THE IONIC MECHANISMS OF EXCITATORY AND INHIBITORY SYNAPTIC ACTION

There are two major classes of synapses that act by chemical transmission: the excitatory and inhibitory. Recently a third type of chemical synapse has been discovered-the presynaptic inhibitory synapse, but there is as yet no information on its ionic mechanism. Electron-microscopy is now revealing that the excitatory and inhibitory synapses are very similar in structure, though there is evidence of finer structural differences. In both, the presynaptic components have synaptic vesicles that are in part concentrated on active sites on the membrane fronting the narrow synaptic cleft (about 200 A across) separating the presynaptic and subsynaptic components of the synapse. Both these components display a specialization of structure revealed by membrane thickenings. Moreover, it is believed that excitatory and inhibitory synapses have a similar mode of operation. In both, the presynaptic impulse propagates up to the presynaptic terminal, as has now been conclusively demonstrated for the neuromuscular synapse (Katz & Miledi, 1965), and within a fraction of a millisecond causes the liberation of a jet of transmitter in a packaged release of so many quanta. This transmitter then diffuses across the synaptic cleft and acts on specific receptor sites on the subsynaptic membrane, causing transmembrane currents to flow. These subsynaptic currents may flow either inwards or outwards, so removing charge or adding charge to the whole postsynaptic membrane-according to whether the synapses are excitatory or inhibitory. Some years ago (Fatt & Katz, 1951; del Castillo & Katz, 1954; Katz, 19588; Coombs et al., 195%; Eccles, 1957) a simple hypothesis was proposed for the ionic mechanisms concerned in excitatory synaptic action. It was postulated that the excitatory transmitter makes the receptive patches of a subsynaptic membrane permeable to all diffusible ions on both sides of the subsynaptic membrane, which consequently run down their electrochemical gradients causing a virtual shortcircuit of the charge on the postsynaptic membrane. This explanation was in accord with the observed null-point (about 15 mV) for reversal of the excitatory subsynaptic currents, which is approximately the same as the calculated value for the liquid-junction potential between the interior of the postsynaptic cell and its external environment (Nastuk & Hodgkin, 1950). A complementary hypothesis was developed for inhibitory synaptic action (Coombs et al., 1953, 1955b; Eccles, 1957), the inhibitory transmitter being assumed to make the subsynaptic membrane permeable to small hydrated ions such as potassium or chloride, but not to the larger hydrated sodium ions. This hypothesis was in accord with the observed null-point for reversal, about 80 mV, approximately half way between the equilibrium potentials for chloride and potassium ions. It was also in accord with the changes that injection of small anions into the postsynaptic cell caused to occur in size and direction of the inhibitory subsynaptic currents; whereas anions that were larger than a critical size in the hydrated state were ineffectual. It was thus postulated that the essential difference between the subsynaptic membrane of excitatory and inhibitory synapses was that, under the action of the appropriate transmitter, the former developed holes or pores large enough to allow the passage of hydrated sodium ions, whereas with the