Fig. 4

Factors regulating initiation and maintenance of reverberating activity within the CA3 network. a Schematic shows initial network states (30 and 70% hyperpolarized) and the distribution of simulated pyramidal cells (PCs) with relatively depolarized (red) or hyperpolarized (blue) membrane potentials. After initial hyperpolarization, 70% of PCs were randomly selected, every 50 ms, to receive 50-ms injections of hyperpolarizing current for the duration of the run. Note, some PCs (triangles) do not change membrane potential with any given transition between states. b Exemplar sections (2 s) of simulated (top) and current clamp (Iclamp; bottom) recordings from a CA3 PC illustrating similar amplitude fluctuations in membrane potential during periods without action potentials. Bar graph summarizes the mean (±SEM) maximum change in membrane potential for simulated (n = 6), clamp-recorded CA3 (n = 6) and CA1 (n = 7) PCs. Scale bars: y = 10 mV, x = 200 ms. c Representative sections (4 s) from 4 simulated PCs demonstrate that prolonged firing was only observed when the network was in the 30% hyperpolarized state at time of stimulation (Cue). Scale bars: y = 2 mV, x = 500 ms. d Graphs summarizing how probability (P) of prolonged firing (top) and duration of PC spiking (bottom) are influenced by the magnitude of hyperpolarizing currents in the two initial network states. e The hippocampal simulation displays catastrophic collapse following a small (~15–20%) reduction in excitatory transmission in CA3. f Representative CA3 fEPSPs (left) and traces of PC spiking (right) in the absence (black) and presence (red) of kynurenic acid (KYN; 500 µM). The sharp negative deflection is the antidromic potential. Scale bars: fEPSP y = 1 mV, x = 10 ms; PC spiking y = 100μV, x = 200 ms. g Graph summarizing the unequal effects of KYN upon fEPSP amplitude and PC spike frequency (n = 4): **p = 0.0014 for effect on fEPSPs; ***p = 0.0002 for effect on PC spiking; ***p = 0.0007 for difference in KYN effect on fEPSP versus spiking. h Time course of KYN effects upon fEPSP amplitude and PC spiking. i Graph of the reduction of PC firing versus the decrease in fEPSP size during KYN infusion illustrates the catastrophic collapse in CA3 activity; as in the simulation, a <20% reduction in excitatory transmission caused a drastic reduction in PC spiking (n = 4)