Nonlinear systems analysis of the hippocampal perforant path-dentate projection. III. Comparison of random train and paired impulse stimulation.

1. The transformational properties of the network of hippocampal neurons activated monosynaptically and polysynaptically by electrical stimulation of the perforant path were analyzed using random impulse train and paired impulse stimuli. In response to both types of input, the amplitudes of granule cell population spikes evoked in the dentate gyrus were used as the measure of network output. The random stimulus train consisted of a series of 4,064 electrical impulses, with interimpulse intervals determined by a Poisson distribution; the mean interimpulse interval of the train was 500 ms, and the range was 1-5,000 ms. Paired impulse stimuli consisted of pairs of impulses separated by 10-1,200 ms; impulses pairs were delivered once every 20 s. The procedures were applied to both anesthetized and chronically implanted, unanesthetized preparations. 2. Nonlinear systems analysis of population spike responses evoked during random train stimulation revealed that dentate granule cell output to any impulse was highly dependent on the interval since a prior impulse. Data from anesthetized animals showed that population spike amplitudes were markedly suppressed in response to intervals less than 50 ms, facilitated in response to intervals of approximately 100 ms, suppressed slightly in response to intervals of 300-700 ms, and unaffected by intervals greater than 700 ms. Data from unanesthetized animals showed similar results except that facilitation rather than suppression of spike amplitude was observed in response to intervals of 300-700 ms, and could extend to intervals as great as 1,000 ms. 3. The results of paired impulse stimulation applied to the same preparations also showed that granule cell response was highly dependent on interimpulse interval. However, nonlinearities observed with paired impulse stimulation differed from those revealed by a random impulse signal. Compared to results of random train stimulation, a paired impulse format produced greater magnitude spike suppression in response to short interimpulse intervals (e.g., 10-20 ms), maximum facilitation in response to shorter interstimulus intervals (50 ms rather than 100 ms), greater magnitude spike facilitation, and greater suppression in response to intervals greater than or equal to 300 ms. Furthermore, there were virtually no differences in the nonlinearities of granule cell response recorded from anesthetized and unanesthetized animals when a paired impulse format was used, whereas several differences were observed with random train stimuli. 4.(ABSTRACT TRUNCATED AT 400 WORDS)