No abstract

No abstract

No abstract

Changes in the amplitudes of signals conveyed at synaptic contacts between neurons underlie many brain functions and pathologies. Here we review the possible determinants of the amplitude and plasticity of the elementary postsynaptic signal, the miniature. In the absence of a definite understanding of the molecular mechanism releasing transmitters, we investigated a possible alternative interpretation. Classically, both the quantal theory and the vesicle theory predict that the amount of transmitter producing a miniature is determined presynaptically prior to release and that rapid changes in miniature amplitude reflect essentially postsynaptic alterations. However, recent data indicates that short-term and long-lasting changes in miniature amplitude are in large part due to changes in the amount of transmitter in individual released packets that show no evidence of preformation. Current representations of transmitter release derive from basic properties of neuromuscular transmission and endocrine secretion. Reexamination of overlooked properties of these two systems indicate that the amplitude of miniatures may depend as much, if not more, on the Ca(2+) signals in the presynaptic terminal than on the number of postsynaptic receptors available or on vesicle's contents. Rapid recycling of transmitter and its possible adsorption at plasma and vesicle lumenal membrane surfaces suggest that exocytosis may reflect membrane traffic rather than actual transmitter release. This led us to reconsider the disregarded hypothesis introduced by Fatt and Katz (1952; J Physiol 117:109-128) that the excitability of the release site may account for the "quantal effect" in fast synaptic transmission. In this case, changes in excitability of release sites would contribute to the presynaptic quantal plasticity that is often recorded.

1. Hippocampal neurones cultured from the 18-day-old embryonic rat for 3 days to 3 weeks were recorded with Cl--filled patch pipettes. Spontaneous synaptic currents, which reversed at the equilibrium potential for Cl-ions (Ec1) and were blocked by the GABAA (y-aminobutyric acid) receptor antagonists bicuculline or picrotoxin, were recorded in every culture. At 25°C and -80 mV they decayed with a time constant > 20 ms that invariably increased at positive potentials. After 2 weeks, 50-75% of all neurones were GABA immunoreactive. 2. In pairs-recordings, coincident synaptic currents in both cells were either spontaneous or evoked by stimulation of one cell. In the presence of tetrodotoxin and using pipettes containing lidocaine (lignocaine) N-ethyl bromide, coincident spontaneous Cl-transients still occurred in both neurones far more frequently than expected by chance. 3. Holding the potential of one neurone at a positive value reversed the synaptic transients in that cell and, in half of the cells, increased the frequency of coincident events in both cells. 4. In neurones where depolarization increased the frequency of coinciding events and all regenerative current apparent at the soma was abolished, short depolarizing pulses occasionally evoked all-or-none, pre-and postsynaptic currents with matching transmission failures and identical delays in transmission. 5. The results suggest that the same pulse of GABA simultaneously activates GABAA receptor-coupled Cl-channels on both sides of the same synaptic cleft, producing immediate auto-transmission in the absence of collaterals or interneurones.