DIVERSE NEURONAL POPULATIONS MEDIATE LOCAL CIRCUIT EXCITATION IN AREA CA3 OF DEVELOPING HIPPOCAMPUS

1. Studies were undertaken to better understand why the developing hippocampus has a marked capacity to generate prolonged synchronized discharges when exposed to gamma-aminobutyric acid-A (GABA(A)) receptor antagonists. 2. Excitatory synaptic interactions were studied in small microdissected segments of hippocampal area CA3. Slices were obtained from 10-to 16-day-old rats. Application of the GABA(A) receptor antagonist penicillin produced prolonged synchronized discharges in minislices that were very similar, if not identical, to those recorded in intact slices. The sizes of minislices were systematically varied. Greater than 90% of those that measured 600 mu m along the cell body layer produced prolonged synchronized discharges, whereas most minislices measuring 300 mu m produced only brief interictal spikes. 3. Action potentials in the majority (75%, 158 of 254) of cells impaled with microelectrodes were able to entrain the entire CA3 population. They were also able to increase (on average 26%) the frequency of spontaneous population discharges. The population discharges were followed by a refractory period that lasted 5-60 s, during which single cells were unable to initiate a population discharge. 4. The majority (87%) of neurons with intrinsic burst properties were found to entrain the CA3 population. The electrophysiological characteristics of these cells were reminiscent of recordings obtained from more mature rats. Action potentials were quite prolonged and demonstrated a secondary shoulder or hump on the down-slope of the spike. 5. When bursting cells were filled with Lucifer yellow and imaged during recording sessions by videomicroscopy and later using confocal microscopy, they showed the anatomic features of CA3 hippocampal pyramidal cells. Confocal microscopy permitted detailed characterization of individual neurons and showed substantial variation in cellular microanatomy. 6. Another class of cells that were found to entrain the CA3 population but did not demonstrate intrinsic bursts were termed regular-firing cells. These cells possessed many of the anatomic and physiological features of bursting cells with the exception of burst firing. They were rarely encountered in intracellular recordings. 7. The third physiological class of cells was termed fast-spiking cells. These had action potentials that were shorter in duration than the other two cell types. They were distinct in the rapid rate of spike repolarization. They demonstrated modest degrees of spike frequency adaptation and fired repeatedly and at relatively high frequencies. Compared with reports on fast-spiking cells in mature hippocampus and neocortex, action potentials appear to be slower and repetitive discharging appeared to be of a lower frequency. 8. Anatomically, fast-spiking cells were classified as ''simple'' pyramidal cells or multipolar stellate cells. They had anatomic features of hippocampal interneurons, yet both cell types were able to entrain the CA3 population. 9. Paired intracellular recordings showed 25% of the cells that entrained the CA3 population produced excitatory postsynaptic potentials (EPSPs) in nearby cells. Indeed, in many instances the onset of synchronized discharging was preceded by EPSPs recorded in the second cell of a pair. 10. Of bursting cells, 25% produced EPSPs in nearby cells. Polysynaptic EPSPs, produced by two to four action potentials, were large (2.0 +/- 0.3 mV, mean +/- SD), with time to onset of 11.1 +/- 3.5 ms and time to peak amplitude of 30.1 +/- 9.9 ms. Their probability of occurrence was surprisingly high, on average 0.82. 11. Dye-coupling was observed in 65% of cases after intracellular injection of Lucifer yellow. As many as six neurons were recovered after a single injection. Dye-coupling was evident not only in confocal images but during recording sessions where cells were viewed with fluorescent videomicroscopy. In this regard, during dual recordings, 4% of cell pairs (6 of 149) were found to be electrotonically coupled. 12. Seventy-three percent of fast-spiking cells studied entrained the CA3 population. Thirteen percent produced EPSPs in nearby pyramidal cells. Compared with pyramidal cells, more presynaptic action potentials were required of fast-spiking cells to either entrain the CA3 population or generate a polysynaptic EPSP. Time to EPSP onset was longer (40-70 ms) than that recorded from bursting cells. 13. Results suggest that the recurrent excitatory networks in area CA3 are very well developed early in postnatal life. In addition to the pyramidal cell to pyramidal cell connections, a novel form of local circuit excitation may exist that is mediated by local circuit interneurons. Local circuit excitation likely contributes importantly to hippocampal hyperexcitability during this critical period in development.