ROLE OF EPSPS IN INITIATION OF SPONTANEOUS SYNCHRONIZED BURST FIRING IN RAT HIPPOCAMPAL-NEURONS BATHED IN HIGH POTASSIUM

Spontaneous discharges that resemble interictal spikes arise in area CA3 b/c of rat hippocampal slices bathed in 8.5 mM [K+](o). Excitatory postsynaptic potentials (EPSPs) also appear at irregular intervals in these cells. The role of local synaptic excitation in burst initiation was examined with intracellular and extracellular recordings from CA3 pyramidal neurons. Most (70%) EPSPs were small (<2 mV in amplitude), suggesting that they were the product of quantal release or were evoked by a single presynaptic action potential in another cell. It is unlikely that most EPSPs were evoked by a presynaptic burst of action potentials. Indeed, intrinsic burst firing was not prominent in CA3 b/c pyramidal cells perfused in 8.5 mM [K+](o). The likelihood of occurrence and the amplitude of EPSPs were higher in the 50-ms interval just before the onset of each burst than during a similar interval 250 ms before the burst. This likely reflects increased firing probability of CA3 neurons as they emerge from the afterhyperpolarization (AHP) and conductance shunt associated with the previous burst. Perfusion with 2 μM 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX), a potent quisqualate receptor antagonist, decreased the frequency of EPSPs in CA3 b/c neurons from 3.6 ± 0.9 to 0.9 ± 0.3 (SE) Hz. Likewise, CNQX reversibly reduced the amplitude of evoked EPSPs in CA3 b/c cells. Spontaneous burst firing in 8.5 mM [K+](o) was abolished in 11 of 31 slices perfused with 2 μM CNQX. When bursts persisted in the presence of CNQX, their frequency was slowed to 76 ± 6% control, and the intensity of the burst waveform was reduced to 54 ± 10% control. In contrast, the N-methyl-D-asparate (NMDA) receptor antagonist, D-2-amino-5-phosphonovaleric acid (10 μM), had no effect on burst frequency. The amplitude of the paroxysmal depolarization (PDS) underlying the interictal burst was reduced by CNQX in a graded fashion to 24 ± 4% control, as expected if the PDS itself were a synaptic potential. The data support the hypothesis that EPSPs in CA3 pyramidal cells contribute to the emergence of spontaneous, synchronized bursts in high [K+](o). EPSPs might originate from mossy fiber synapses or, more likely given their appearance just before initiation of each interictal burst, synapses made by recurrent excitatory collaterals from neighboring CA3 cells. A simulation that incorporates the essential features of spontaneous burst firing in high [K+](o) supports this interpretation.