AXON TERMINAL HYPEREXCITABILITY ASSOCIATED WITH EPILEPTOGENESIS IN-VITRO .1. ORIGIN OF ECTOPIC SPIKES

1. Intracellular and extracellular recording techniques were used to study the increase in ectopic (i.e., nonsomatic) action-potential generation occurring among CA3 pyramidal cells during the kindling-like induction of electrographic seizures (EGSs) in this subpopulation of the hippocampal slice. Kindling-like stimulus trains (60 Hz, 2 s) were delivered to s. radiatum of CA3 at 10-min intervals. As EGSs developed, the frequency of ectopic firing increased markedly (by 10.33 +/- 3.29 spikes/min, mean +/- SE, P much less than 0.01). Several methods were applied to determine the initiation site for these action potentials within the cell (axons vs. dendrites). 2. Collision tests were conducted between known antidromic and orthodromic action potentials in CA3 cells to determine the critical period, c, for collision. Attempts were then made to collide ectopic spikes with known antidromic action potentials. At intervals less than c, ectopic spikes failed to collide with antidromic ones, in 5 of 10 cases. In these cells, this clearly indicates that the ectopic spikes were themselves of axonal origin. In the remaining five cases, ectopic spikes collided with antidromic action potentials at intervals approximately equal to c, most likely because of interactions within the complex system of recurrent axon collaterals in CA3. 3. Action potentials of CA3 pyramidal cells were simulated with the use of a compartmental computer model, NEURON. These simulations were based on prior models of CA3 pyramidal neurons and of the motoneuron action potential. Simulated action potentials generated in axonal compartments possessed a prominent inflection on their rising phase (IS-SD break), which was difficult to appreciate in those spikes generated in somatic or dendritic compartments. 4. An analysis of action potentials recorded experimentally from CA3 pyramidal cells also showed that antidromic spikes possess a prominent IS-SD break that is not present in orthodromic spikes. In addition to identified antidromic action potentials, ectopic spikes also possess such an inflection. Together with the predictions of computer simulations, this analysis also indicates that ectopic spikes originate in the axons of CA3 cells. 5. Tetrodotoxin (TTX, 50 muM) was locally applied by pressure injection while monitoring ectopic spike activity. Localized application of TTX to regions of the slice that could include the axons but not the dendrites of recorded cells abolished or markedly reduced the frequency of ectopic spikes (n = 5), further confirming the hypothesis that these action potentials arise from CA3 axons. 6. Changes in the excitability of the Schaffer collateral axons of CA3 cells were estimated by measuring the stimulus threshold required to directly evoke an antidromic action potential. Measurements of this threshold were made before and after EGS induction, at two separate sites along the course of the Schaffer collaterals: in the nonterminal region passing through s. radiatum of area CA2, and in the region of the synaptic terminals in s. radiatum of CA1. The antidromic stimulus threshold in the synaptic terminal region was decreased after EGS induction (mean change - 103.75 +/- 27.49 muA), but no significant change was detected in the nonterminal region. 7. Our data indicate that the kindling-like induction of electrographic seizures in the hippocampal slice is accompanied by a lasting increase in the excitability of CA3 axon terminals, which is manifested by an increase in action-potential initiation at this site. The significance of these findings is discussed with regard to possible mechanisms underlying the hyperexcitability and the functional role of this unusual form of neural plasticity in epileptogenesis and seizure expression. In the companion paper, we present pharmacological evidence that strengthens the link between axon terminal hyperexcitability and the expression of electrographic seizures, and we further address its underlying mechanisms.