A SIMULATION OF ACTION-POTENTIALS IN SYNAPTIC BOUTONS DURING PRESYNAPTIC INHIBITION

1. During presynaptic inhibition, an increased conductance in the membrane of the presynaptic bouton is presumed to reduce the action potential, thereby reducing transmitter release. The object of the simulation has been to determine the magnitude of a chloride conductance required to reduce transmitter release, for various diameters of synaptic boutons, connected to axons with diameters in the range 0.1-1.0 mu m. 2. A propagating action potential was simulated in axons connected to either side of a hemispherical bouton. The axons could be myelinated or unmyelinated, while the bouton membrane could be passive, a node of the myelinated nerve, or have the same active properties as the attached unmyelinated nerve. Membrane properties of the axons were derived from mammalian data and scaled to 37 degrees C. 3. A steady-state chloride conductance was included in the bouton membrane, with E(cl) = -40 mV. The amplitude of the action potential in the bouton was calculated for different diameters of axon and bouton and for different magnitudes of chloride conductance. 4. Using published data on the relationship between the amplitude of a presynaptic action potential and the resulting postsynaptic potential, the relationship between the chloride conductance and the postsynaptic response was calculated for different geometries. Transmitter release was reduced when an action potential was 90 mV or smaller, with no transmission for action potentials smaller than 50 mV. 5. Conductance increases in the range 3 to 10 nS were required to reduce the action potential to 90 mV, depending on the diameter of the axon (0.5-1.0 mu m), diameter of the bouton (3-6 mu m), whether the bouton had passive or active membrane, and whether the axon was myelinated or unmyelinated. A 3 mu m passive bouton connected to a 0.5 mu m myelinated axon was most sensitive to the effects of a chloride conductance, while a 6 mu m active bouton connected to a 1 mu m myelinated nerve was least sensitive to the effects of a chloride conductance. 6. The reduction in the action potential was compared when E(cl) = -40 mV and when E(cl) = E(rest) = -80 mV. inactivation of the sodium conductance by terminal depolarization was the dominant influence on the amplitude of the action potential. 7. Conductances that were sufficient to completely block synaptic transmission at a bouton were insufficient to prevent the spread of the action potential away from that bouton. 8. Schemes involving three boutons en passant, or three boutons terminating an axon, with the boutons linked by small diameter(0.l-1.0 mu m) axons of length 10 mu m, required conductances in the range 200 pS-3 nS on all three boutons to reduce the action potential to 90 mV. 9. These calculations are integrated with the quantal conductance for gamma-aminobutyric acid (GABA), and the convergence of axoaxonic contacts onto presynaptic terminals to determine whetherthe conductance increases required for presynaptic inhibition are likely to occur. It is suggested that it will be difficult to achieve a sufficiently large chloride conductance to make a significant reduction in transmitter release. However, the depolarization associated with the chloride conductance may have a direct inactivating action on high threshold calcium channels in the terminal membrane, thereby contributing to presynaptic inhibition.