic currents (GPSCs). This effect was associated with a decrease in failure rate and increase in PPR, indicating a presynaptic type of action. GluK1 receptors were found to be tonically activated by glutamate present in the extracellular space, since decreasing the extracellular concentration of glutamate with a glutamate scavenger system prevented their activation and mimicked the effects of KAR antagonists. The depressant effect of GluK1 on GABA release was dependent on pertussis toxin (PTx)-sensitive G protein-coupled kainate receptors since it was prevented when hippocampal slices were incubated in the presence of a solution containing PTx. This effect was presynaptic since application of UBP 302 to cells patched with an intracellular solution containing GDP S still potentiated synaptic responses. In addition, the depressant effect of GluK1 on GABA release was prevented by U73122, which selectively inhibits phospholipase C, downstream to G protein activation. Interestingly, U73122, enhanced the probability of GABA release, thus unveiling the ionotropic type of action of kainate receptors. In line with this, we found that GluK1 receptors enhanced MF excitability by directly depolarizing MF terminals via calcium-permeable cation channels. We also explored the possible involvement of GluK1 in spike time-dependent (STD) plasticity and we found that GluK1 dynamically regulate the direction of STD-plasticity, since the pharmacological block of this receptor shifted spike-time dependent potentiation into depression. The mechanisms underlying STD-LTD at immature MF-CA3 synapses have been investigated in detail in the second paper. STD-plasticity is a Hebbian form of learning which consists in bi-directional modifications of synaptic strength according to the temporal order of pre and postsynaptic spiking. Interestingly, we found that at immature mossy fibers (MF)-CA3 synapses, STD-LTD occurs regardless of the temporal order of stimulation (pre versus post or viceversa). However, as already mentioned, while STD-LTD induced by positive pairing (pre before post) could be shifted into STD-LTP after blocking presynaptic GluK1 receptors, STD-LTD induced by negative pairing (post before pre) relied on the activation of CB1 receptors. At P3 but not at P21, endocannabinoids released by the postsynaptic cell during spiking-induced membrane depolarization retrogradely activated CB1 receptors, probably expressed on MF terminals and persistently depressed GABA release in the rat hippocampus. Thus, bath application of selective CB1 receptor antagonists prevented STD-LTD. Pharmacological tools allow identifying anandamide as the endogenous ligand responsible of activity-dependent depressant effect. To further assess whether STD-LTD is dependent on the activation of CB1 receptors, similar experiments were performed on WT-littermates and CB1-KO mice. While in WT mice the pairing protocol produced a persistent depression of MF-GPSCs as in rats, in CB1-KO mice failed to induce LTD. Consistent with these data, in situ hybridization experiments revealed detectable levels of CB1 mRNA in the granule cell layer of P3 but not of P21mice. These experiments strongly suggest that at immature MF-CA3 synapses STD-LTD is mediated by CB1 receptors, probably transiently expressed, during a critical time window, on MF terminals. In the third paper, I studied synaptic transmission and activity dependent synaptic plasticity at immature MF-CA3 synapses in mice devoid of the prion protein (Prnp0/0). The prion protein (PrPC) is a conserved glycoprotein widely expressed in the brain and involved in several neuronal processes including neurotransmission. If converted to a conformationally altered form, PrPSc can cause neurodegenerative diseases, such as Creutzfeldt-Jakob disease in humans. Previous studies aimed at characterizing Prnp0/0 mice have revealed only mild behavioral changes, including an impaired spatial learning, accompanied by electrophysiological and biochemical alterations. Interestingly, PrPC is developmentally regulated and in the hippocampus its expression parallels the maturation of MF. Here, we tested the hypothesis that at immature (P3-P7) MF-CA3 synapses, PrPC interferes with synaptic plasticity processes. To this aim, the rising phase of Giant Depolarizing Potentials (GDPs), a hallmark of developmental networks, was used to stimulate granule cells in the dentate gyrus in such a way that GDPs were coincident with afferent inputs. In WT animals, the pairing procedure induced a persistent increase in amplitude of MF-GPSCs. In contrast, in Prnp0/0 mice, the same protocol produced a long-term depression (LTD). LTP was postsynaptic in origin and required the activation of cAMP-dependent PKA signaling while LTD was presynaptic and was reliant on G protein-coupled GluK1 receptor and protein lipase C downstream to G protein activation. In addition, at emerging CA3-CA1 synapses of PrPC-deficient mice, stimulation of Schaffer collateral failed to induce LTP, known to be PKA-dependent. Finally, we also found that LTD in Prnp0/0 mice was mediated by GluK1 receptors, since UBP 302 blocked its induction. These data suggest that in the immature hippocampus PrPC controls the direction of synaptic plasticity.